bSfeoRKBOTANIC^^S CATALOGUE FOSSIL PLANTS GLOSSOPTERIS FLORA DEPARTMENT OF GEOLOGY, CATALOGUE OF THE FOSSIL PLANTS OF THE GLOSSOPTERIS FLORA IN THE DEPARTMENT OF GEOLOGY LIBRARY BRITISH MUSEUM BOTANICAL (NATURAL HISTORY). GARDEN MONOGRAPH OF THE PERMO-CARBONIFEROUS FLORA OF INDIA AND THE SOUTHERN HEMISPHERE. BY E l . A. NEWELL ARBER, Ml, F.L.8., F.G.S., TKINITV COLLEGE, CAMBRIDGE, UNIVERSITY DEMONSTRATOR IX PALEOBOTANY. LONDON: PRINTED BY ORDER OF THE TRUSTEES. SOLD BY LONGMANS AND CO., 39, PATERNOSTER ROW. B. QUAHITCH, 15, PICCADILLY. DULAU AND CO., 37, SOHO SQUARE, W. KEGAN PAUL AND CO., 43, GERKARD STREET, SOHO, W. AND AT THE BRITISH MUSEUM (NATURAL HISTORY), CROMWELL ROAD, 8.W. {All rights reserved. mo. 5. I-O" .87 HERTFORD STEPHEN AUSTIN AND SONS, LTD. PRINTERS. PREFACE Since the publication of Mr. Kidston's Catalogue of the Palaeozoic Plants in 1886, extensive additions have been made to the collection in the British Museum, and much research has been devoted to these fossils. The time seems therefore to have arrived for a series of more detailed Catalogues. The Glossopteris Flora has proved to be especially important, and it is not only of great geological and botanical interest, but also of practical value to the prospectors for coalfields in India and the southern hemi- sphere. As this Flora has not hitherto been the subject of any comprehensive treatise, it is thus appropriately dealt with in the present volume, which is the first special Catalogue of Palaeozoic Plants issued by the British Museum. The work has been prepared with much care by Mr. E. A. Newell Arbor, who has been able to add somewhat to the existing knowledge, which he has critically summarised. A. SMITH WOODWARD. Department of Geology. October, 1905. CD CD en AUTHOR'S PREFACE. In this Catalogue I have made an attempt towards a complete summary of what is at present known on the subject of the Glossopteris Flora. It is hoped that reference will here be found to all the more important memoirs which have been published ; a literature now not only large, but widely scattered. I have endeavoured to give some account of the less abundant, and consequently the less known species, in addition to the more characteristic types represented in the British Museum collection, and to revise, so far as possible, the older records in the light of our present knowledge. The task has not proved a light one. In many instances the rarer fossils are unrepresented in any European collection, and I have had to rely solely on the original descriptions and figures, both sometimes very inadequate, which have been published. I trust that, whatever imperfections this volume may be found to contain, it may be of use to those who are interested in the study of this ancient flora, especially to residents in our Greater Colonies of India, Australia, and South Africa, which at one time formed the principal provinces of a great southern continent, Grondwana- land, whose flora is the subject of the present work. I have to express my acknowledgments to many friends to whom I have turned for information or advice at different times during the preparation of this volume. To Mr. A. C. Seward, F.R.S., I am especially indebted, not only for the information which he has on all occasions placed at my disposal, but for much kindly sympathy and encouragement. AUTHOR S PREFACE. To Dr. I). II. Scott, F.R.S., Professor F. W. Oliver, F.R.S., and Professor Zeiller of Paris, I would express my sincere thanks for valuable suggestions, or other aids. Mr. C. D. Sherborn has rendered me great services in connection with the literature on this flora, and to him I would return my thanks. I am similarly indebted to Mr. B. B. Woodward for assistance in the study of the periodical literature under his charge at the Natural History Museum. My thanks are also due to Mr. Holland, F.R.S., Director of the Geological Survey of India, for particulars of the specimens in the Calcutta Museum, and to Mr. W. S. Dun, of the Geological Survey of New South Wales, for valuable information on the subject of the Australian fossils. To Mr. A. G. Hamilton, of New South Wales, I am indebted for the loan of a collection of sections of petrified woods from that colony for comparison with those here described. The Council of the Geological Society of London have kindly given me permission to reproduce Text-figs. 12-15 from my recent paper in the Quarterly Journal, and to them I would express my thanks. For many of the other text- figures I have drawn largely on the illustrations published by the late Dr. Ottokar Feistmantel, the source in each case being duly acknowledged. Lastly, I have to record my thanks to Dr. Smith Woodward, F.R.S., Keeper of the Geological Department, British Museum, for valuable suggestions, and to the Staff of that Departmenl for the kindness with which they have always facilitated my work at the Museum. I would also express my thanks to Miss G. M. Woodward for drawing the illustrations included in this volume. E. A. NEWELL ARBER. Trinity College, Camhkidge. July, 1905. NOTE. The numbers in brackets after the Authors' names in the footnotes refer to the year of publication of the memoirs quoted. Full references to all the works cited will be found in the Bibliography at the end of the volume. LIST OF FIGURES IX THE TEXT. PAGE Map of the Distribution of the Northern and Southern Types of Perm o- Carboniferous floras xix Fig. 1-4. ScMzoneura gondwanensis, Feist 6-9 5. \_A.etinopteris\ bengedensis, Feist 15 6. Plnjllotheca indica, Bunb 21 7. P. deliquescens (Gopp.) 23 8. P. robusta, Feist 26 9. P. Ether idgei, sp. nov 27 10. \_Annularia (?)] australis, Feist 31 11. Sphenophyllum, sp 37 12-15. Sporangiutn-like organs of Glossopteris Brown iana, Brong 41-43 16. Glossopteris Browniana, Brong 51 17. G. indica, Sehimp 66 18. G. indica, Schiinp., and G. angastijolia, Brong. . 68 19. G. angustifolia, Brong 74 20. G. ampla, Dana 80 21. G. retifera, Feist 84 22. G. conspicua, Feist 86 23. G. divergent, Feist 89 24. G. decipiens, Feist 91 25. G. orbicularis, Feist 93 26. Gangamopteris cyclopteroides, Feist 107 27. G. (?) buriadica, Feist., and G. cyclopteroides, var. major, Feist 112 28. Neuropteridium validum, Feist 118 29. Tceniopteris Feddeni (Feist.) 123 30. Palceovittaria Kurzi, Feist 129 XII LIST OF FIGUKES IX THE TEXT. TAGE Fi<;. 31. Sphenoptens polymotpha, Feist 132 ,, 32. S. Hughesi (Feist.) 134 ,, 33. Cludophlebis Rogki, Arber 143 ,, 34. Merianopteris major. Feist 145 ,, 35. Belemnopteris Wbod-Masomana, Feist 146 ,, 36. Bothrodendron Leslii, Seward 167 ,, 37. Noeggerathiopsis Hislopi (Bunb.) 182 ,, 38. N. (?) Sfoliczhina (Feist.) 188 ., 39. [N. ?] lacerata, Feist 189 ,, 40-43. Badoxylon austr ale, sp. noy 1 02—196 ,, 44. Cardiocarpus indtcus, Zeiller 205 ., 45. Pterophyllum (Anomoza ■mites) Balli (Feist.) . . . 209 ,, 46. Rhipidopsis ginkgoides, Schmal 211 ,, 47. Psygmophyllum Kidstoni, Seward 213 ,, 48. Vbltua heterophylla, Brong 217 ,, 49. Cyclopitys dichbtoma, Feist 219 ,, 50. Brachgphyllum (?) australe, Feist 221 ,, 51. Bicty opteridium sporiferum, Feist 221 CONTENTS. PAGE Preface v Author's Preface vii Note ix List of Figures in the Text xi Introduction xvii (1) Botanical Affinities of the Glossopteris Flora xx (2) Distribution in Space xxxii (3) Age and Distribution in Time xxxviii Historical Sketch xliv History of the Collection lxxi Algjs 1 Reinschia australis, Bertr. & Ren 1 Pila australis, Bertr 2 Equisetales 3 Schizoneura gondicanensis, Feist 5 S. Wardi, Zeiller 12 8. (?) africana, Feist 13 [Actinopteris] lengalensis, Feist 14 Phyllotheca australis, Brong 17 P. indica, Bunb 20 P. deliquescens (Gopp.) 22 P. Griesbachi, Zeiller 25 P. robusta, Feist 25 P. Etheridgei, sp. nov 26 P. Zeiller i, Etheridge, jun 28 Phyllotheca, sp 29 [Annularia ?] australis, Feist 30 \JEauisetites ?] Morenianus, Kurtz 32 Obscure Articulated Casts 33 XIV CONTENTS. PAGE Sphenophyllales 34 Sphenophyllum speciosum (Royle) 35 Sphenophyllum, sp 36 Filicales (?) 37 Glossojjleris 38 The Fructification of Glossopteris 39 Species of Glossopteris 45 Synopsis of Species of Glossopteris 47 G. Browniana, Brong 48 G. indica, Schimper 64 G. angustifoUa, Brong 72 G. stricta, Banbury 76 G. ampla, Dana 78 G. retifera, Feist 83 G. conspicua, Feist 86 G. formosa, Feist 87 G. tortuosa, Zeiller 88 G. di verge ns, Feist 89 G. decipiens, Feist 90 G. longicaulis, Feist 92 G. orbicularis, Feist 92 Vert eh ear ia indica, Royle 97 Boots and Rootlets of uncertain affinity 101 Gangamopteris cydopteroides, Feist 104 G. cydopteroides, var. major, Feist 113 G. angiistifolia, McCoy 113 G. Wldttiana, Feist 114 67. (?) bur indica, Feist 115 Neuropteridium ralidum, Feist 116 Tamioptcris danceoides (Boyle) 121 T. Feddeni (Feist.) 123 T. spathulata, M'CleUand 124 Tceniopteris cf. M'Clellandi (Old. & Morr.) 126 Taniopteris, sp 128 Paheorittaria Kurzi, Feist 130 Sphenopteris polymorpha, Feist 131 S. Uughesi (Feist.) 133 CONTENTS. XV PAGE Sphenopteris lobifotta, Morris 135 S. data (Brong.) 138 Sphenopteris, sp 139 Pecopteris phegopteroides (Feist.) 1-40 Cladophlebis Roylei, Arber 142 Cladophlebis, sp 144 Merianopteris major, Feist 144 Belemnopteris Wood- Mason iana, Feist 147 Psaronius brasiliensis, Unger 148 [ Caulopteris ?] Adamsi, Feist 152 Obscure Stem Structures. Incertce Sedis 153 Lycopodiales 153 Lepidodendron Pedroanum (Carr.) 156 L. Derbyi (Renault) 159 L. (Knorria), sp 162 Obscure Lepidodendroid fossils 162 Lepidopliloios laricinus, Sternb 163 Bothrodendron Leslii, Seward 166 Sigillaria Brardi, Brong 170 Lycopodean Spores 175 COROAITALES 177 Npeggerathiopsis Hislopi (Bunbury) 179 JV. Whittiana (Feist.) 186 N. (?) Stoliczlcana (Feist.) 187 [JIT. ?] lacerata, Feist 188 Dadoxylon australe,_sp. nov 191 i). P«&w, Zeiller 199 D. Maiilandi (Shirley) 201 J). Binneyi (Shirley) 201 B. Williamsoni (Shirley) 201 J). brisbanense (Shirley) 202 Dadoxylon, sp 202 Incepts Sedis 203 Petrified Woods 203 Gymnospermous Seeds, possibly of Cordaitean affinity . 204 Cardiocarpus indicus, Zeiller 205 XVI CONTENTS. PAGE C. (?) Milleri (Feist.) 205 Cardiocarpus, sp 206 Cycadophyta 208 Pterophyllum {Anomozamites) Balli (Feist ) 208 Cycadites, sp 210 ? GlNKGOALES 210 Rhipidopsis ginkgoides, Schuial 211 it. densinervis, Feist 212 Psygmophyttum Kidstoni, Seward 213 Ottokaria bengalensis, Zeiller 215 CONIFEKALES 215 Voltzia heterophylla, Brong 216 Seeds possibly belonging to Voltzia 218 Albertia(?),8$ 218 Cyclopitys dichotoma, Feist 220 Brachyphyttum (?) australe, Feist 221 Araucarites Oldhami, Zeiller 222 [ Walchia], sp 222 Scale-leaves of Gymnosperrnous affinity 223 Plants Incert^e Sedis 223 Conites, sp 223 Diciyopteridiiim sporiferum, Feist 224 Appendix 225 Sphenophyllum speciosum (Boyle) 225 Gangamopteris kashmiremis, Seward 225 Bibliography 227 Index 245 Explanation of the Plates. INTRODUCTION. The oldest assemblage of plant-remains with which we are acquainted, in number sufficiently numerous to rank as a flora, is that of the Upper Devonian and Lower Carboniferous rocks. 1 On the present evidence, the floras of these two geological periods constitute one great paia^obotanical epoch, characterised by remarkable uniformity throughout the world. 2 But when we pass to the vegetation of the succeeding Upper Carboniferous and Permian deposits we no longer find that it exhibits such a worldwide uniformity of distribution. The plant-remains of these two periods, although everywhere constituting another great epoch in the history of the vegetable kingdom, are grouped naturally in two well-marked botanical provinces. Thus one flora existed in Europe, Northern Asia, and North America, more or less contemporaneously with a dissimilar type of vege- tation, confined for the most part to India and the Southern Hemisphere. The southern flora differed remarkably in certain features from that of the Northern Hemisphere. It is now generally accepted that these two types of Permo-Carboniferous vegetation flourished on two great continental regions, for the most part, but not entirely, isolated, and widely separated from one another. The : The earliest known fossil plants are of Silurian ag-e. 3 Seward (03-), p. S31. b XV111 INTRODUCTION. great Southern Pernio - Carboniferous continent lias been dis- tinguished by Suess ] as Gondwanaland, a term derived from tbe great series of fresh-water sediments in India, to which Medlicott gave the name, Gondwana system. Only the lower part, however, of the Gondwana rocks of India belong to the period under discussion. It is with the fossil plants of Gondwanaland that we are here concerned. This flora is especially characterised by the frequent occurrence of Glossopteris, a fern-like plant not only of wide distribution but often found in such extreme abundance that some of the rocks of Gondwanaland appear to be largely composed of impressions of its fronds. Thus the southern type of Permo- Carboniferous flora has come to be known as the Glossopteris flora, a name first used by Neumayr ~ to distinguish it from the dissimilar, but more or less contemporaneous flora of the Northern Hemisphere. The Glossopteris flora is found typically developed in four great provinces of Gondwanaland, viz., India, Australia with Tasmania, Southern Africa, and South America. The accompanying map shows roughly the relative distribution of the Northern and Southern Permo-Carboniferous floras. At the close of the Permo-Carboniferous period, some members of the flora had become extinct, while others still survived and con- tributed to the vegetation of Triasso-Rhaetic times, though in a subordinate degree ; the survivors being no longer dominant elements in the vegetation of the early Mesozoic rocks. In discussing this flora three main points have to be considered : (1) its botanical affinities, (2) its distribution in space, (3) the evidence as to its age and distribution in time. We may commence with a brief account of the botanical affinities of the Glossopteris flora. At the same time it must be borne in mind that our knowledge at present on this point is still very imperfect, and consequently some of the conclusions arrived at must be regarded as of a provisional nature. 1 Suess [85), vol. i. p. 76S. 2 Neumayr (87), p. 191. INTRODUCTION. 1. BOTANICAL AFFINITIES OF THE GLOSSOPTERIS FLORA. In order to understand the botanical affinities of the Glossopteris flora it will be necessary to compare it somewhat closely with the Northern type of Pernio - Carboniferous vegetation, and also with its immediate predecessor in geological time, the flora of the Upper Devonian and Lower Carboniferous rocks, which, as we have pointed out, was characterised by worldwide uniformity of distribution. The plant-remains of the Upper Devonian period are un- fortunately scanty, but within recent years our knowledge of these ancient types of vegetation has been considerably increased. It is interesting to note that beds of this age within the Arctic regions have been found to furnish more valuable material in this respect than those developed elsewhere. The Devonian floras of several Arctic lands are now known, thanks especially to the researches of Nathorst. 1 Among the other sediments of this age which have yielded plant-remains, the more important are those of Southern Ireland, 2 Belgium, 3 Austria, 4 Russia, 3 and Canada. 6 In the Lower Carboniferous deposits of Europe and North America, plant-remains are more numerous and less fragmentary, and con- sequently the flora of this period is better known. It appears, however, to be essentially similar in type to that of the Upper Devonian rocks, and in all probability we may regard the Upper Devonian and Lower Carboniferous floras as constituting one great botanical epoch. Plant-remains belonging to the latter period 1 Nathorst (94), (00), (02 1 ), (04). 2 See Kidston (86), pp. 228-236. 3 Crepin (74), (75). 4 Stur (81) ; Potonie & Bernard (04). 5 Schmalhausen (94). 6 Dawson (59), (71). (81). INTRODUCTION. XXI have been described from Britain, 1 France, 2 Austria, 3 Russia, 4 China, 5 North America, 6 the Arctic regions, 7 and elsewhere. 80 far we have mentioned only certain areas lying within the Northern Hemisphere. But it is particularly interesting to find that there is a considerable accumulation of evidence as to the existence of plants belonging to this same epoch in some of the provinces in which the (Jlossopteris flora flourished at a later period. Thus either Upper Devonian or Lower Carboniferous plant- remains have been described from New South Wales, 8 Victoria, 9 Queensland, 10 and Argentina." With the exception of the last- mentioned, floras belonging to both these periods have been recorded. In India, unfortunately, the earliest plant- bearing sediments are those in which the Glossopteris flora is found, and in South Africa few, if any, plants are known from rocks of earlier age than Permo-Carboniferous. On the present evidence, however, the plant-remains of the Lower Carboniferous and Upper Devonian rocks of the other regions in which Glossopteris flourished at a later period appear to be genetically, and often specifically, identical with those of the Northern Hemisphere. We may briefly sum up the botanical affinities of the Lower Carboniferous and Upper Devonian flora as follows : — It consists, so far as is known, of representatives of six great groups. Of these the Equisetales, to which the modern Horsetails belong, are chiefly represented by an ancient and somewhat archaic type, Archceoealamites. The genus Catamites is also found, but com- paratively rarely, in the Lower Carboniferous rocks. The great group of Lycopods, including the modern Club-mosses, is represented 1 Kidston (03). - Vaffier (01). A Stur (75). 4 Eichwald (60). 5 Schmalhausen (S3 1 ), p. 433. 6 Dawson (73). ' Nathorst (94). » Feistmantel (90) ; Etheridge (90-) ; David & Pittman (93) ; Dun (99) 9 McCoy (74), Prod. i. iv. 10 Feistmantel (90), p. 54 ; Carruthers (72). 11 Szajnocha (91). INTUODrCIIOX. by two genera, Lepidodendrdn and Bothrodendron, while Lepido- phloios, and to some extent also Sigillaria, appear in Lower Carboniferous times. The Fern-like plants, whether they be true Ferns (Filicales), or fern-like Spermophytes (Pteridospermeae), are numerous. Only one genus, Adiantites, a characteristic type, is at present known to have belonged to the latter group, while the true affinities of others, such as Archteopteris, Ca/rdiopteris, li/iacopfcn's, and Sphenopten*, have yet to be ascertained. There is, however, evidence, from specimens in which the internal structure is preserved, that the Cycadofilices, of which some members are known to have possessed the seed-bearing habit (Pteridosperms), were not uncommon at this period. There remain two important groups, the Sphenophyllales, of which the genus Sphenophyttum is the principal representative, and the Cordaitales, the latter of Gymno Members of the Glossopteris flora have also been recognised in Kashmir 4 (Permo-Carboniferous), Afghanistan and Persia 5 (Permo- Carboniferous), Asia Minor 6 (Upper Carboniferous), and the Altai and Siberia 7 (Permian). Survivals of the Glossopteris flora also occur in the Triasso-Rhaetic rocks of India (Panchet division), Tonquin, 8 and China. 9 1 Araalitzky (97), (01). - Schmalhausen (79). ! Feistmantel (79 3 ). 4 Noetlinj? (03) ; Seward & Woodward (05). 5 Griesbach (85). 6 Zeiller (99). 7 Schmalhausen (79), (S3 1 ) ; Zeiller (96 3 ). * Zeiller (02 3 ). '■> Krasser (01); Zeiller (00-). INTRODUCTION. XXXlll Australasia : New South Wales} Wianamatta Series ... j Triasso _ paretic, with a few survivals of Hawkesbury Senes, with glacial ■ ^ G1 teris Flora> deposits ... ... ... I Newcastle Series ... —Glossopteris Flora'j i Upper Marine Beds I Permo- Muree Series ] "Lower Coal Measures" — Glossopteris Flora j Carboniferous. ( Lower Marine Beds ... ... ... ...J Glacial deposits Victoria. 2 Bacchus Marsh Sandstones ... Permo-Carhoniferous :— Glossopteris Flora. Bacchus Marsh Glacial deposits. Queensland. 3 Bowen River Series ... Perrao-Carboniferous : — Glossopteris Flora. Glacial deposits. Western Australia. 4, — Members of the Glossopteris flora have been recognised in the Collie River coalfield, and at the Gascoyne River. Tasmania. 5 Mersey River Series ... Permo-Carboniferous : —GlossopterLs Flora. Glacial deposits. Southern Africa 6 : Cape Colony, Natal, Transvaal, Orange Hirer Colon//. !Stormberg Series ... Triasso-Rhsetic Flora. Beaufort Series ... Pernio- Carboniferous : — Glossopteris Flora. Ecca Series ... Permo-Carboniferous : — Glossopteris Flora. Ihvyka Conglomerate (Glacial). The Glossopteris flora has also been recorded from Rhodesia, 7 and from German and Portuguese East Africa. 8 South America : Brazil, 9 Argentina} — A list of the plant-remains occurring in each of these regions of Gondwanaland will be found in the Historical Sketch. 1 Feistmantel (90). 2 McCoy (74), (98) ; Feistmantel (90). 3 Carruthers (72), (80) ; Feistmantel (90) ; Jack & Etheridge (92) ; Shirley (98), (02). 4 Etheridge, sen. (93) ; Etheridge, jun., & David (94). 5 Johnston (86), (88), etc. ; Feistmantel (90). 6 Feistmantel (89) ; Zeiller (96 1 ) ; Seward (97 1 ) ; Etheridge (01). 7 Arber (03). 8 Potonie (99), (00). 9 Carruthers (69) ; Zeiller (95 2 ). 10 Kurtz (94 3 ); Bodenbender (95), (96). c "Fntqg ■umbnoj, •noisiAig ^atpuBj : Bipni ::«::::::■:■:: = ■::: •BTip-uasay •JIZB.ljJ "Boujy ^Bg; i)saii.Snj.u) ( i ptiB nimi.Mf ) ■Bisapoq'jj : : : x : : : * ■fUBASiraij, pus 'I ,: 1 1; X AnO[OQ 8cIbq •BiumusBX ■ptiBjsnaanf) "BUapjA •sauag a[}s«0Avax; : * * : h : x x X X ... X X ■saiiag aampj •noisiAiQ Bpn:ut!| | x : x .>. x a. : x : x x x x x x ■noisiAiQ iiqopsj, •Bissrvg :adoanjj .2 ,« = e s -8 : g £ e : « € ~ C ~ ;c -~ 15 "~ e* « e e t ■ «» S "° ■** "~ s Si £ i s cs =5 « a.-S ~.s a,S 2 -5 ^ = & « s»-S ■« <2> £K»N S 4 s &^ ST g -sex.©*. =5 S 2§ =q •< iS ^ =; ^ ^ ^ ^' ^' JjS^bo SoJOsOJ Oj to a- X X • * X • X X xxxx : :xxx : :x : :xx :.-v. :xxx : :xxxxxx : xx .xxxx -XX : S "S > : : : « ■ § »l - • • • s> 8 S "= S .2 -S fe S « -1 ^2 -2 -~ J <3 ?. ; v s s o s, « ^ ; » o t, ; sj ? s C; "* J* |1 3 R, R..S = ° ~ "» ' 1 § .2 .2 | C $ g I J | 2 2 2 •- a, - * f> ^^= v £ 2 o 8 •a (A o OQ OQ .5 H •BtliqQ •nmbuoj, •nuisiAiQ ^oqjuuj : uipuj so cs o !- '3 o d o o i V •■Bni^uaSiy •jizB-ig 3S8T1 "botjjy ;se;j • instpoq^j •[BBAStUUJ, pun "I^BX \\no[OQ acluQ "BIUBraSBJ, •pn^jsnaan^) : : : : x x x x x ::«.::::::::::::: : ■BUO^OI^ CD •S9IJ3g 9J^SB0Ai8X •sauag aa.inj^ p •noisiAiQ upninEQ •noisiAiQ .nqojBj, x x : : : : : : : x x * : * : : : x x * : . : " : x CD c nc o 3 1 «5 1 III." JJ ' I s^* : -S : : '. : : : s : : S :=• -;2 2 sj :*^^s : ;», ■~c -~ 1 . Oldham (86). 1 Waagen (86) : W. T. Blanford (86). Xl INTRODUCTION. important discovery was the first direct evidence to be gained from India bearing on the age of the Glossopteris flora. In the previous year, Griesbach ' had shown that Vertebraria occurs in beds which overlie similar glacial deposits in Afghanistan. Still more recently, the existence of a well - marked Glacial period at the beginning of Periuo-Carboniferous times has received yet wider confirmation. The Dwyka conglomerate of South Africa is of glacial origin, and occupies a similar position with regard to the Glossopteris - bearing beds of that province as in India and Australia. Further, in 1888, Derby 2 showed that glacial deposits belonging to the same period are also developed in Brazil. Within the last two years, Noetling 3 has found Gangamopteris in the Salt Range of Kashmir in beds below those containing Permian invertebrata. Amalitzky 4 has also discovered the Glossopteris flora in Russia on a horizon intermediate between a plant-bearing series of Lower Permian age and deposits containing Upper Zechstein mollusca. Further, there is the botanical evidence of the Glossopteris flora itself. The representatives of two groups associated with that flora, the Lycopods and the Sphenophyllales, are generically, and sometimes even specifically, identical with those of the northern type of vegetation. Among the earlier types of a Mesozoic facies, such as Tceniopteris, Voltzia, and Rkipidopxis, which begin to come in during Permo-Carboniferous times in both the northern and southern floras, there is also generic identity. Also the essentially Palaeozoic types, which at this period were dominant elements in the flora, all belong, so far as their botanical affinities are known, to the same great classes of plants as the members of the flora of the Northern Hemisphere. At the close of Permo-Carboniferous times, a change took place in the flora of Gondwanaland. It is of great interest to find that we have an almost continuous succession of plant-bearing sediments in more than one province of this continental region, in part of Palaeozoic, and in part of Mesozoic age. This is the case in India 1 Griesbach (85), p. 62. 2 See W. T. Blanford (95), (96). 3 Noetling (03) ; Seward & Woodward (05). 1 Amalitzky (97), (01). INTRODUCTION. xli and New South Wales, and possibly also in South Africa. "\Ve are able therefore to trace to some extent the fate of the various elements of the Palaeozoic flora as recorded in rocks of Triasso- Ehaetic age. We find that by this time some members of the flora had become extinct, though representatives of perhaps the majority of the more important genera still survived. These, in some cases, inhabited regions more or less conterminous with the pre-existing Palaeozoic continent, but the more marked characteristic of this period is the evidence of a widespread migration i of many of the Palaeozoic genera far beyond the boundaries of Gondwanaland. This conclusion is not based on the distribution of such elements of a Mesozoic facies as are present in the Glossopteris flora, as well as in the northern type of Palaeozoic vegetation. We recognise such elements in the genus Teemopteris among Fern- like plants, the Cycadophyta, the Ginkgoales, and the Coniferae, such as Voltzia, associated in India and elsewhere with Glossojrteris and its allies. The presence of these types, in common with the marked migration in early Mesozoic times of certain genera essentially of a Palaeozoic facies, serves but to confirm the con- clusion that the change in the vegetation from a Palaeozoic to a Mesozoic facies was of an extremely gradual nature. In the Triassic rocks (excluding the Khaetic) we have a flora which is essentially a Transition flora, in which the older Palaeozoic types one by one die out, and gradually give place to newer types of a Mesozoic facies, whose first incomings, as in the case of the above- mentioned families, we can trace back as far as the Carboniferous rocks. Put when we turn to the truly Palaeozoic genera, such as Sehizoneura, Phyttotheca, and NeuroptericUum, we find in several cases that they have spread as far as Western Europe in Triassic times. It is a moot point at present how far the presence of such survivals of the Glossopteris flora may be regarded as dominant members of the flora of the Triassic (excluding the Phaetic) period. The difficulty arises from the fact that the vegetation of the Triassic rocks of Europe is still very imperfectly known. Put whether we regard such survivals as characteristic and dominant elements in the flora of that Transition period or not, there can be no doubt Seward (03-), p. 835. xlii INTRODUCTION. that they did not play such a part in the true Mesozoic floras. The Mesozoic facies of vegetation is first found fully developed in Rhaetic times, and persisted in the Jurassic and "Wealden. It is especially characterised by a general absence of the older Palaeozoic types ; such as existed being numerically inferior and quite over- shadowed by the development of new types of plant life. It is true that in Tonquin, China, Europe, and elsewhere a few survivals, such as Schizoneura, Glossopteris, etc., have been found associated with the Rhcetic flora, but these, as recent research has shown, are all of secondary importance as compared with the other elements of the flora. Similarly, Phyllotheea survived as late as the Jurassic period in Italy, whereas almost all the other elements of the Jurassic flora are essentially of a Mesozoic type. It has been assumed by many palaeobotanists in the past that the Mesozoic facies of the Glossopteris flora is beyond dispute, since several of its most characteristic members are found in the Trias, and even extend to the Rhsetic. I have endeavoured to show that this conclusion is incorrect. Further evidence may be found in the general absence of Mesozoic types from the rocks of Gondwana- land, with the exception of a few Ferns, Cycads, Ginkgoales, and Coniferae, which both in the Northern and the Southern Hemi- spheres arc among the earliest arrivals of a Mesozoic facies. "We find practically no representatives of such typical Mesozoic genera as JEquisetites, Clathropteris, Laccopteris, Dictyophyllum, Sagenopteris, WiUiamsonia, Otozamites, Podoza mites, Nilssonia, Ginkgo, and Baiera, among many others which might be mentioned as having a worldwide distribution. In the past, the evidence of the Mesozoic facies of this flora has been based parti}' on inaccurate generic determinations. I have already noticed (p. xxx) some of these in relation to the Cycads, a group which in reality is very scantily represented in the Glossopteris flora. Similarly, the name Glossopteris lias been applied by some of the older authorities on paleobotany to Mesozoic plants which do not belong to that genus. 1 The converse has also occurred in the case of Sagenopteris, a typical frond of a Mesozoic facies, which has been recorded from 1 Schimper (69), vol. i. pp. 641-2. INTRODUCTION. Xllll Grondwanaland by Feistmantel, but wbicli iu reality is entirely unrepresented in that flora. The attributions to the genera Actinopteris and Asplenium are other cases in point. With regard to the nature of the climate of Gondwanaland there is no evidence to be gained from the plant-remains as they are at present known. This is particularly unfortunate, since it would- seem to be impossible to account for tbe differences between the northern and southern floras by the mere fact of isolation alone. It would appear that climate as well as isolation must have had a determining influence on the distribution of Perino- Carboniferous vegetation. 1 Probably the existence of widespread glacial conditions immediately preceding the deposition of the earlier Glossopteris-beaxing sediments had a marked influence in this connection. 2 Dr. Blanford 3 has suggested there is "some evidence in favour of the view that the transfer of the southern plants to the Northern Hemisphere was caused b\- a period of low temperature that drove a southern temperate flora northward to the equator." He adds that "it is highly probable that many other forms of terrestrial life besides the Mesozoic flora originated in the Southern Hemisphere ; and unless a very considerable area of what is now deep ocean was occupied by land in Mesozoic and Palaeozoic times, a change in favour of which there appears but slight evidence, it is far from improbable that the Antarctic continent was the original area of development." 1 W. T. Blanford (90), p. 96. • Sec Seward (03-),' pp. 834-5 ; II. F. Blanford (75), pp. 534, 510. 3 W. T. Blanford (90), pp. 105-6. xliv HISTORICAL SKETCH. HISTORICAL SKETCH. The various regions of the world, from which members of the Glossopteris flora have been recorded, are so scattered that it will perhaps be most convenient, in a brief historical sketch of the rise and progress of our knowledge of this flora, to consider each area separately. I. Southern Asia. («) India, The first Indian specimens of Glossopteris, obtained from "Rana- Gunge, near Rajeinahl," were described by Brongniart ' in 1828. At the same time another important genus, afterwards known as Schizoneura, was founded. 2 Glossopteris Browniana, sp. no v. Zeugophyllites — ? Schizoneura.* Conral/arites = Schizoneura. In 1830 Brongniart, 4 in his " Histoirc des Vegetaux fossiles," described and figured two Indian species of Glossopteris. Glossopteris Browniana, var. indica = G. indica, Schimper. G. angustifolia, sp. nov. Boyle, 5 in the first part of his " Illustrations of the Botany of the Himalayan Mountains," published in 1833, figured several important types from the Bard wan Coalfield (Raniganj Group, Damuda Series) of India. These specimens, now in the British Museum, were re-examined by the present author in 190 1, 6 and are also described in this volume. Trizygia speciosa, sp. nov. = Sphenophyllum speeiosum (Royle). Vertebraria indica, sp. nov. Glossopteris danaoides, sp. nov. = Tceniopferis danceoides (Royle). Pecopteris Lindleyana, sp. nov. = Cladophlebis Roylei, Arber. Pustularia ealdcriana, sp. nov. = Nomen nudum. 1 Brongniart (28 1 ), p. -54. - Brongniart (28 1 ), pp. 121, 128, 175. ■ Arber (02 l ), pp. 18, 19. 4 Brongniart (28 2 ), pp. 223-4, pi. lxii (pars) ; pi. lxiii, tig. 1. ' Boyle (33). ,; Arber (01). HISTORICAL SKF.TCIT. xlv In 1836, Goppert 1 described and figured Glossopteris in his work "Die fossilen Farnkrauter." In 1850, M'Clelland 2 figured several new specimens from the Bardwan Coalfield, India, in addition to examples of three species which had been previously described. His drawings, however, were so carelessly executed, and the descriptions of the specimens so imperfect, that it is practically impossible to recognise any of M'Clelland' s species, with the exception of Taniopieris danmoides. Zamia burdwanensis, sp. nov. = ? Tceniopteris, sp. 3 Sphenophylluni fusciculatum, sp. nov. = ? Poneites muricata, sp. nov. = ? P. minor, sp. nov. = ? Glossopteris acaulis, sp. nov. = Glossopteris, sp. G.frondosa, sp. nov. = Glossopteris, sp. G. reticulata, sp. nov. = Glossopteris, sp. Tceniopteris danceoidts (Royle). Pecopteris ajfinis, sp. nov. = ? Sphenopteris polymorpha, Feist. 4 Fucoides venosus, sp. nov. = ? Glossopteris, sp. Sir Charles Bunbury, 5 in 1861, described a number of fossil plants from Nagpur in Central India, of which several were new species. These important type-specimens are now in the Museum of the Geological Society of London. Glossopteris Browniana, var. indica, Brong. = G. indica, Schimper. G. Browniana, var. australasica, Brong. = G. Browniana, Brong. G. muscefolia, sp. nov. = G. ampla, Dana. G. leptoneura, sp. nov. = G. angustifolia, Brong. G. stricta, sp. nov. Pecopteris (?) = Sphenopteris polymorpha , Feist. Cladophlebis = Sphenopteris polymorpha, Feist. (?). Tceniopteris danceoides, M'Clell. (?) = ? T. cf. WGlellandx (Old. & Morr.). Filiates = ? Glossopteris, sp. Filicites (qu. Glossopteris, sp. ?) = Glossopteris, sp. Nbeggerathia? (Cyclopteris?) Hislopii, sp. nov. = Noeggeruthiopsis H'tslopi (Bunb.). Phyllotheca indica, sp. nov. Vertebraria (?). Knorria? (conifer?) = ? Bothrodendron , sp. Ywcites (?) = ? 1 Goppert (36), p. 346, pi. xxi, figs. 9, 10. 2 M'Clelland (50), pp. 53-6, pis. xiv-xvii. 3 See p. 122. 4 See Feistmantel (80), p. 77 ; and p. 132 of the present work. 5 Bunbury (61). xlvi HISTORICAL SKETCH. Feistmantel, 1 in 1876, described the following new species from the Damuda formation : — Schizoneura gondwanensis, sp. now Gangamopteris cyclopteroid.es, sp. now Sagenopteris, sp. = Glossopteris, sp. Neuropteris valida, sp. now = Neuropteridium vdlidum, Feist. Actinopteris bengalensis, sp. nov. = \_A.~\ bengalensis, Feist. In the same year Feistmantel 2 published an important paper in the Journal of the Eoyal Asiatic Society of Bengal, in which the following additions to the Glossopteris flora were described : — Sphenopteris polymorpha, sp. nov. Alethopteris phegopteroides, sp. nov. = Pecopteris phegopleroides (Feist.). Palteovittaria Kurzi, sp. nov. Belemnopteris Wood- Masoniana, sp. nov. Gangamopteris Whittiana, sp. nov. Glossopteris communis, sp. nov. = G. indica, Schiraper. Sagenopteris poly phylla, sp. nov. = Glossopteris retifera, Feist. In 1877 the following species were added : — 3 Glossopteris stenoneura, sp. nov. 3 = G. indica, Schimper. Ehizomopteris Balli, sp. nov. Feistmantel, 4 in 1879, published the first part of his great work on the Fossil Flora of the Lower Gondwanas, which included numerous figures of almost all of the Permo-Carboniferous plants known from India. The first section, together with a supplement, was devoted to the lower series of Glossopteris -hewing rocks, known as the Talchir-Karharbari beds. The following species were described : — Schizoneura cf. Meriani (Brong.) = Schizoneura, sp. (?). S. gondwanensis, Feist. Yertebraria indica, Boyle. Neuropteris valida, Feist. = Neuropteridium validum, Feist. Glossopteris communis, Feist. = G. indica, Schimper. G. damudica, sp. nov. = G. ampla, Dana. G. decipiens, sp. nov. Gangamopteris (?) bitriadica, sp. nov. G. major, sp. nov. = G. cyclopteroides, var. major, Feist. G. ci. angustifolia, McCoy. G. cyclopteroides, Feist. 1 Feistmantel (76 1 ). - Feistmantel (76 2 ). 3 Feistmantel (77 1 ), p. 74. 4 Feistmantel (79 1 ). IIISTOKICAL SKETCH. xlv Gangamopteris cyclopteroides, var. subaurkulata, Feist. = G. cyclopteroides, Feist. G. cyclopteroides, var. areolata, Feist. = G. cyclopteroides, Feist. G. cyclopteroides, var. attenuata, Feist. = 6r. cyclopteroides, Feist. Sagenopteris (?) StoliczJcana, sp. now = ? Glossopteris decipiem, Feist. Glossozamites Stoliczkanus, Feist. = Noegyerathiopsis('i) Stoliczlcana (Feist.). Koeggerathiopsis Hislopi (Bunb.). iV. Hislopi, var. subrhomboidalis, Feist. = iV. Hislopi (Bunb.). Euryphyllum WAittianum, sp. nov. = Noeggerathiopsis Whittiana (Feist.). Voltzia het erophylla, Brong. Samaropsis of. parvula, Heer. Carpolithes Jl/t/eri, sp. nov. = Cardiocarpus (?) Milleri (Feist.). In the second and third parts of the Lower Gondwana flora, published in 1880-1, Feistmantel dealt with the fossil plants from the Damuda-Panchet divisions. The following Pernio ■ Carboniferous species are figured : — x Schizoneura gondwanensis, Feist. S. cf. Meriani (Brong.) = Schizoneura, sp. (?). Phyllotheca indica, Bunb. P. robusta, sp. nov. Trizygia speciosa, Royle = Sphenophyllum speciosum (Boyle). Vertebra ria indica, Royle. Cyathea cf. Tchihatcheffi, Schmal. = cf. Sphenopteris polymorpha, Feist. Sphenopteris polymorpha, Feist. Dicksonia Hughesi, Feist. = Sphenopteris Hughesi (Feist.). Asplmium. whitbyeme, Goepp. = Cladophlebis, sp. Altthopteris Lindleyana, Royle = Cladophlebis Roylei, Arber. A. phegopteroides, Feist. = Pecopteris phegopteroides (Feist ). Merianopteris major, sp. nov. Macrotecniopteris danceoides (Royle) = Tamiopteris danecoides (Royle). M. Feddeni, Feist. = Tceniopteris Feddtni (Feist.). Paheovittaria Karzi, Feist. Angiopteridium cf. M'Clellandi (Old. & Morr.) = Tceniopteris cf. M'Clellandi (Old. & Morr.). A. infarction, sp. nov. = cf. Tceniopteris M' Clcllandi (Old. & Morr.). Glossopteris communis, Feist. = G. indica, Schimper. G. communis, var. stenoneura, Feist. = G. indica, Schimper. G. intermittens, sp. nov. = G. Browniana, Brong. G. stricla, Bunb. G. museefolia, Bunb. = G. ampla, Dana. G. indica, Schimper. G. Browniana, Brong. G. retifera, sp. nov. G. conspicaa, sp. nov. 1 Feistmantel (80). xlviii nrsmurcAL sketch. Glossopteris dirergens, sp. uov. G. damudica, sp. nov. = G. ampla, Daua. G. angustifolin, Brong. G. leptoneura, Buub. = G. angastifolia, Brong. G. formosa, sp. nov. G. orbicularis, sp. nov. G. decipicns, Feist. Gangamopteris anthrophyoides, sp. nov. = ? G. Whittiana, Feist. G. Hughesi, Feist. = G. cyclopteroides, Feist. G. cyclopteroides, Feist. G. cyclopteroides, var. subauriculata, Feist. = G. cyclopteroides, Feist. G. cyclopteroides, var. areolata, Feist. = G. cyclopteroides, Feist. G. cyclopteroides, var. attenuata, Feist. = G. cyclopteroides, Feist. Belemnopteris Wood-Masoniana, Feist. Sagenopteris longifolia, sp. nov. = Glossopteris angustifolia, Brong. S. (?) polyphylla, sp. nov. = Glossopteris retifera, Feist. S. cf. rhoifolia, Presl = Glossopteris, sp. S. (?) Stoliczkaua, Feist. = ? Glossopteris decipicns, Feist. Actinopteris bengalensis, Feist. = [-<4.?] bengalensis, Feist. Pterophyl/um burdwanense (M'Clellandj = ? Tceiriopteris, sp. Glossozamites Stoliczkanus, Feist. = Xoeggerathiopsis (?) Stoliczkaua (Feist.). Koeggerathiopsis Hislopi (Buub.). Euryphyllum Whittianum, Feist. = Xoeggerathiopsis Whittiana (Feist.). Rhipidopsis densinervis, sp. nov. Yoltzia heterophylla, Brong. In 1881, Feistraantel 1 also instituted the species — Cyclopitys dichotoma, sp. nov. The first part of the fourth volume of the Gondwana Flora, published in 1882, contained descriptions of plant-remains from the South Rewah Gondwana basin. Feistmantel, 2 in addition to further figures of several plants already well known, describes the following new species : — Glossopteris cordata, sp. nov. = cf. G. Browniana, Brong. G. tcenioides, sp. nov. = cf. G. avgusti folia, Brong. Xoeggerathiopsis lacerata, sp. nov. = \N. ?] lacerata, Feist. In 1886, Feistraantel 3 completed his Flora of Lower Gondwanas by a memoir on the plants obtained from some of the coalfields in Western Bengal. In addition to figures of several plant-remains previously recorded, the following new varieties were described : — Gangamopteris cyclopteroides, var. acuminata, Feist. = G. cyclopteroides, Feist. G. cyclopteroides, var. cordifolia, Feist. = G. cyclopteroides, Feist. 1 Feistmautel (81 1 ), p. 257. 2 Feistmantel (82'). 3 Feistmantel (8G). HISTORICAL SKETCH. xlix In the second edition of the "Geology of India," edited by R. D. Oldham ' and published in 1893, a list of Permo-Carboniferous plants is given, as well as figures of the more important species. The most recent memoir 3 on the Glossopteris flora of India contains the following new species instituted by Professor Zeiller. 3 as well as many figures of some of the earlier described plants : — Glossopteris tortuosa, sp. nov. Schizoneura Wardi, sp. nov. Phylhtheca Griesbachi, sp. nov. Cycadites(?), sp. Ottokaria bei/i/a/ensis, sp. nov. Araucarites Oldhami, sp. nov. Cardiocarpus indicus, sp. nov. Cardiocarpiis, sp. (b) Kashmir. Dr. Noetling 4 has recently discovered a Gangamopteris in beds overlain by the Permian rocks of the Salt Range Series at Khumnu, in the Vihi Valley, Kashmir. This is an especially interesting discovery, as it confirms the long disputed fact that the Glossopteris flora existed in India in Permo-Carboniferous times, which had previously only been inferred from indirect evidence. Mr. Seward 3 has described this plant as Gangamopteris kashmirensis, sp. nov. (c) Afghanistan and Persia. Griesbach 6 has described Vertebraria, the rhizome of Glossopteris, from rocks believed to be of Permo-Carboniferous age in Afghanistan, and Glossopteris from the province of Khorassan in Persia. ((/) Asia Minor. It is interesting to record that the most satisfactory specimens showing the fructification of Phyllotheca are those described by Professor Zeiller 7 from the Upper Carboniferous rocks of the coal- basin of Heraclea in Asia Minor. 1 Oldham (93), pp. 162-3, and two plates. 2 See Appendix. 3 Zeiller (02 1 ). i Noetling (03), p. 22. 3 Seward & Woodward (05) ; and the Appendix to this volume, p. 225. 8 Griesbach (85), pt. i, p. 62, 1885 : pt. ii, p. 59, 1886. 7 Zeiller (99 . J 1 HISTORICAL SKETCH. II. Eastern Asia, etc. (a) Tonquiu. The flora of the plant-bearing beds of Tonqnin has been described by Zeiller 1 in a series of papers published from 1882 onwards. Quite recently these results have been collected in the form of a monograph. 2 The beds are regarded as of Rhaetic age, and it will, therefore, only be necessary to notice such members of the Glossopteris flora as have been recorded in association with the Rhaetic flora. Glossopteris indica, Schimper. G. angustifolia, Brong. Nbeggerathiopsis Hislopi (Bimb.). Taheovittaria Kurzi, Feist. Teeniopteris spatulata, M'Clelland. T. cf. M'Clellandi (Oldham & Morris). Daiueopsis cf. Haghesi, Feist. = Sphenopteris cf. Hughesi (Feist.). (b) China. The fossil plants of Palaeozoic age described by Schmalhausen, 3 Newberry, 4 Schenk, 5 Abbado, 6 and Zeiller 7 from different parts of China are all typical members of the flora of the Northern Hemisphere, whereas the earlier memoirs of Newberry and Brongniart relate to fossils of apparently Jurassic age. Krasser 8 has figured several species from China and Central Asia collected by M. Obrutschew during the years 1893-1894. These plants are of different ages, and from various localities. Some of them, such as Lepidodendron cf. JIaidingeri, Ett., and Cordaites cf. principalis (Germ.), are members of the northern type of Permo-Carboniferous flora. But from San-schi-li-pu, in the province of Schen-si, two typical members of the Glossopteris flora were obtained, namely, Danceopsis Hughesi, Feist., and leaves resembling Noeggerathiopsis Hislopi (Bunb.). Zeiller (82 1 ), (82'-), (86 1 ), (86-). 2 Zeiller (02 :i ). Schmalhausen (83'), p. 432. 4 Newberry (83). Schenk (83), (85). 6 Abbado (00), p. 127. Zeiller (01), p. 431. 8 Krasser (00). HISTORICAL SKETCH. li Also in beds regarded as Rhoetic in age, near the village of Hsu-kia-ho, in the province of Sz'-tschwan, a Schizoneura was found which has since been named by Mr. Seward 1 S. Krasseri. Zeiller 2 has recently published an account of some fossils from Southern China. The Rhaetic flora of Tai-Pin-Tchang contains Glossopteris indica, Feist., in association with Cladophlebis Roesserti (Presl), a typical Rhsetic fern-like plant. Glossopteris indica is also recorded from Kiang-Ti-Ho, and Professor Zeiller states that this flora compares very closely with that of Tonquin. ((?) Borneo. Tenison-AVoods 3 has recognised in a collection of fossils from the Sarawak Coalfield in Borneo, Phjllothcca australis and Virtebraria. The latter is now known to be the rhizome of Glossopteris. III. Central and Northern Asia. The Altai and Siberia. Fossil plants from the Altai and Siberia have been described by Goppert, 4 Eichwald, 5 Geinitz, 6 Schmalhausen, 7 and Kosmovsky. 8 Schmalhausen regarded this flora as of Jurassic age, but Zeiller 9 has since shown that it is more closely related to that of the Permian rocks of Europe and North America. Phyllotheca, however, is associated with this flora; one species, P. deliquescens (Gopp.), described by Goppert and Schmalhausen, being also recorded from Australia. 10 Zeiller 11 has also compared P. Stschuronowskn, Schmal. with P. robusta, Feist, from the Lower Gondwanas of India. In a recent paper, the same author 12 has concluded that the plants described by Goppert and Geinitz under the name Noeggerathia, and by Schmalhausen as Rhiptozamites, should be referred to Cordaites. 1 Seward (03 ), p. 48. • Zeiller (00-). 3 Tenison-Woods (85), p. 584. J Goppert (45). 5 Eichwald (60). 6 Geinitz (69), (71). " Schmarhausen (79), (83 1 ). 8 Kosmovsky (91) ; see Zeiller (96-), p. 48 >. 9 Zeiller (96'-), (02-). 10 Arber (02 1 ), pp. 17, 22. 11 Zeiller (96-'), p. 472. '- Zeiller (02 2 ), pp. 889-90 ; (96-), p. 486. Ill HISTORICAL SKETCH. IV. Europe. (a) Russia. In 1897, Amalitzky 1 announced the discovery of Glossopteris in the Upper Permian beds of the Soukhona and Petite Dwina, Russia. Attention was called to this discovery by a short note published by Zeiller 2 in the following year. In 1901, Amalitzky 3 recorded several further members of the Glossopteris flora from beds inter- mediate between a Lower Permian horizon, with Lep telodendron and Callipteris eonferta, and sediments containing Upper Zechstein Mollusca. Glossopteris indica, Schimper. G. angustifolia, Feist. G. strieta, Bunb. Gungamopteris major. Feist. = G. cyclopter aides, var. major. Feist. G. eyclopteroides, Feist. Vertebraria. It may perhaps be mentioned that Schmalha.u sen i had already recorded Phyllotheca from the Permian of Northern Russia. (b) Spitsbergen. In 1897, Messrs. Newton & Teall described a few plants collected between Cape Grant and Cape Stephen in Franz Josef Land. Among them was one named Rhiptozamites (?) cf. Gopperti, 5 a species somewhat similar to certain leaves occurring in the rocks of Gondwanaland. It seems, however, more probable that this fossil is identical with the Mesozoic type of leaf now known as Photnicopsis elongatus (Morris). 6 The other associated specimens recall the plant -remains of a somewhat later period than the Permo-Carboniferous, and in all probability are not of that age. V. Australasia. («) New South Wales. Glossopteris Broivniana was first described from Australia by Brongniart 1 in 1828 from specimens derived from the coal-mines 1 Amalitzky (97). - Zeiller (98-). 3 Amalitzky (01), p. 592. 4 Schmalhausen 7!' . 5 Newton & Teall (97), p. 504, pi. xli, figs. 6, 7. ,; Seward (03'), p. 67. ' Brongniart (28'), p. 54 : (28 3 ), p. 223, pi. lxii, figs. I. la. HISTORICAL SKETCH. lili on the Hawkesbury River, 10 miles to tbe north of Port Jackson. From the same locality the fern-like plant, figured by Brongniart ' as Pecopteris data, and now known as Sphenopteris alata (Brong.), was recorded. The genus Pkyttotheca, and the species P. australis, were also founded by Brongniart 2 from Australian specimens. In 1831, Nicol 3 described sections of petrified woods from Tasmania, and two years later he figured other sections from New South Wales. Tbe occurrence of petrified wood in Australia had also attracted the attention of Clarke 4 among the earlier observers. The first collection of fossil plants from New South Wales was described by Morris 5 in 1845. These specimens are now partly in the British Museum, and partly in the Museum of the Geological Society of London. The following species were obtained from the Newcastle Series: — Sphenopteris lobifolia, sp. nov. S. alata, var. exilis, Morr. = S. alata (Broug.). Glossopteris Browniana, Broug. Phyllotheca aus trails, Broug. Some two years later, McCoy 6 published an important paper dealing with the plants obtained from several different horizons in New South Wales by the Rev. W. B. Clarke during the years 1839 to 1844. The Clarke collection was presented to the Woodwardian Museum, Cambridge, in 1844, and is now preserved in the Sedgwick Museum of that University. These specimens were re-examined and described more fully in 1902. 7 The following species were determined from the Newcastle Series by McCoy : — Glossopteris Browniana, Brong. G. linearis, sp. nov. = G. Browniana, Brong. Vertebraria australis, sp. nov. = V. inclica, Boyle. Sphenopteris alata (Brong-.). S. ger mantis, sp. nov. = S. lobifolia, Morris. S. plumosa, sp. nov. = S. lobifolia, Morris. S.flexuosa, sp. nov. = S. lobifolia, Morris. S. hast at a, sp. nov. = S. lobifolia, Morris. S. lobifolia, Morris. Gangamopteris angustifolia, sp. nov. Phyllotheca ramosa, sp. nov. = P. australis, Brong. 1 Brongniart (28 1 ), p. 58 ; (28 2 ), p. 3C1, pi. cxxvii. See Arber (02 1 ), p. 10. : Brongniart (28 1 ), pp. 152, 175. 3 Nicol (31), (33). 4 Clarke (43). 5 Morris (45). « McCoy (47). ' Arber (02'). liv HISTORICAL SKETCH. P. Bookeri, sp. nov. (pars) = P. australis, Brong. P. Hookeri, sp. nov. (pars) = P. deliqucscens (Giipp.). Zeugophyllites elongatus (Morris) = Noeggerathiopsis Hislopi (Bunb.). McCoy also described two species from beds at Arowa, on a lower horizon than tbo Newcastle Series : — Otopteris ovata , sp. nov. = Aneimites ova/a (McCoy). Phyllotheca Hookeri, sp. nov. = P. deliquescent! (Gopp.) . The two following species were recorded in 1902, as new from New South Wales, from among the undescribed specimens in the Clarke collection at Cambridge : — l Sphennpteris polymorpha, Feist. Car dioear pits, sp. Dana, 2 in 1849, published a further list of fossils from New South Wales, collected during the United States Exploring Expedition between the years 1838 and 1842 : — Noeggerathia spatulata, sp. nov. = Noeggerathiopsis Hislopi (Bunb.). N. media, sp. nov. = Noeggerathiopsis Hislopi (Bunb.). N. elongata, sp. nov. (nm Morris) = Noeggerathiopsis Hislopi (Bunb.). Sphenopteris lobifolia, Morris = ? 8. lobifolia, Morris. Glossopteris Browniana, Brong. G. ampla, sp. nov. G. reticulum, sp. nov. = G. Browniana, Brong. G. elongata, sp. nov. = G. Browniana, Brong. G. (?) cordata, sp. nov. = G. ampla, Dana. G. linearis, McCoy = G. Browniana, Brong. Phyllotheca australis, Brong. Clasteria australis, sp. nov. = Vertebraria indica, Royle. Anarthrocanna australis, sp. nov. = ? Cystoseirites (?) = ? roots. Austrella rigida, sp. nov. = ? roots. Confervites (?) tenella, sp. nov. = ? roots. Feistmantel, 3 in 1878, described the following species from the "Lower Coal Measures" of New South Wales, which form the lower series of Ghssopten's-bear'mg rocks in that Colony : — Glossopteris Browniana, Brong. G. Browniana, var. precursor, Feist. = G. Browniana, Brong. G. primeeva, sp. nov. = cf. G. indica, Sehimper. G. Clarkci, sp. nov. = cf. G. Browniana, Brong. G. elegans, sp. nov. Noeggerathiopsis prisca, sp. nov. = cf. N. Hislopi (Bunb.). 1 Arber (02 1 ). 2 Dana (J9). 3 Feistmantel (78). HISTORICAL SKETCH. Iv In addition to fossils previously recorded, Feistmantel described the following species from the Newcastle Scries of New South Wales : — Glossopteris teeni'opteroides, sp. nov. = G. indica, Schimper. G. Wilkinsoni, sp. nov. = G. Browniana, Brong. G. parallela, sp. uov. = cf. G. Browniana, Brong. Gang a mo pterin Clarkeana, sp. nov. = G. cyclopteroides, Feist. Gaulopteris (?) Adamsi, sp. nov. = \Caulopteris?~\ Adamsi, Feist. Brachijphijlliim (?) austral?, sp. uov. In 1883, Tenison -Woods ' published an account of the literature and previous records of fossil plants from the coal deposits of Australia. Several new species, some of which undoubtedly require revision, were described by the same author, but these appear to belong to the Triasso-Rhsetic or Jurassic floras, and consequently need not be considered here. Feistmautel, 2 in 1890, contributed an important monograph on the Palaeozoic and Mesozoic plants of Australia, founded on his previous work published in 1878. It contains a full account of the literature on the subject, with descriptions and, in many cases, figures of the Glossopteris flora from New South Wales, and elsewhere in Australasia. The following new species were described : — Glossopteris gangamopteroides, sp. nov. = G. Browniana, Broug. G. spathulato-cordata, sp. nov. = cf. G. orbicularis, F'eist. About this period, the occurrence of organic remains in the kerosene shales of New South Wales received considerable attention at the hands of two French Palseobotanists, Professor Bertram! and the late Dr. Renault. A species of Reinschia 3 was described by these authors conjointly in 1892, and in 1896 Professor Bertram! described Fila australis.* In 1893, Mr. E,. Etheridge, jun., 5 described the first specimens of Schizoneura from New South Wales, further and more satisfactory examples being obtained in 1903. Schizoneura australis, Etheridge = S. gondicanensis, Feist. 1 Tenison- Woods (83). 2 Feistmantel (90). 3 Bertrand & Renault (92), (93) ; also Bertrand (93), (00). 4 Bertrand (96). 5 Etheridge (93), (03). lvi HISTORICAL SKKTCH. The same author 1 also described in 1895 an interesting fossil which he referred to the genus Phyllotheca, and which is here described as P. Etheridyei, sp. now In 1897, Dun 3 recorded two new species of Glossopteris from the Newcastle Series : — Glossopteris rectinervis, sp. nov. = cf. G. Browniana, Brong. G. acuta, sp. nov. (h) Victoria. The genus Glossopteris has not, so far as I am aware, been recorded from Victoria, although Gangamopteris has been known from the colony for many years. The first mention of the latter genus was apparently that by Selwyn 3 in 1866, who identified it from the Bacchus Marsh beds. In 1875, McCoy 4 described the following forms of Gangamopteris from the same horizon and locality : — Gangamopteris angustifolia, McCoy. G. spatulata, sp. nov. = G. cyclopteroides, Feist. G. obliqua, sp. nov. = G. cyclopteroides, Feist. In 1878 and 1890, FeistmanteP described the above species, without any further additions to the flora except Phyllotheca australis, Brong., from Cape Paterson, Victoria. In 1898, the late Sir Frederick McCoy 6 described a fragmentary portion of a Tseniopteroid frond as Tamiopteris Siceeti, which was obtained from the Bacchus Marsh Sandstone. (c) Queensland. In 1872, Carruthers 7 recorded Glossopteris Broicniana from Queensland. The same author, 9 in 1880, determined the following species in a collection of plants obtained by Mr. Jack from the Bowen River Coalfield, North Queensland : — Glossopteris Browniana, Brong. G. linearis, McCoy = G. Browniana, Brong. G. cf. G. ampla, Dana. Phyllotheca australis, Brong. Araucarioxylon Nicoli, sp. nov. = nomen nudum. 1 Etheridge (95). - Dun (97), p. 64. 3 Selwyn & Ulrich (66), p. 16. 4 McCoy (74). 5 Feistmantel (78), pp. 101-2; (90), pp. 130-33. 6 McCoy (98). ' Carruthers (72), p. 354. 8 Carruthers (80), p. 325. HISTORICAL SKETCH. lvii Jack & Etheridge 1 recorded the following species from Queensland in 1892:— Phyilotheca aus trails, Brong. Sphenopteris lobifolia, Morris. S. jitxuosa, McCoy = S. lobifolia, Morris. S. crebra, T. - Woods (?). Glossopteris Browniana, Brong. G. linearis, McCoy = G. Browniana, Brong. G. ampla, Dana. Araucarioxylon Nicoli, Carr. = nomen nudum. In 1898, Shirley 2 described a number of petrified -woods from Queensland. Araucarioxylon Maitlandi, sp. now = Ladoxylon Maitlandi (Shirley). A. Binneyi, sp. nov. = I). Binneyi (Shirley). A. H'illiamsoni, sp. nov. = D. Williamsoni (Shirley). A. brisbanense, sp. nov. = I). brisbanense (Shirley). More recently the same author 3 has figured several species new to Queensland. Glossopteris Browniana, var. precursor, Feist. = G. Browniana, Brong. G. communis, Feist. = G. indica, Schimper. G. parallela, Feist. = cf. G. Broicniana, Brong. G. elegans, Feist. G. Wilkiiisoni, Feist. = G. Browniana, Broug. Sphenopteris alata (Brong.). Cycadospermum Dawsoni, sp. nov. = Cardiocarpus (?) Daivsoni (Shirley). Noeggerathia (?) = ? Noeggerathiopsis, sp. A number of specimens from the Baron River, Queensland, in the National Museum, Melbourne, were described by Chapman in a paper published in 1904. 4 Phyilotheca australis, Brong. Glossopteris Browniaua, Brong. G. ampla, Dana. G. parallela, Feist. = cf . G. Browniana, Brong. Araucarioxylon Daintreei, sp. nov. = Dadoxylon Dainfreei (Chapman". 1 Jack & Etheridge (92), p. 189, etc. - Shirley (98). 3 Shirley (02). 4 Chapman (04). I did not have an opportunity of seeing this paper until most of the systematic portion of this work had passed through the press. Consequently these records are not noticed under the specific headings. lviii HISTORICAL SKETCH. This paper also includes a discussion on the fructification of Phyllotheca australis, and an account of the distribution of silicified woods in Australia. It has been stated by several observers that Glossopteris occurs in the Desert Sandstone of Queensland, a deposit of Cretaceous age. It was first recorded from these beds by Norman Taylor ' in the year 1874. In 1890, Rands 2 discovered some specimens, wbich were identified by Mr. Etheridge, jun. as undoubtedly fronds of Ghssopteris, and identical with specimens of Pe rmo-Carboniferous age, from beds resembling the Desert Sandstone, and stated to overlie the Rolling Downs formation, at a locality known as Betts Creek, within a mile of Conglomerate Gully in the Cape Goldfield. Dr. Jack 3 afterwards visited this locality, and emphatically confirmed the conclusion that the rocks developed there belong to the Desert Sandstone formation. Despite the authority of these statements, one cannot help feeling that this matter requires yet further confirmation, since this evidence is entirely opposed to that of all the other regions from which Lower Cretaceous plant-remains are known ; a flora remarkable for its worldwide uniformity of distribution, and one in which Glossopteris does not occur. That genus is unknown elsewhere, in beds of later age than the Rhastic. (d) Western Australia. The records of the Glossopteris flora from Western Australia are few in number. In 1893, the late Mr. Robert Etberidge, 4 sen., detected portions of Glossopteris and Noeggerathia in coal from the Collie River Coalfield. In the following year, Professor Edgeworth David 5 stated that Mr. B. H. Woodward had recognised Glossopteris Browyiiana from the Gascoyne River, West Australia. It maybe also added that Mr. Kidston 6 has figured imperfect fragments of a Lepidodendron, Stigmaria, and Cype rites-like leaf from Yarralla Hill, near the mouth of May River (King Sound). ' See Jack & Etheridge (92), p. 559. 2 Rands (91), p. 10 ; see also Jack & Etheridge (92), p. 518. 3 Jack & Etheridge (92), pp. 518-520, 558 ; Etheridge & David (94), p 4 Etheridge, sen. (93), p. 241. 5 David, in Etheridge & David '94). p. 256. 6 Kidston (90), p. 102, pi. iv, tigs. 4, la, 5, (3, 6*(. 7, 7". 8, 8«. HIST0K1CAL SKETCH. lix So far as I am aware, no members of the Glossopteris flora have been discovered as jet in South Australia. (c) Tasmania. The earlier discoveries of plant-remains in Tasmania were of fossils chiefly Khaetic in age. The Glossopteris flora was first recognised by Brough Smyth and McCoy 1 in 1874. Johnston, 2 in 1886, mentions the following species as occurring in the Mersey River and Don Coal-basins : — Glossopteris Browniana, Brong. G. ampla, Dana. Gangamopteris spathulala, McCoy = G. ct/clopteroides, Feist. G. angustifolia, McCoy. G. obliqua, McCoy = G. ct/clopteroides, Feist. G. Clarkeana, Feist. = G. cydopteroides, Feist. Noeggerathiopsis media '(?), Dana = N. Hislopi (Bunli.). N. spathulata, Dana = A T . Hislopi (Bunb.). iV. elongata, Dana = JY. Hislopi (Bunb.). iV. prisca, Feist. = cf. N. Hislopi (Bunb.). Schizoneura, sp. (?) = ? Carpolithus (?) tasmanicus, sp. nov. = ? Cardiocarpus, sp. Tasmanites punctatus, Newton. In rocks regarded as Triassic in age, the following species, common also to the Permo-Carboniferous flora, Avere found: — Phyllotheca australis, Brong. Sphenopteris alata (Brong.). S. plamosa, McCoy = S. lobifolia, Morris. S. lobifolia, Morris. Johnston, 3 in 1887, recorded the following species from Bruin Island, Southern Tasmania : — Gangamopteris spathulata, McCoy = G. ct/clopteroides, Feist. G. obliqua, McCoy = G. cyclopteroides, Feist. Glossopteris Browniana, var. prccvnrsor, Feist. = G. Browniana, Brong. In the same year Johnston i also described a number of new species, but the age of the beds from which they were obtained is not very clear, and none of the specimens were figured. In his work on the Geology of Tasmania, 6 published in 1888, figures of 1 See Smyth (74), p. 24. 2 Johnston (86), p. 362. •* Johnston (87 1 ), p. 21. i Johnston (87-), p. 160. 5 Johnston (88), pp. Ill, 134, pis. viii-x. lx HISTORICAL SKETCH. Tasinanian specimens of Glossopteris, Gangamopteris, and other- genera are given. In 1890, Feistinantel ' published the following list of plants from the Mersey River Beds : — Thyllothcca australis, Biong. Glossopteris communis, Feist. = G. indica, Schiruper. G. Browniana, Brong. 67. spathulato-cordata, Feist. = cf. G. orbicularis, Feist. G. reticulum, Dana = G. Browniana, Brong. Gangamopteris cyclopteroides, Feist. G. cyclopteroides, var. subauriculnta, Feist. = G. cyclopteroides, Feist. G. cyclopteroides, var. attenuata, Feist. = G. cyclopteroides, Feist. G. angustifolia, McCoy. G. spathulata, McCoy = G. cyclopteroides, Feist. G. obliqua, McCoy = G. cyclopteroides, Feist. Noeggerathicpsis Hislopi (Bunb.). N. spathulata (Dana) = N. Hislopi (Bunb.). Squamae gymnospermarum. Samaropsis, sp. (?). In 1892, Johnston 2 described some further plants from the Henty River. Glossopteris Browniana, Brong. G. ovata, sp. nov. — ? G. ampla, Dana. Johnston, 3 in 1894, described a new plant from Ida Bay, South- port, Tasmania, as Pecopteris lunensis, which was found associated with Vertebraria australis. He states that this species occurs in beds of both Permo-Carboniferous and Mesozoic age in Tasmania. Judging from his drawing, it would appear to be of a type more in common with members of the Triassic flora than with the Permo-Carboniferous. Two years later, 4 the same author gave a complete list of the Palaeozoic and Mesozoic plants known from Tasmania, and also figures a fine specimen of Sphenopteris lobifolia, Morr., from Seymour, under the name, S. Iforrisiana. 1 Feistmantel (90), p. 60. 2 Johnston (92). s Johnston (94), pt. i, p. 170, pi. i, figs. 5-7. 4 Johnston (94), pt. ii, table opposite p. 62, and p. 58, figs. 14. 15. HISTORICAL SKETCH. I XI (,/') New Zealand. The Glossopteris flora is not known with certainty from New Zealand. 1 Sir James Hector, 2 in 1878, stated that the heds of Mount Potts are full of the leaves of Glossopteris. Further, in 1886, he says that "at the base of the Kaihiku Series are the Glossopteris heds of Mount Potts, and in the Kaihiku district Glossopteris occurs in the lower beds as developed in Popotunea Gorge." Crie 3 has also recorded Glossopteris from beds believed to be of Triassic age at Wairoa. Neither of these discoveries has been since confirmed, and the other genera known from New Zealand appear to be typical of a later period than the Permo-Carboniferous. VI. South Africa. {a) Cape Colony, Natal {including Zululand), and the Orange River Colony. Apparently the earliest record of fossil plants from South Africa is that of specimens discovered by Bain in the Roggeval (Fish River), which were described by Hooker 4 in 1856, and are here noticed under the name Schizonetcra(?) africana, Feist. Wyley 5 referred the beds from which these fossils were obtained to the Koonap division of the Karoo Series. Plant-remains from Natal were mentioned by Sutherland as far back as 1855. 6 The first discovery of the Glossopteris flora in South Africa was, however, made by Rubidge 7 in 1859, who obtained specimens from Bloemkop, which he compared with those occurring in India. Tate, B in a paper published in 1867, figured Glossopteris and other genera from South Africa. These specimens are now in the Museum of the Geological Society of London. The chief localities were Heald Town (near Fort Beaufort), Bloemkop (near the Sunday's River), Graaf Reinet, and East London. ' See Etheridge & David (94), p. 2,30. 2 Hector (78), p. 533 ; (86), p. 7 3 Crie (88), p. 1014. 4 Hooker (56), p. 227, pi. xxviii, fig-. 1. 5 See Tate (67), p. 172. 6 Sutherland (55), p. 4 66. 7 Rubidge (59), p. 19S. s Tate (67), p. 140. lxii HISTORICAL SKETCH. Glossopferis Browniana, Brong. G. Satherlandi, sp. nov. = G. Browniana, Brong. Rubidgea Machayi, sp. nov. = ? Dictyopteris (?) simplex, sp. nov. = ? Glossopteris retifera, Feist. Fhyllotheca, sp. = ? In a synopsis of the Karoo Beds by Professor Eupert Jones, included in Tate's paper, some of the above-mentioned species are stated to be derived from the Beaufort Series. Fossil wood is also mentioned as common. In the same paper, Lepidodendron, etc., are recorded from the Wittenberg, Znurberg, and elsewhere by Wyley. 1 This genus had previously been stated to occur at Jackal's Kop, on the eastern side of the Stormberg range, by Kubidge, 2 who also mentions Calamite-like plants from the western part of the Zuurbergen. It is doubtful, however, if these early determinations of Lepidodendron were correct, and the doubt is even stronger with regard to the discoveries next to be described. Grey, 3 in some remarks on South African plants, published in 1871, recorded Lepidodendron, Sig Maria, and Pecopteris from beds associated with the coal at Andries's Nek, on the north-east margin of the Stormberg range, and 25 miles north-east of Queenstown. Carruthers examined these specimens, and determined the Lepido- dendron to be L. crenatum (Sternberg), which, he observed, was associated with " stems of Catamites, perhaps of three species, such as those which have a very slender periphery." Carruthers also records Pecopteris Cistii (Brong.) ?, " Alethopteris Longchitidis, Stbg., and Asterophyllites equisetiformis, Brg." Grey mentioned a Stigmaria from " the Carboniferous rocks of Lower Albany," but Carruthers decided that, although this has somewhat the appearance of a Cyclodigma, it is a true Sigillaria. Professor Rupert Jones 4 also mentions that Mr. Bristow had recognised Sigillaria, Stigmaria, Lepidostrobus, JLalonia, and Selagi?iites in the micaceous shales of Port Alfred, on the Albany Coalfield. These conclusions were reviewed by Zeiller 5 in 1883. Griesbach, 6 in 1871, also recorded Calamites, Equisetum, and ' See Tate (67), p. 173. 3 Rubidge (56), p. 237. 3 Grey (71). * See Grey (71), p. 51. 5 Zeiller (83). ,; Griesbach (71/, p HISTORICAL SKETCH. lxiii Zepidodendron from the sandstones associated with the coal at Tulberg, Cape Colony. It must, however, be pointed out that, despite the records of Rubidge, Wyley, Grey, Carruthers, and Griesbach, there is still grave doubt whether such typical genera of the Northern type of Permo-Carboniferous flora as Calamites, Zepidodendron, etc. (except SigiUaria and JBothrodendron), have been really found to occur in South Africa in all these localities. Professor Rupert Jones 1 has expressed hesitation as to whether the plants described by Grey were really derived at all from South Africa, and Mr. Seward 8 has recently supported this conclusion. There are specimens in the British. Museum collections (V. 235, V. 2390, and V. 3267) of Zepidodendron, Calamites, and Calamocladus, said to have been derived from the Stormberg Beds, which seem to me to be preserved in shales very unlike any rocks which I have seen from South Africa, and more closely similar to the Coal Measure deposits of this country. It is possible that in several cases plant - remains which we should now assign to the genus Phyllotheca have been mistaken for Calamites. 3 It is also probable that badly-preserved specimens of SigiUaria or Bothrodendron may have been identi6ed in some cases as Zepidodendron. These two genera, characteristic of the Northern type , V. 7101//, V. 710U (Bard wan Coalfield, India). Fronds of Glossopteris angusti folia from South Africa. V. 3613. Figured by Seward (97 1 ), p. 321, pi. xxi, fig. 4a. An imperfect fragment of a linear frond, about 11 '5 cm. long, in association with Noeggerathiopsis Iiislopi. Casey's Township, Transvaal. Pres. bg D. Draper, Esq., 1897. V. 3132tf. Among a number of scattered fronds shown on this specimen are some probably belonging to this species, associated with G. Browniana and G. indica. Natal. Pres. by the Natal Government, 1894. 4. Glossopteris stricta, Bunbury. (PI. IV, Fig. 1.) 1861. Glossopteris stricta, Bunbury, Quart. Journ. Geol. Soc, vol. xvii, p. 331, pl. ix, fig. 5. 1869. A»giopteridium (?) strictum, Schimper, Traitc, vol. i, p. 606. 1876. Glossopteris stricta, Feistmantel, Eec. Geol. Surv. India, vol. ix, pt. 3, pp. 73, 74. 1880. G. stricta, Feistmantel, Flora Gomlw. Syst., vol. iii, pts. 2, 3, p. 100, pl. xxxviiA, fig's. 1-2 ; pl. xxxviiiA, fig. 3. 1882. G. stricta, Feistmantel, ibid., vol. iv, pt. 1, p. 33, pl. xxi, fi»'. 11. 1889. ? G. stricta, Feistmantel, Abhand. b6hm. Gesell. Wiss. Prag, sir. vu, vol. iii, p. 4.5, pl. iv, figs. 6, 6a. 1901. G. stricta, Amalitsky, Compt. Rend., vol. cxxxii, p. 591. Tgpe. No. II. 10,363, Museum Geol. Soc. London. GLOSSOPTERIS. 77 Frond large, elongate-lanceolate. Apex acuminate. Midrib strong. Secondary nerves arched near their point of origin, and forming two or more series of comparatively broad and short polygonal meshes bordering on the midrib, then bent abruptly, and passing very obliquely towards the margin in a direction almost perpendicular to the midrib. Except near the midrib, the secondary nerves are numerous and close, forming very narrow and trans- versely-elongate meshes. The fronds measure 25-35 cm. or more in length, and are often 4 cm. broad. This species is distinguished from 6. indioa by the strictly lanceolate form of the frond, tapering more gradually towards the apex and base ; by the contrast in the size and shape of the meshes bordering on the midrib with those of the rest of the lamina ; and by the very oblique course of the veins towards the margin at a short distance from the midrib. It is only known from a very few localities in India, of which the Nagpur district is the more important. It is somewhat doubtful if the fronds from Cape Colony ascribed to this species by Feistinantel are identical. G. stricta has, however, been recently reoorded by Amalitsky from the Permian (Northern Type) of Kussia. V. 7145. PL IV, Fig. 1. A slab showing four almost perfect fronds, of which the most nearly complete is 25 cm. long. Part of one of these is figured on PI. IV, Fig. 1, showing the midrib composed of a bundle of parallel nerves, and the characteristic lateral nervation. The basal portions of several other leaves are also seen. Silewada, 12 miles north of Nagpur, India. Hunter Coll, 46,705. A large slab sb owing several fairly complete fronds, one of which is 23 cm. long, and almost perfect except the apex. The contracted base is well seen. The midrib is broad, and is made up of many parallel, anastomosing nerves. The lateral nervation is strongly arched. Another almost complete frond on the same slab measures 33 cm. long, and is 4 cm. broad at the widest point. The nervation is less distinct than in the former specimen. The margin is in places slightly incurved. Several basal portions of fronds also occur on the same slab, as well as a median portion showing the nervation excellently. Silewada, India. Hunter Coll. <0 GLOSSOFTKMS. 43,705/'. A slab showing a number of fronds in which the nervation is somewhat indistinct, some of which probably belong to G. stricta, while others may possibly be referred to other species. Silewada, India. Hunter Coll. V. 7146. Several fragments of long, lanceolate fronds, probably of this species, showing the nervation clearly. The midribs are thick. The broad meshes bordering on the midrib are well marked. Near Nagpur, India. Hunter Coll. V. 7144. Very fragmentary fronds, but interesting as probably belonging to this species. The comparatively broad meshes near the midrib are well seen, also the bundles of parallel, anastomosing nerves of which the midrib is composed. Silewada, India. Hunter Coll. 5. Glossopteris ampla, Dana. (Text-fig. 20.) 1849. Glossopteris ampla, Dana, in Wilkes' U.S. Explor. Exped , vol. x, p. 717, pi. xiii, figs, la, lb. G. (?) cordata, Dana, ibid., p. 718, pi. xiii, fig. 5. 1861. G. muscefolia, Buubury, Quart. Jourri! Geol. Soc, vol. xvii, p. 329, pi. viii, fig. 6. 1869. Macrotesitiopteris mmce-folia, Scliimper, Traite, vol. i, p. 612. 1876. Glossopteris mime folia, Feistniantel, Rec. Geol. Surv. India, vol. ix, pt. 3, pp. 73, 74. 1878. G. ampla, Feistniantel, Paheontogr., Suppl. iii, p. 91, pi. xi, fig. 2 ; pi. xii, fig. 7. G. cordata, Feistniantel, ibid., p. 92. 1879. G. damudica, Feistniantel, Flora Gondw. Syst., vol. iii, pt. 1, p. 17. 1880. G. damudica, Feistniantel, ibid., vol. iii, pts. 2, 3, p. 105, pi. xxxa, figs. 1, 2 ; pi. xxxi.v, figs. 1-3 ; ? pi. xxxua, fig. 1 ; ? pi. xIa, fig. 6. G. lousrcfolia, Feistniantel, ibid., p. 101. 1882. G. damudica, Feistniantel, ibid., vol. iv, pt. 1, p. 35. 1883. G. ampla, Tenison- Woods, Proc. Linn. Soc. Xew South Wales, vol. viii, p. 124. G. (?)cordata, Tenison-Woods, ibid., p. 124. 1886. G. ampla, Johnston, Papers and Proc. 11. Soc. Tasmania for 1885, p. 377. G. cordata, Johnston, ibid., p. 378. II. damudica, Feistmantel, Flora Gondw. Syst., vol. iv, pt 2, p. 28, pi. iA, fit;-. 3 ; pi. iva, tig. 1 (right-hand figure) ; pi. va, fig. 6. 1888. G. ampla, Johnston, Geol. Tasmania, p. Ill, pi viii, fig. 3; Ppl. ix, tig. 3; Ppl. x, fig. 1. GLOSSOPTEltlS. 79 1S89. Glossopteris damudica, var. stcnomura, Feistmantel, Abhand. bohm. Gesell. Wiss. Prag, ser. vn, vol. iii, p. 46, pi. iv, figs. 7, 7a. 1890. G. ampin, Feistmantel, Mem. Geol. Surv. New South Wales, Pal., No. 3, p. 122, pi. xix, figs. 1, 2. G. cordata, Feistmantel, ibid., p. 124. 1892. G. ampla, Jack & Etheridge, Geol. and Pal. Queensland, p. 195, pi. xv, fig. 7. ? G. ovata, Johnston, Papers and Proc. R. Soc. Tasmania for 1891, p. 13. 1901. G. damudica, var. slenonenra, Etheridge, jun., in Anderson, 1st Rep. Geol. Surv. Natal, p. 70. 1902. G. damudica, Zeiller, Pal. Indica, n.s., vol. ii, p 13, pi. iv, figs. 5 (pars), 6, 7. G. ampla, Arber, Quart. Journ. Geol. Soc, vol. lviii, p. 7. Types. The whereabouts of Dana's type is unknown to me. Type of G. muscefolia, Bunbury, INV R. 10,359, Mus. Geol. Soc. London. Types of G. damudica, Feistmantel, Xos. 5262-6, 5306, Mus. Geol. Surv. India, Calcutta. Frond usually large, broadly obovate. Apex obtuse or emar- ginate. Midrib stout, especially in the lower portion of the frond, but not persisting quite to the apex. Lateral nerves arched at midrib, forming one or two series of comparatively broad and short meshes, and then subdividing into a number of close, almost parallel veins, often very oblique, forming extremely narrow, elongate meshes. The characters which especially distinguish this species are the breadth of the fronds, the stoutness of the midrib, and the contrast between the fine, close, and pseudo-parallel nervation of the greater portion of the lamina, and the nervation bordering on the midrib. It seems clear from a re- examination of these broad -leaved Glossopterids that G. damudica, Feist., must be united with Dana's G. ampla. Although Feistmantel's name is the best known, Dana's, as being the older, should have priority. The only difference which I can find between these fronds is that the nervation in the Indian leaves (G. damudica) is rather more acute than in the Australian specimens, where the obliquity is very marked in the greater portion of the frond. I am unable, however, to attach any weight to this character. I have had an opportunity of comparing the fine specimens of the Australian fronds in the Clarke Collection of the Sedgwick Museum, Cambridge, with the 80 GLOSSOPTER1S. British Museum examples of the Indian frond, and this comparison has tended to confirm this conclusion. I have recently re examined the type-specimen of Bunhury's Fig. 20. — Glossopteris ampla, Dana. After Feistmantel. -f nat. size. G. muscefolia, which I believe to be identical with Feistmantel' s species and with G. ampla. As in other species of Glossopteris, the fronds of G. ampla show CLOSSOFTERIS. 11 considerable variations in size. Zeiller 1 has recently figured some Indian specimens, barely 2 cm. broad, pointing out, however, that the type of nervation remains fairly constant. The specimen figured by Dana 2 as G. (?) cordata is simply a basal portion of a frond of this species. Dana himself admitted that the nervation is very similar to that of G. ampla. I think it probable that the frond termed G. orata by Johnston 3 should also be included here. No figure has been published of this Tasmanian leaf, and it is therefore not easy to arrive at any decision on this point, but the author states that it appears to differ only in details from G. ampla. Johnston's specimen was obtained from the Henty River, Tasmania. Distribution. — Permo- Carboniferous (Glossopteris flora) : — India, in the Talchir and Damuda divisions ; New South Wales, in the Newcastle Series ; Queensland ; Tasmania (Mersey River) ; Cape Colony, Natal, ? Rhodesia. Fronds of Glossopteris ampla from India. V. 7141. A fragment of a frond, measuring 7 - 5 cm. in length, and rather more than 5 cm. across. The midrib is stout. The characteristic nervation of this species, consisting of a few rows of comparatively broad meshes bordering on the midrib, and a much finer, closer, and sub-parallel nervation towards the margin, is well seen in this specimen. Nagpur, India. Hunter Coll. 46,705ff. A fine frond, measuring 27 cm. long and 8 cm. broad at its widest part. The midrib is stout, the lateral nervation very oblique and almost at right angles to the midrib. The larger meshes near the midrib are well seen. Portions of other fronds occur on the same specimen, which probably also belong to this species ; one being a fairly complete leaf, and another a basal portion of a frond. Silewada, 12 miles north of Nagpur, India. Hunter Coll. 1 Zeiller (02 1 ), p. 14. pi. iv, figs. 5 (pars), 6, 6a. 2 Dana (49), p. 718. pi. xviii, fig. 5. s Johnston (92), p. 13. 82 GLOSSOPTKRIS. Fronds of Ghssopteris ampin from Australia. V. 2108rY. A specimen showing several imperfect fragments of large fronds probably identical with this species. One of these measures 23 cm. in length, and 12 - 5 cm. across. The meshes are very elongate, the nerves being close, and almost parallel. Pres. by Sir C. Purdon Clarke, 1889. V. 2108/>. Similar fragments in which the nervation is more clearly preserved. The meshes are broader near the petiole, but very elongate towards the margin. Pres. by Sir C. Purdon Clnrke, 1889. V. 2109. Fragments of large fronds probably belonging to this species. Pres. by Sir C. Purdon Clarke, 1889. Fronds of Ghssopteris ampin from Tasmania. V. 3776/.:. Part of a broadly oval frond, 12-5 cm. long by 10 cm. broad. The midrib is strongly marked, but the lateral nervation is not very clear. Fragments of Gangamopteris also occur on the same specimen. Mersey River. Pres. by T. Stephens, Esq., 1898. V. 3776/>. A slab of shale with many impressions of Ghssopteris and other leaves, among which there is a frond probably belonging to this species. Mersey River. Pres. by T. Stephens, Esq., 1898. V. 3776«. Fragments of fronds of this species in association with G. indica. Mersey River. Pres. by T. Stephens, Esq., 1898. V. 3776/. Part of a frond, probably of this species. Mersey River. Pres. by T. Stphens, Esq., 1898. Frond of Ghssopteris ampla (?) from Rhodesia. V. 7593. This specimen shows a broad fragment of a frond, probably an apical portion, measuring 8 cm. in length, with a maximum breadth of 4*5 cm. The nervation is fairly well preserved, the lateral nerves being acute and pseudo-parallel, forming very narrow and elongate meshes. Sisi Siding, Peehuanaland Railway. Pres. by A. J. C. Molyneux, Esq., 1901. GLOSSOPTERIS. 83 Frond of Glossopteris, sp., from Natal. V. 2903. This specimen is labelled Tamiopteris, but it is obviously a Glossopteris, since the lateral nerves clearly anastomose. It is a broad frond, which must have been 10 cm. or more across. The midrib is missing. The nerves are nearly parallel, and anastomose apparently at long intervals. Although it is not sufficiently perfect to identify specifically, it may be compared with G. ampla. Farm Glencalder (at 6,700 feet s.m.), Newcastle Division. Pres. by B. Draper, Esq., 1893. 6. Glossopteris retifera, Feistmantel. (Text-fig. 21.) 1867. ? Dietyoptcris (?) simplex, Tate, Quart. Journ. Geol. Soc, vol. xxiii, p. 141, pi. vi, fig. 6. 1876. Sagenopteris polyphylla, Feistmantel, Journ. Asiat. Soc. Bengal, vol. xlv, pt. 2, p. 377, pi. xx, figs. 5, 6. 1880. Glossopteris retifera, Feistmantel, Flora Gondw. Syst., vol. iii, pts. 2, 3, p. 103, pi. xxviiiA, figs. 2, 7, 10; pi. xliA, fig. 9. Sagenopteris polyphylla, Feistmantel, ibid., p. 113, pi. xliA, figs. 3, 4. 1882. Glossopteris retifera, Feistmantel, Flora Gondw. Syst., vol. iv, pt. 1, p. 35. 1886. G. retifera, Feistmantel, ibid., vol. iv, pt. 2, p. 29, pi. iva, fig. 1 (left-hand figure). 1889. G. retifera, Feistmantel, Abhand. bohm. Gesell. Wiss. Prag, ser. vn, vol. iii, p. 46, pi. iv, tig. 3. ? G. Tatei, Feistmantel, ibid., p. 44, pi. iv, fig. 8. 1896. G. retifera, Bodenbender, Zeitschr. deutsch. geol. Gesell., vol. xhiii, table opposite p. 772. 1901. G. retifera, Etheridge, jun., in Anderson, 1st Kep. Geol. Surv. Natal, p. 70, pi. xiii, figs. 7, 8. Type. jSTos. 5243, 5251, Mus. Geol. Surv. India, Calcutta. Frond petiolate, of medium size, lanceolate, or oval-lanceolate. Apex acuminate, or (?) obtuse. Midrib distinct, longitudinally furrowed. Lateral nerves arched, forming open, broadly polygonal mesbes, not much longer than broad, and approximately of equal size throughout the lamina. Feistmantel ' has figured, under the name of Sagenopteris poly- phylla, impressions of several Glossopterid fronds radiating from 1 Feistmantel (80), pi. xliA, figs. 3, 4. 84 GLOSSOPTERIS. a stem-like structure. I have no hesitation in regarding these as fronds of G. retifera attached to a rhizome. Zeiller 1 has already Fig. 21. — Giossopteris retifera, Feistmantel. After Feistmantel. Nat. size, suggested that these specimens should be included in the genus Giossopteris. 1 Zeiller (%'), p. 371. GLOSSOPTERIS. 85 The imperfect specimen with a very oblique nervation figured by Tate ' in 1867 as Dictyopteris (?) simplex, may be compared with this species as regards the characters of the meshes. A re- examination of the type, preserved in the Museum of the Geological Society of London (No. R. 11,104), has shown that it is probably a frond of Glossopteris, but that the fossil is too fragmentary to permit of specific determination. Distribution. — Pernio- Carboniferous (Glossopteris flora) : — India, in the Damuda division ; Cape Colony, Natal, Orange River Colony, and Argentina. Fronds of Glossopteris retifera from Natal. V. 2904. Small fragments occur on this specimen in association with G. Browniana. Farm Glencalder (at 6,700 feet s.m.), Newcastle Division. Pres. by D. Draper, Esq., 1893. V. 29015. Several well-preserved fragments showing the nervation clearly. One of these is a median portion of a leaf, 5 cm. long and 3 cm. across. Farm Glencalder (at 6,700 feet s.m.), Newcastle Division. Pres. by D. Draper, Esq., 1893. V. 2904c. Basal portions of fronds of this species in association with G. Browniana. Farm Glencalder (at 6,700 feet s.m ), Newcastle Division. Pres. by D. Draper, Esq., 1893. Fronds of Glossopteris retifera from the Or any e River Colony. V. 2902a. A number of fragments of fronds of this species in association with those of G. conspicua. Mill River Ford, Harrisrnith. Pres. by D. Draper, Esq., 1893. V. 2465. Among the fronds shown on this specimen there are some possibly identical with G. retifera, in association with fronds of G. Broivtiiana and G. conspicua. Mill River Ford, Harrismith. Pres. by D. Draper, Esq., 1890, 1 Tate (67), pi. vi, fig. 2 ; Feistmautel (89), pi. iv, fig. 8. 86 GLOSSOPTERIS. 7. Glossopteris conspicua, Feistinantel. (Plate III, Fig. 3 ; Text-fig. 22.) 1881. Glossopteris conspicua, Feistmantel, Flora Gondw. Syst., vol. iii, pt. 3, p. 104, pi. xxviiiA, figs. 1, 5, 6, 8, 9. 1886. G. conspicua, Feistmantel, ibid., vol. iv, pt. 2, p. 29. 1893. G. conspicua, Oldham, Man. Geol. India, 2nd pi. opp. p. 162. Fig. 22. — Glossopteris conspicua, Feist. After Feistmantel. Nat. size. Type. Nos. 5242 and 5247-50, Mus. Geol. Surv. India, Calcutta. Frond fairly large, spathulate, or oval -lanceolate. Midrib distinct. Secondary nerves forming very large, open, elongate GLOSSOrTFRIS. 87 meshes of approximately the same size throughout the lamina. Meshes oblong-polygonal, transversely elongate, much longer than broad. 67. conspicua recalls 67. retifera, but the leaf is generally larger, and the meshes much longer than broad, whereas in the latter species they are comparatively smaller, and more nearly of equal length and breadth. The meshes in G. conspicua are probably larger, and more open (i.e. broader in comparison with their length) than in any other known representative of this genus. G. conspicua has so far been recorded only from the Raniganj group of the Damuda division in India. Some specimens from the Orange River Colony are, however, here attributed to this species. Fronds of Glossopteris conspicua from the Orange River Colony. V. 2465. PI. Ill, Fig. 3. A fairly large, well-preserved, oval-lanceolate frond, in association with fronds of G. Browniana and G. retifera, of which only a portion is figured on PI. Ill, Fig. 3. The leaf measures 9 cm. in length, and has a maximum breadth of 4-4 cm., the basal part being absent. The nervation is very clear, the long but very broad nets recalling the Indian fronds described by Feistmantel. A smaller fragment also occurs on the same specimen. Mill River Drift, Harrismith. Pres. by D. Draper, Esq., 1890. V. 2902«. Fragments of fronds of this species, occurring in association with 67. retifera. Mill River Drift, Harrismith. Pres. by D. Draper, Esq., 1893. 8. Glossopteris formosa, Feistmantel. 1881. Glossopteris formosa, Feistmantel, Flora Gondw. Syst., vol. iii, pt. 3, p. 10G, pi. xxxixa, figs. 3-7. 1882. G. formosa, var. major, Feistmantel, ibid., vol. iv, pt. 1, p. 36, pi. xxi, fig. 12. 1886. G. formosa, Feistmantel, ibid., vol. iv, pt. 2, p. 26. Type. Nos. 5292-4, Mus. Geol. Surv. India, Calcutta. Leaves linear or linear-lanceolate. Apex obtuse. Midrib slender. 88 GLOSSOrTERIS. Secondary nervation acute, forming broad, oblong-polygonal meshes, of approximately the same size throughout the lamina. G. formosa appears to be a narrow-leaved type of frond corresponding to G. retifera or G. conspicua, in the same way as G. angustifolia probably corresponds to G. indica. It is possible that in each case both the narrow and broad-leaved fronds may have belonged to the same plant. G. formosa is known only from the Raniganj group of the Damuda division in India. Not represented in the British Museum collection. 9. Glossopteris tortuosa, Zeiller. 1902. Glossopteris tortuosa, Zeiller, Pal. Indica, n.s., vol. ii, p. 14, pi. iii, figs. 2, la. Type. No. 7267, Mus. Geol. Surv. India, Calcutta. Frond of medium size, contracting gradually towards the base ; median nerve strong, with longitudinal striations ; secondary nerves thick, spreading, irregularly sinuate, forming on each side of the midrib two series of large, polygonal meshes, the succeeding meshes being narrower, more elongate, polygonal, or trapezoidal. Zeiller has recently founded this species on a single specimen from India. It recalls the fronds of G. Browniana or G. indica, but the median nerve is stronger and is longitudinally striated, while the lateral nerves, which are flat and broad near the midrib, thinning gradually towards the margin, are almost at right angles to the midrib. The lateral nerves are sinuate, and anastomose at intervals, giving rise to a network, of which the two series of meshes next the midrib are broadly polygonal, the more distal being narrower and more elongate, sometimes truncate, sometimes narrowed to a point at their extremities. In their nervation, these fronds may be also compared with the Indian leaves referred to Feistmantel's species G. damudica, which is here included with G. ampla, Dana. Zeiller does not, however, regard them as identical. Glossopteris tortuosa is known only from the Raniganj group of the Damuda division in India. Not represented in the British Museum collection. GLOSSOrTEKIS. 89 10. Glossopteris diver gens, Feistmantel. (Text-fig. 23.) 1881. Glossopteris diver gens, Feistmantel, Flora Gondw. Syst., vol. iii, pt. 3, p. 104, pi. xxviiiA, figs. 3, 4. Type. Nos. 5244-5, Mus. Geol. Surv. India, Calcutta. This species is known only from a single imperfect specimen, of which the following appear to be the chief characters. Frond (?) ovate. Midrib strong. Secondary nerves very oblique, almost at right angles to the midrib in the median portion of the Fig. 23. — Glossopteris divergens, Feist. After Feistmantel. Nat. size. frond, becoming even more oblique towards the base, but less so towards the apex. The course of the secondary nerves is sinuous or tortuous, the meshes being of irregular shape, but fairly broad, and rather longer than broad. Feistmantel in his description of this species does not mention the flexuous character of the secondary nerves, which must be very conspicuous if the drawing which he gives (Text-fig. 23) is faithful to the original. Provided also this feature is not due 90 GLOSSOPTERIS. to an accident of preservation, it probably forms an important character of this frond. That the nerves in certain Glossopterids may be sinuous has been recently shown by Zeiller (see G. tortuosa). G. divergens is known only from the Kaniganj group of the Damuda division in India. Not represented in the British Museum collection. 11. Glossopteris decipiens, Feistmantel. (Text-fig. 24.) 1879. Glossopteris decipiens, Feistmantel, Flora Gondw. Syst., vol. iii, pt. 1, p. 17, pi. xviii, figs. 3-5, pi. xxiv, fig. 6. ? Sagenopteris (?) Stoliczkana, Feistmantel, ibid., p. 18, pi. xiii, fig. 4. 1880. Glossopteris decipiens, Feistmantel, ibid., vol. iii, pts. 2, 3, p. 107. ? Sagenopteris (?) Stoliczkana, Feistmantel, ibid., p. 114. 1886. Glossopteris decipiens, Feistmantel, ibid., vol. iv, pt. 2, p. 29. Type. Nos. 5025 and ?5012, Mus. Geol. Snrv. India, Calcutta. Frond narrowly spathulate, base truncate (?), basal angles rounded, or sub-auriculate. Midrib strong, hardly extending for more than two-thirds of the length of the leaf, breaking up above into radiating and anastomosing secondary nerves. Secondary nervation arising at an acute angle, forming narrow, oblong meshes. The general character of the secondary nervation appears to me to be closely similar to that of G. indica. The frond is, however, distinguished by the fact that the midrib does not extend for more than two-thirds of the length of the leaf, and by the truncated base, the lateral angles of which are slightly auriculate. As Feistmantel has pointed out, this species seems to present a transitional type between Glossopteris and Gangamopteris. It is also especially interesting as being one of the earliest members of the genus to appear in India, in rocks which are probably equivalent in age to the Upper Carboniferous deposits of Europe. It is remarkable that among the earliest Glossopterid fronds from Australia, there also appear to be some in which the midrib is impersistent, and dissolves into lateral veins at some distance from the apex. It is not improbable that when these early forms are better known, this character may be found to be of some GLOSSOrTKIUS. 91 Fig. 24. — Glossopteris decipiens, Feist. After Feistniantel. J nat. size. significance. Feistniantel 1 has figured a very imperfect fragment from the "Lower Coal Measures" of Greta, New South Wales, 1 Feistniantel (90), p. 124, pi. xii, figs. 3, 3«, and (78), p. 1.55, pi. viii, figs. 2, la. Also identified by Johnston (86), p. 378 ; Tenison- Woods (83), p. 125 ; and Shirley (02), p. 13. 92 GLOSSOPTEKIS. under the name G. elegans, which, in the characters of the midrib and nervation generally, offers a close comparison with the Indian species under consideration. It is, however, impossible, on such imperfect evidence, to arrive at any conclusion as to the identity of these two fronds, and for the present G. elegans can hardly rank as a well-defined species. With these species may be also compared Feistruantel's G. gatigamopteroides (see p. 53) from New South Wales. The Indian specimens figured by Feistmantel 1 as Sagenopteris (?) Stoliczhma are undoubtedly Glossopteris fronds, and very possibly leaves of G. decipiens. Glossopteris decipiens is known only from the Karharbari Series of the Talchir division of India. Not represented in the British Museum collection. 12. Glossopteris longicaulis, Feistmantel. 1881. Glossopteris longicaulis, Feistmantel, Flora Gondw. Svst., vol. iii, pt. 1, Suppl., p. 53, pi. xxxi, figs. 1, 3. Type. Nos. 5084 and 5086, Mus. Geol. Surv. India, Calcutta. Frond oblong-oval, with a long petiole. Midrib distinct in the lower part of the frond, but evanescent above. Secondary nervation arising at a sub-acute angle, forming somewhat broad, oblong meshes. This frond is only known from the Karharbari Series in India. Its chief distinguishing characters appear to be the long petiole, and impersistent midrib. Not represented in the British Museum collection. 13. Glossopteris orbicularis, Feistmantel. (Text-fig. 25.) 1880. Glossopteris orbicularis, Feistmantel, Flora Gondw. Syst., vol. iii, pts. 2, 3, p. 107, pi. xliA, figs, l, 2. Type. No. 5307, Mus. Geol. Surv. India, Calcutta. Frond orbicular, sub-emarginate, petiolate. Median nerve strong 1 Feistmantel (79 1 ), p. 18, pi. xiii, fig. 4. VERTEBRA RIA. 93 in the lower portion of the leaf, but dissolving into fine veins towards the apex. Lateral nervation acute, arched. Meshes wide, oblong-polygonal. Feistrnantel has suggested that this almost circular frond may- be only an immature stage of some other species. It is known from two localities in the Raniganj group of the Damuda division of India. The same author l has, however, described, under the name G. spathulato-cordata, some very similar fronds from the Newcastle Series of New South Wales, and from Tasmania. So far Fig. 25. — Glossopieris orbicularis, Feist. After Feistrnantel. Slightly reduced. as one can judge from a comparison of his figures, the latter specimens, although smaller, have a similar nervation, and may eventually prove to be identical. Not represented in the British Museum collection. THE 11HIZOME OF GLOSSQPTERIS. Genus VERTEBRARIA, Royle, 1833. [Illust. Bot. Himal. Mounts., p. xxix*.] Elongate, flattened, cylindrical casts, simple or branched, often bearing root-like organs. In specimens preserved more or less perpendicular to the bedding- 1 Feistrnantel (90), p. 128, pi. xx, figs. 5, 7, 8. 94 . VERTEBRARIA. planes of the rock, the axis, as seen in transverse section, is composed of a number of compact, wedge-like sectors, radiating from a common centre. Between the sectors, and sometimes external to the broad faces of the wedges, a thin film of carbonaceous material is found. In specimens preserved approximately parallel to the bedding- planes, the cast, as seen in surface view, is formed of two or three longitudinally disposed series of oblong or almost square areas. When two series of areas only are present they are separated by a longitudinal groove or ridge. Where three series are found two longitudinal grooves may occur, but they are often less well-marked than in the former case. In many specimens the areas of one series slightly overlap those of another. The areas are often striated longitudinally, the strias being fairly distant. The areas of each longitudinal series are separated from one another by transverse grooves, the grooves of any two parallel series being very rarely, if ever, opposite. In most specimens each area has a well-marked transverse ridge, which varies greatly in position. The length of the area, i.e. the distance between two transverse grooves, also varies considerably in the same series, and in two parallel series. The same cast may show three series of areas in one part, and only two in another. The root-like organs arise at the transverse folds, and are highly branched structures. Vertebraria occurs in the form of casts presenting features quite unlike any of the associated fossils, and much speculation has arisen as to the affinities of this plant. Royle, who instituted the genus, speaks of it as the " Ranigunj Reed." Ennbury 1 was inclined to regard it as a root or rhizome of an Equisetalian plant, a view which Peistmautel at one time also shared to some extent. McCoy 2 suggested that these fossils originated from plants bearing slender, jointed stems, surrounded by compact whorls of six to twelve wedge-shaped leaves. All doubts were, however, set at rest in 1896, when Zciller 3 discovered that Vertebraria was neither more nor less than the rhizome of 1 Bimbury (Gl), p. 339. 2 McCoy (47), p. 146. a Zeiller (9G 3 ) and (9G 1 ), pi. xv, figs. 8, 9. YKRTEBKARIA. 95 Glossopteris, a discovery which was confirmed almost simultaneously by Oldham. 1 In both cases fronds of Glossopteris were found in continuity with Vertebraria. 2 Previously to Zeiller's discovery, several observers had described fronds of Glossopteris attached to an axis, usually extremely fragmentary, or at any rate not possessing the well-known features of Vertebraria. McCoy, 3 in 1847, stated that he believed he had "ascertained the rhizoma of this species" (G. Browniana); but he did not give any particulars of his discovery. In 1849 Dana 4 figured a number of petiolate fronds of Glossopteris attached to a rhizome, and, in his great memoir on the Flora of the Lower (londwana Series, Feistmantel ° figured several others. But the figures published by Etheridge 6 in 1894 are more important. The specimen described was obtained from Mudgee, New South Wales, and showed several leaves of Glossopteris Browniana in continuity with a stem structure, described as a caudex by that author, which presented none of the features characteristic of Vertebraria. This specimen will be further discussed here. The explanations offered of the peculiar features exhibited by this fossil have been varied. That suggested by Solms-Laubach 7 was that these rhizomes were cylindrical organs with a solid axis surrounded by lacunae, which were bridged here and there by transverse diaphragms of tissue, connecting the cortical tissues with the central axis. That offered by Zeiller 8 in 1896 is much more elaborate. He concludes that the features of Vertebraria are derived as casts of the external surface of a winged rhizome. He compares it with the rhizome of such a fern as Struthiapteris germanica, Willd., in which a variable number of prominent, projecting, longitudinal wings occur, which anastomose with one another at different levels. The areas which form such a characteristic feature of the fossil as seen in surface view, are the interpolations of the matrix of the rock between the wings 1 Oldham (97), pi. iii. 2 A specimen showing continuity between these two organs has since been found in the Clarke Collection in the Sedgwick Museum, Cambridge. 3 McCoy (47), p. 151. ~ i Dana (49), p. 716, pi. xii, fig. 13c. 5 Feistmantel (80), pi. xl a, fig. 1 ; pi. xli a, fig. 3. 6 Etheridge & David (94). 7 Solms-Laubach (91), p. 366. 8 Zeiller (9G 1 ), p. 3J1 ; (96'). 96 VEBTEBBABIA. of the rhizome, which in the fossil state are sometimes represented by a small amount of carbonaceous material, while the transverse grooves may be explained as the points of anastomosis of two of the wings at short intervals. Zeiller also figures fronds attached at the transverse grooves, and in other instances he finds a print of the attachment of a leaf in a similar position. In 1897 Oldham 1 published an account of a specimen of Verte- Iraria from India, to which a number of fronds, probably of Glossopteris indica, were attached, thus confirming Zeiller's con- clusion as to the nature of this fossil. At the same time Oldham pointed out that his specimens of Vertebraria 2 did not agree in all particulars with those described by Zeiller, and he doubted whether the explanation of the morphological features of the fossil put forward by that author was entirely correct. In particular, he pointed out that in some of the Indian specimens, as seen in transverse section, a shell of coaly matter, which was apparently wanting in the South African specimens, was present outside the wedge-like sectors. He concludes that Vertebraria must have been a stem with a central axis connected with some outer tissues by radiating, longitudinally disposed plates or septa, which were usually eight in number, in addition to numerous transversely placed septa, which divided up the lacunae between the longitudinal septa into chambers. In 1902 Zeiller 3 replied to Oldham's criticism, and maintained his explanation, pointing out that the small traces of a carbonaceous layer outside the wedges, which are found in some specimens as seen in transverse section, correspond with the points of anastomosis of two wings of the rhizome. It must be admitted that Zeiller's explanation of Vertebraria is the best which has so far been put forward. Yet, without attempting to criticise it in detail, I am not entirely convinced that the morphological features of that fossil are directly due to the structure of the external surface of a rhizome. It must be remembered that Etheridge, 4 in 1894, figured a specimen showing some fronds of Glossopteris Browniana attached to an axis, which 1 Oldham (97), pi. lii. 2 Oldham (97), pis. iv, v. 3 Zeiller (02 1 ), p. 17. * Etheridge & David (94), p. 228, pi. xviii, fig. 1 ; pi. xix, figs. I, 2. VERTKBRAEIA. 97 appears to me to be of the nature of a rhizome, but which, obviously, is not a Vertebraria. The Australian specimen has exactly the appearance that one would expect the external surface of a rhizome to present. It is a cylindrical axis on which may be seen numerous prints or leaf-scars. Similar organs have also been described elsewhere in association with the Glossopteris flora. Feistmantel : has figured one from India ; and one from Port Stephens, New South Wales (V. 7206, Odinheimer Coll.), is figured here on Plate VI, Pig. 4. The latter specimen occurs on a piece of yellowish-white shale, on the back of which are found several fragments of fronds of G. Browniana, and a small portion of a Vertebraria. The two cylindrical bodies figured appear to be axes, on which may be seen numerous depressions having all the characteristics of leaf-scars, the vascular prints being clearly visible here and there. They were no doubt stem structures bearing crowded, almost overlapping leaves, and there is every probability that an impression of the true external surface is preserved. There is, however, no absolute proof that this specimen belonged to Glossop- teris, although it somewhat resembles that figured by Etheridge. In view of this evidence it would appear to me more probable that the characteristic features of Vertebraria may be due to the internal, and not to the external structure of a rhizome ; a structure with which we are at present entirely unacquainted, unfortunately, owing to the absence of petrified material. In this case we should be dealing with an internal and not an external cast. It may be that impressions of the external features of these rhizomes were rare as compared with the internal, just as in the case of the stems of Catamites. AVhether this provisional suggestion is likely to prove correct remains to be seen. In the meanwhile there is room for still further research into the origin of the features presented by this puzzling fossil, Vertebraria. Vertebraria indica, Eoyle. (PL IV, Pigs. 2-4.) 1833. Vertebraria indica, Eoyle, Illust. Bot. Hirual. Mounts., p. xxix*, pi. ii, figs. 1-3. V. radiata, Royle, ibid., p. xxix*, pi. ii, figs. 5-7. 1 Feistmantel (79'), pi. xiii, fig. 6. yb VKKTEBRAB.IA. 1847. Vertebraria aus trails, McCoy, Aim. & Mag. Nat. Hist., vol. xx, p. 147, pi. ix, fig - . 1. 1849. Cluster ia australis, Dana, iii Wilkes' U.S. Explor. Exped., vol. x, p. 719, pi. xiv, figs. 3-5. 1850. Vertebraria indica, M'Clelland, Rep. Geol. Surv. India, pi. xiv, figS. 1, 1(7. Sphenophyllum australe, Unger, Gen. et Spec. Plant, foss., p. 72. S. indicum, Unger, ibid., p. 71. S. radiation, Unger, ibid., p. 71. 1861. Vertebraria (?), Bunbury, Quart. Journ. Geol. Soc, vol. xvii, p. 338, pi. xi, fig. 3. 1876. Vertebraria indica, Feistrnantel, Journ. Asiat. Soc. Bengal, vol. xiv, pt. 2, p. 347, pi. xv, fig. 3 ; pi. xvi. fig. 4. 1S78. V. australis, Feistrnantel, Paheantogr., Suppl. iii, p. 84, pi. vi, figs. 1, 2; pi. vii, figs. 1, 2. 1879. V. indica, Feistrnantel, Flora Gondw. Syst., vol. iii, pt. 1, p. 8, pi. i, figs. 8, 9. 1880. V. indica, Feistrnantel, Flora Goudw. Svst., vol. iii, pts. 2, 3, p. 72, pi. xiiA, figs. 10, 11; pi. xiiiA, figs. 1-6; pi. xiv a, figs. 1-4; pi. xiv a bis, fig. 3. 18S2. V. indica, Feistrnantel, Flora Gondw. Syst., vol. iv, pt. 1, p. 22, pi. xi, figs. 1-4. 1S83. V, australis, lYnison-Woods, Proc. Linn. Soc. New South Wales, vol. viii, p. 75. 1886. V. indica, Feistrnantel, Flora Gondw. Syst., vol. iv, pt. 2, p. 22, pi. ivA, figs. 4, 5, 7-11 ; pi. v a, fig. 1 ; pi. xiiiA, fig. 8. V. australis, Jobnston, Papers and Proc. It. Soc. Tasmania for 1885, p. 364. 188S. V. indica, Scbenk, Foss. Pflanzenr., p. 188. 1890. V. australis, Feistrnantel, Mem. Geol. Surv. New Soutb Wales, Pal. Xo. 3, p. 87 ; pi. xiv, fig. 6 ; pi. xv, figs. 1-3. 1891. V. sp., Solms-Laubach, Foss. Botany, Eng. ed., pp. 365-6. 1893. J", indica, Oldham. Man. Geol. India, pi. opp. p. 162. 1896. J', indica, Zeiller, Bull. Soc. Geol. France, ser. in, vol. xxiv, p. 351, pi. xv, figs. 1-9. 1897. V. indica, Oldham, Rec. Geol. Surv. India, vol. xxx, pt. 1, pp. 45, 49, pi. iv, figs. 1-5 ; pi. v, figs. 1-4. 1900. 7'. indica, Zeiller, Elem. Paleobot., p. 114, text-fig. 87. ? V. sp., Potonie, in Deutsch. Ost-Afrika, vol. vii, p. 499, figs. 23, 24. 1901. V. indica, Arber, Geol. Mag., dec. iv, vol. viii, p. 547. 1902. V. indica, Zeiller, Pal. Iucbca, n.s., vol. ii, p. 17, pi. V, figs. 1-7. 1903. V. australis, Arber, Quart. Journ. Geol. Soc, vol. 1 viii, p. 8. Type. No. V. 4189, Geol. Dept. British Museum (Nat. Hist.). There is, I believe, only one species of Vertebraria known, V. indica, and consequently the characters of the species are those of the genus. VERTEBRARIA. 99 I have here united V. australis with V. indica, since there do not appear to be any good characters which clearly separate them. Even Feistmantel admitted that these two species were closely similar. Tenison - Woods 1 has described two species from the Mesozoic rocks of Queensland, but it is more than doubtful if they have any claims to be included in the genus, as Zeiller has already pointed out. The fossil described by Schmalhausen 2 as V. (?) petschorensis is almost certainly not a true member of the genus. Royle, who first described the Indian species, assigned two separate names, V. indica and V. radiata, to the same fossil as seen in surface view and transverse section respectively. Curiously enough the same thing happened with regard to the Australian fossils, McCoy's V. australis being a specimen showing the transverse section, and Dana's Clasteria australis representing the same plant as seen in surface view. Specimens of Vertebraria indica from India. V. 4189. PI. IV, Fig. 2 ; also figured by Royle (33), pi. ii, fig. 1. Type. Of the six specimens of Vertebraria figured by Royle, only one has been identified, and this is the more important of those described under the name V. indica. It measures 15 cm. in length, and 1*8 cm. across. The median groove is well marked, and the areas between the transverse furrows on either side are almost square (9 mm. long and broad), or slightly oblong (6x16 mm.). Transverse ridges also occur. The upper portion of the cast is probably imperfect. Bardwan Coalfield. V. 7191. A specimen rather more than 10 cm. long, and about 2 cm. broad, showing the longitudinal and transverse grooves and the transverse ridges. Near Nagpur. Sanlcey Cull. Specimens of Vertebraria indica from New South Wales. V. 7208. PL IV, Fig. 3. A specimen measuring nearly 9 cm. in length, and 2*5 cm. in 1 Tenison- Woods (83), pp. 80-1, pi. i, figs. 1-4. 3 Schmalhausen (79), p. 53, pi. vii, figs. 14-18. 100 VERTEBRARIA. breadth, with a well-marked, somewhat flattened median ridge, formed by the compression of a median series of areas. The transverse ridges and grooves of the lateral areas alternate on either side of the ridge, and occur at very unequal distances from each other. On the right-hand side of the specimen as seen in the figure, the oblong areas between the four transverse grooves overlie the areas of the median ridge. Four transverse ridges are also seen which appear to correspond with the transverse ridges of the median series, whereas the grooves do not correspond. On the left-hand side three ridges and grooves are seen. The longitudinal ridge is flattened over some of the areas on this side. Port Stephens. Odinheimer Coll. V. 7212. PI. IV, Fig. 4. A specimen showing two fragments of Vertebraria, of which one is figured here. There is no longitudinal groove or ridge, but a median series of areas, which is overlapped by both the lateral rows of areas. The transverse furrows of the median series do not correspond with those of either of the lateral series. Three ridges are seen in the median row. As figured here, the right-hand lateral series shows five areas, and the left-hand three, on which well-marked, but distant strife occur. Port Stephens. Odinheimer Coll. V. 4286. Two series of areas, separated by a median groove, are seen in this specimen for a considerable portion of its length, but at one end there is simply a single row of larger areas, which are striated longitudinally. Port Stephens. Odinheimer Coll. V. 7210. A small piece of a rhizome, showing a longitudinal groove and two series of areas with transverse ridges and grooves. The areas are faintly striated longitudinally. Fragments of scale- fronds are associated with this specimen. Port Stephens. Odinheimer Coll. V. 7209. A specimen of Vertebraria, some 9 cm. in length. The longitudinal groove is indistinct, but there are two series of areas, with well-marked transverse grooves and ridges. The areas are also striated longitudinally, the striae being fairly distant from each other, about I mm. apart. Several fronds of Glossopteris Browniana are also associated. Port Stephens. Odinheimer Coll. VKRTEBRARIA. 101 V. 4203. A fragment of a rhizome with plainly-marked median and lateral grooves, associated with fronds of Glossopteris Browniana. Port Stephens. Odinheimer Coll. V. 7290. A portion of a Vertebraria with rootlets, also associated with fronds of Glossopteris Browniana, showing the nervation clearly. Port Stephens. Odinheimer Coll. Other specimens :— V. 4285, V. 7203, ? V. 7223. Port Stephens, New South Wales ; Odinheimer Coll. Roots and Rootlets of Uncertain Affinity. Specimens of roots or rootlets are occasionally found associated with members of the Glossopteris flora. Some of these no doubt may be attributed to Vertebraria, while others belong to plants whose nature it is impossible to decide. Some of the specimens to which Dana gave the names Austrella rigida, 1 Confervites(?) tenella, 2 Cystoseirites, sp., 3 are probably of this nature. The term Pustularia calderiana, Royle, 4 is a nomen nudum. Specimens of Roots, etc. V. 7192. A piece of sandstone with a number of highly branched roots or rootlets, with some trace of a main axis. ? India. V. 7134. Fragments of branched root-like structures. Near Kamthi, India. Hunter Coll. V. 7198. A root, bearing rootlets. Nagpur, India. Sankey Coll. V. 239. A root or rhizome. Beaufort Series, Zwartkei River, Cape Colony. V. 2112. Fragments of rootlets resembling the specimens from India, above described. Bowenfels, New South "Wales. 1 Dana (49), p. 720, pi. xiv, figs. 7, 8. 2 Dana (49), p. 720, pi. xiv, fig. 9. 3 Dana (49), p. 720, pi. xiv, fig. 6a. 4 Royle (33), p. xxix* ; see also Arber (01), p. 549. 102 GANGAMOPTERIS. V. 2909. Several root-like structures with, median grooves, associated with fragments of rootlets. Farm Glencalder (at 6,700 feet s.m.), Newcastle Division, Natal. Pres. by D. Draper, Esq., 1893. V. 2906. Fragments of a rhizome-like structure, and of rootlets. Farm Glencalder (at 6,700 feet s.m.), Newcastle Division, Natal. Pres. by B. Draper, Esq., 1893. V. 2905, V. 2905«, V. 2905^. Numerous fragments of rootlets, in association with Glossopteris. Farm Glencalder (at 6,700 feet s.m.), Newcastle Division, Natal. Pres. by D. Draper, Esq., 1893. V. 8318. A piece of sandy shale, largely composed of root-like structures, hearing lateral rootlets. The details of the external morphology are not very clear. Candiota, Rio Grande do Sul, Brazil. Pres. by J. Mawson, Esq., 1894. V. 8316, V. 8316ff, V. 8317. Smaller fragments, similar to the last sjiecimen, consisting chiefly of lateral rootlets. Candiota, Bio Grande do Sul, Brazil. Pres. by J. Mawson, Esq., 1894. Genus GANGAMOPTEEIS, McCoy, 1861. [Trans. Eoy. Soc. Yictoria, vol. v (1860), p. 107, note.] Frond simple, entire, greatly varied in shape, elliptical, broadly and elongately ohovate, broadly lanceolate, sub-linear, etc., contracted towards the base. Midrib absent. Median nerves more or less parallel, anastomosing ; lateral nerves arising by repeated dichotomy from the median nerves or from the base, arched, bifurcating, and anastomosing to form a network. The chief feature in which Ganyamopteris differs from Glossopteris is the absence of a definite midrib, the median portion of the leaf being usually traversed by a group of almost parallel, anastomosing nerves. The anastomosing lateral nervation is very similar in both genera, and almost as great variations are found in Ganyamopteris as in Glossopteris, in the habit and details of the nervation, even GANGAMOPTEEIS 103 in fronds which were probably borne by one and the same plant. In some cases it is questionable whether it is possible to distinguish clearly between the two genera. 1 Nothing is known as to the fructification of any member of the genus, nor has it been ascertained whether the fronds are dimorphic like those of Glossopteris. Feistmantel 2 has figured some very small fronds, recalling the scale-fronds of Glossopteris, which he referred to Gangamopteris, but whether they should be regarded as belonging to the former genus rather than the latter there is no evidence at present to show. The specimen originally described by McCoy in 1847 as Cyclopteris angustifolia, on which the genus is founded, is pre- served in the Sedgwick Museum, Cambridge. Distribution. — Gangamopteris is perhaps as widely distributed in the Permo-Carboniferous rocks of Gondwanaland as Glossopteris, although it is not of such common occurrence nor so generally abundant. In South America it occurs in both Brazil and Argentina, whereas Glossopteris has been only comparatively recently discovered in the latter country. In Victoria, Australia, Glossopteris is unknown, while several species of Gangamopteris are found. In India it appears to be more abundant than Glossopteris in the lowest member of the Gondwana Series, the Talchir- Karharbari Beds, though in the "Lower Coal Measures" of New South Wales, of approximately the same age, it is unknown, whereas several Glossopterids occur. Otherwise it occurs in almost all the divisions of the Permo-Carboniferous rocks in various parts of the world from which the Glossopteris flora is known. It has not, however, been recorded from beds of Mesozoic age. Permo-Carboniferous (Glossopteris flora): — India, in the Talchir and Damuda divisions ; Kashmir ; New South Wales, in the Newcastle Series; Victoria; Tasmania; S. Africa, Transvaal; South America, Brazil and Argentina. Permian (Northern Type) : — Russia. 1 See Etheridge & David (94), pp. 240-1. 2 Feistmantel' (79 1 ), pi. ix, fig. 4 ; pi. x, fig. 3 ; (81), pi. xxxixA, fig. 9. 104 GANGAMOrTERIS. 1. Gang-aniopteris cyclopteroides, Feistmantel. (PI. Ill, Figs. 4, 5 ; Test-fig. 26.) 1869. Noeggerathia obovata, Carruthers, Geol. Mag., vol. vi, p. 155, pi. vi, fig. 1. 1875. Gangamopteris spatulata, McCoy, Prodr. Palaeont. Victoria, dec. n, p. 12, pi. xiii, figs. 1, la. G. obliqua, McCoy, ibid., p. 13, pi. xii, figs. 2-4. 1876. G. cyclopteroides, Feistmantel, Pec. Geol. Surv. India, vol. ix, pt. 3, pp. 73, 78. 1878. G. spathulata, Feistmantel, Palasontogr., Suppl. iii, p. 102. G. obliqua, Feistmantel, ibid., p. 102. G. Glarkeana, Feistmantel, ibid., p. 93, pi. xv, fig. 9. 1879. G. cyclopteroides, Feistmantel, Flora Gondw. Syst., vol. iii, pt. 1 p. 12, pi. vii ; pi. viii; pi. ix, figs. 1-3. 6; pi. xi, figs. 2-4 pi. xii, figs. 2, 3 ; pi. xiii, figs. 1,5; pi. xxvi, figs. 1,3; pi. xxvii figs. 2, 2a, 3. G. cyclopteroides, var. subauriculata, Feistmantel, ibid., p. 13, pi. x figs. 1, la, lb; pi. xiii, fig. 2 ; pi. xv, figs. 1-3 ; pi. xvi, fig. 3. G. cyclopteroides, var. areolata, Feistmantel, ibid., p. 14, pi. x, fig. 2 pi. xvi, figs. 4, 4a. G. cyclopteroides, var. attenuata, Feistmantel, ibid., p. 14, pi. xi fig. 1 ; pi. xii, fig. 1 ; pi. xiii, fig. 3 ; pi. xiv, figs. 1,2; pi. xvi fig. 5; pi. xxvii, bus. 1, la. 1881. G. obliqua, Feistmantel, Pec. Geol. Surv. India, vol. xiv, pt. 3, p. 242 pi. ii, fi Fig. 28. — Neuropteridium validum, Feist. After Feistmantel. Nat. size. NEUROPTEETDIUM. 1 1 9 It may be added that a species described by Johnston l from Hobart, Tasmania, in 1887, as Neuropteris tasmaniensis, Johnst., is stated by that author to be " near to N. valida, Feist." .No figures of this species are, however, published, and the age of the beds is not very clear. Although the fronds described by Carruthers as Odontopteris Plantiana are now known to belong to this genus and species, and are the earliest described examples, I have maintained Feistmantel's specific name here, since any change would tend to create great confusion. Zeiller 2 was the first to point out the identity of these fossils. Fronds of Neuropteridium validum from India. V. 227. PI. VI, Fig. 1. A basal portion of a frond showing a number of oval or orbicular pinnules, rounded at the apex, on either side of a rachis. The pinnules at the base of the frond are very small, entire, almost semicircular, and are attached by the whole base. The larger, oval pinnules are somewhat lobed, and the upper portion of the base is free from the rachis. The nervation is fairly clear, the nerves radiating from the base with frequent dichotomy. Other and smaller fragments of fronds occur besides that figured. ? India. Pres. by C. W. Wilmot, Esq., 1883. V. 227«. A small fragment showing the rachis, and three nearly perfect pinnules, in which nervation is very clear. There is no record of the locality with either of these specimens, but there is little doubt that they were derived from India. ? India. Pres. by C. W. Wilmot, Esq., 1883. Fronds of Neuropteridium validum from Brazil. V. 228tf, Figured by Carruthers (69), pi. vi, fig. 2. A portion of a frond showing lobed pinnules on one side of the rachis. The nervation is fairly clear, the nerves being fine, close, and dichotomising. The impression of the impersistent median nerve can just be made out at the base of the pinnules. Eio Grande do Sul, Brazil. Pres. by N. Plant, Esq., 1869. 1 Johnston (87 2 ), p. 171. ' Zeiller (95 1 ), p. 963 ; (95 2 ), p. 616. 120 T^NIOPTERIS. V. 228. Figured by Carruthers (69), pi. vi, fig. 3. Possibly a large lobed pinnule, but the specimen is not well preserved. Bio Grande do Sul, Brazil. Pres. by X. Plant, Esq., 1869. V. 228^. A very imperfect example of a pinnule. Eio Grande do Sul, Brazil. Pres. by N. Plant, Esq., 1869. Genus TiENIOPTEBIS, Brongniart, 1828. [Prodr. Hist. Veget. foss., p. 61.] Fronds or leaves simple, or more rarely pinnate, ribbon-like, lanceolate, linear-lanceolate, oblong or elliptical, usually entire. Apex acute, acuminate, or obtuse. Median nerve well marked, extending almost to the apex. Lateral nerves arising from the midrib at a very wide angle, numerous, simple or dicbotomising once or several times. A satisfactory definition of this genus is exceedingly difficult if only on account of tbe transitions which some of these fronds seem to present to those of other genera such as Nilssonia. Brongniart's original definition was as follows : — " Fronde simple, entiere, etroite, a bords paralleles, traversee par une nervure moyenne, forte, epaisse, qui s'etend jusqu'a l'extremite ; nervures secondaires presque simples ou bifurquees a la base, presque perpendiculaires sur la nervure moyenne." In 1869, Schimper 1 proposed a more extended classification of fronds of this type, restricting the genus Tceniopteris to certain species of Palaeozoic age, and at the same time instituting a number of other generic names, of which Macrotceniopteris, Oleandridiurn, Angiopteridium, and Panaopsis were the chief. In more recent times this classification has been criticised on the grounds that there is not sufficient evidence to warrant such separation, and that for the present it is safer and more convenient to group these fronds under the original genus Tceniopteris, using that name in a wide sense. This view has been adopted by Xathorst, 2 White, 3 Seward, 4 Zeiller, 5 and others, and is main- tained here. 1 Schiniper (69), vol. i, p. 600, etc. - Nathorst (78). 3 White (93). * Seward (94), p. 124. 5 Zeiller (82 2 ) and (02 s ), p. 59. TJENIOPTEKIS. 121 Tceniopteris is essentially a Mesozoic type of frond in that the genus reaches its maximum development in Triassic and Jurassic times, and extends to the Lower Cretaceous period. It makes its first appearance, however, in the higher horizons of the Upper Carboniferous, and in the Permian rocks of Europe and elsewhere in the Northern Hemisphere, as well as in the Permo-Carboniferous beds of India, ? Australasia, and South Africa. So far we have no knowledge of the fructification of any of the Palaeozoic fronds. Zeiller ' has, however, recently described the fructification of a certain Triassic species, which presents characters closely identical with those of recent members of the Marattiacece. It is possible, however, that other fronds included in this genus, in the wide sense adopted here, may eventually prove to be more closely related to the Cycadean stock than to the true Ferns. 1. Tseniopteris danaeoides (Royle). (PL V, Fig. 1.) 1833. Glossopteris danceoides, Royle, Must. Bot. Hiraal. Mounts., p. xxix*, pi. ii, fig. 9. 1836. Aspidites danceoides, Goppert, Die Foss. Farn., p. 352. 1850. Tceniopteris danceoides, M'Clelland, Rep. Geol. Surv. India, p. 56, pi. xv, figs. 1, la, lb. Pecopteris danceoides, linger, Gen. et Spec. Plant, foss., p. 170. 1876. Tceniopteris danceoides, Feistmantel, Rec. Geol. Surv. India, vol. ix, pt. 3, p. 74. Macrotceniopteris danaoides, Feistmantel, ibid., vol. ix, pt. 4, p. 137. M. danceoides, Feistmantel, Jouru. Asiat. Soc. Bengal, vol. xlv, pt. 2, p. 365, pi. xix, figs. 1, la, 2 ; pi. xxi, fig. 1. 1880. M. danceoides, Feistmantel, Flora Goudw. Syst., vol. iii, pts. 2, 3, p. 88, pi. xx a; pi. xxiA, figs. 1, 2. 1886. M. danceoides, Feistmantel, ibid., vol. iv, pt. 2, p. 24, pi. ivA, figs. 2, 3. 1893. M. danceoides, Oldham, Man. Geol. India, 2nd pi. opp. p. 162. 1901. M. danceoides, Arber, Geol. Mag., dec. iv, vol. viii, p. 548. Type. V. 4191, Geol. Dept. British Museum (Nat. Hist.). Frond large, elliptical or oblong, petiolate, membranous or sub-coriaceous. Midrib thick. Lateral nerves arising at an acute angle, and almost immediately bending towards the margin in 1 Zeiller (02 3 j, p. 63. 122 T.ENIOrTERIS. a direction nearly at right angles to the midrib. Nerves numerous, somewhat distant, strong, dichotomising once, or more rarely twice. The broad form of the frond, and the somewhat distant and coarse lateral nerves, form the chief characteristics of this species. The veins are at least 1 mm. apart. This frond was originally referred to the genus Ghssnpteris by Royle, but M'Clelland, in 1850, transferred it to Tceniopteris. More recently, several authors, including Feistmantel, have adopted Sehimper's generic name Macrotceniopteris for this species, which, as has been already pointed out, is a term of doubtful value. The chief characters distinguishing Macrotceniopteris from Tceniopteris appear to be the larger size of the frond — a point of but little importance — and the more distant secondary nerves. 1 There is also an obscure fragment, the only specimen of its kind known from the Damuda Series, which was originally described by M'Clelland 2 as Zamia lurdwanensis. This specimen, preserved in the Calciitta Museum (No. 5340), was refigured by Feistmantel 3 as Pteroplujllum burdwanense. It would appear doubtful if this is really a Cycadean frond. So far as one can judge from Feistruantel's drawing, it would seem more probable that it may be after all only a torn fragment of a frond of Tceniopteris, a conclusion supported by the irregular margin of the supposed pinnules, as shown in the drawing referred to above. Tceniopteris danceoides occurs in the Raniganj and Earakar Series of the Damuda division in India. V. 4191. PI. V, Fig. 1 ; also figured by Eoyle (33), pi. ii, fig. 9. Type. A fine specimen of a nearly complete oval-lanceolate frond, the apex being wanting. It measures 13 cm. in length, and 6 cm. across at the widest part. The midrib is stout, and the veins, which are very clear, arise almost at right angles to the midrib, remaining simple or branching dichotomously. There is no regular alternation between the simple or branched veins. Bardwan Coalfield, India. 1 Arber (01), p. 548. 3 M'Clelland (50), p. 53, pi. xiv, fig. 4. 3 Feistmantel (80), p. 116, pi. xlviiA, fig. 1. T^NIOrTEETS. 123 2. Tseniopteris Feddeni (Feistmantel). (Text-fig. 29.) 1876. Macrotaniopteris Feddeni, Feistmantel, Rec. Geol. Suit. India, vol. ix pt, 4, p. 137. M. Feddeni, Feistmantel, Journ. Asiat. Soc. Bengal, vol. xlv, pt. 2 p. 367. 1880. M. Feddeni, Feistmantel, Flora Gondw. Syst., vol. iii, pts. 2, 3, p. S9 pi. xxia, fig. 3 ; pi. xxiiA, tigs. 1-4. 1881. M. Feddeni, Feistmantel, Rec. Geol. Surv. India, vol. xiv, pt. 3 p. 255, pi. ii, fig. 1. 1882. M. Feddeni, Feistmantel, Flora Gondw. Syst., vol. iv, pt. 1, p. 31 pi. xxi, fig. 5. 1886. M. Feddeni, Feistmantel, ibid., pt. 2, p. 24, pi. iA, fig. 1. Fig. 29. — Tcenioptcris Feddeni (Feist.). After Feistmantel. Nat. size. Type. Nos. 5200, 5203-6, Mus. Geol. Surv. India, Calcutta. Frond simple, large, elongately elliptical. Apex obtuse or emarginate. Midrib fairly stout, striated. Lateral nerves very numerous, simple or dichotomising close to the midrib, closely approximated, pursuing an almost horizontal direction in the basal portion of the frond, but at a more acute angle to the midrib towards the apex. The nerves at the margin appear to be slightly curved upwards. T. Feddeni is chiefly distinguished from T. danceoides by the closer lateral nervation, and by the thinner midrib. Some of the 124 T.ENIOPTERIS. fronds figured by Feistinantel are more than 25 cm. long, and 12 5 cm. broad. T. Feddeni occurs in the Eaniganj and Barakar Series of the Damuda division in India. Not represented in the British Museum collection. 3. Tseniopteris spathulata, M'Clelland. 1850. ? Tceuiopleris spathulata, M'Clelland, Rep. Geol. Surv. India, p. 53, pi. xvi, fig. 1. 1861. T. Daintreei, McCoy, Trans. R. Soc. Victoria, vol. v, pp. 96-107, 215-217. T. Daintreei, Clarke, ibid., pp. 89-95, 209-214. 1863. Stangerites spathulata, Oldham & Morris, Flora Gondw. Syst., vol. i, pt. 2, p. 34, pi. vi, figs. 1-7. 1867. Taniopteris Daintreei, McCoy, Ann. & Mag. Nat. Hist., ser. m, vol. xx, p. 196. 1869. Angiopteridium spnthulatum, Schimper, Traite, vol. i, p. 605. 1872. Tmiiopteris Daintreei, McCoy, Rep. Westernport Coalfields, p. 6. 1874. T. Daintreei, McCoy, in Smyth's 1st Prog. Rep. Geol. Surv. Victoria, p. 35. 1875. T. Daintreei, McCoy, Prodr. Palaeont. Victoria, dec. n, p. 15, pi. xiv, figs. 1, 2. 1876. T. Daintreei, Feistmantel, Rec. Geol. Surv. India, vol. ix, pt. 4, pp. 122-4. T. {Angi'ipteridium) McClellandi , Feistmantel, ibid., vol. ix, pt. 2, p. 36. 1877. Angiopteridium spathulatum, Feistmantel, Flora Gondw. Syst., vol. i, pt. 3, p. 172, pi. i, figs. 6a, lb. 1878. Ticniopteris Daintreei, Feistmantel, Palaeontogr., Suppl. iii, p. 110, pi. xiv, figs. 2, 3 ; p. 169, pi. xii, figs. 5, 5a. T. Daintreei, Etheridge, jun., Cat. Australian Fossils, p. 100. 1879. Angiopteridium spathulatum, Feistmantel, Flora Gondw. Syst., vol. i, pt. 4, p. 206, pi. i, figs. 8-13, 17, 18; pi. ii, tigs. 3, ? 5, ?6, pi. xv, fig. 11. 1882. Tceniopteris spathulata, Zeiller, Ann. Mines for 1882 (ii), p. 304, pi. x, fig. 8. 1883. T. Daintreei, Tenison- "Woods, Proc. Linn. Soc. New South "Wales, vol. viii, p. 117. 1886. T. Daintreei, Johnston, Papers & Proc. R. Soc. Tasmania for 1885, p. 375. 1890. T. Daintreei, Feistmantel, Mem. Geol. Surv. New South "Wales, Pal., No. 3, p. 114, pi. xxvii, figs. 4, 5 ; pi. xxviii, figs. 6, 6a. 1892. T. Daintreei, Etheridge, jun., Ann. Rep. Dept. Mines New South Wales for 1891, p. 269. T. (Angiopteridium) Daintreei, Jack & Etheridge, Geol. & Pal. Queensland, p. 371. TJENIOriEIUS. 125 1893. Tmriopteris spatulata, Oldham, Man. Geol. India, pi. opp. p. 176. 189K. Aiigiopteridium spathulntum, Dun, Eep. Austr. Assoc. Adv. Sci., p. 390. 1901. A. spathulatum, Etheridge, jun., in Anderson, 1st Eep. Geol. Surv. Natal, p. 72. 1902. Taniopteris spathulata, Zeiller, Flore Foss. Gites Cliarb. Tonkin, p. 74, pi. xiii, fig's. 6-12. Type. M'Clelland's, ? Mus. Geol. Soc. India, Calcutta ; McCoy's (T. Datntreei), ? Melbourne Museum, Victoria. Fronds simple, with parallel sides, entire, strictly linear- lanceolate or linear- spathulate, gradually contracted towards the base, petiole short or almost absent, apex spathulate, rounded or obtusely pointed. Leaves 3-12 mm. broad, 6-15 cm. long. Ilachis between -5 and 1 5 mm. broad, smooth or longitudinally striated, or with fine transverse flutings. Nerves very spreading, generally bifurcating at the base, to the number of 25-30 per centimetre at the margin. Lamina often marked by more or less pronounced transverse folds, dividing it up into compartments which are clearly convex, i-| mm. in height, each traversed by a bifurcating nerve. Zeiller 1 has recently given a full diagnosis of this species, of which the above is a translation, but which applies only to the Indian species T. spathulata. In 1892, however, Etheridge 2 expressed the opinion that McCoy's T. Datntreei, a type of frond from Austi'idasia, is identical with T. spathulata, and Dun 3 in his revision of the Australian Taeniopterids has adopted this conclusion. The full specific diagnosis given by Etheridge agrees very well, in its main essentials, with that of Zeiller. The chief characteristics of this species are the very long, narrow, step-shaped leaves, the thick midrib of nearly equal size throughout the length of the frond, and the lateral nervation, composed of veins distant or close, simple or bifurcating. On the other hand, the specimens described by Carruthers 4 as T. Datntreei, from the Tivoli coal-mines of Ipswich, Queensland, belong to a separate species, for which Tenison-Woods 5 instituted 1 Zeiller (02 3 ), pp. 74, 75. 2 Jack & Etheridge (92), p. 373. 3 Dun (98), p. 390. 4 Carruthers (72), p. 355, pi. xxvii, fig. 6. 5 Tenison-Woods (83), p. 117; see also Feistniantel (90), p. 115; Seward (03 1 ), p. 59. 126 T.EXIOPTERIS. the name T. Carruthersi. Both the species recorded hy Feistmantel 1 from the Cape Colony are also identical with T. Carruthersi. This species is probably more characteristic of the Mesozoic tban of the Palaeozoic rocks. It, however, occurs in Natal, where it has recently been recorded by Etheridge in association with the Glossopteris flora. Distribution. — Pernio- Carboniferous (Glossopteris flora) : — Natal. Triassie, Bkeetic, or Jurassic : — India (Upper Gondwanas), Tonquin, New South Wales, Victoria, Queensland, and Tasmania. Not represented in the British Museum collection from the Palaeozoic rocks. 4. Tseniopteris cf. M'Clellandi (Oldham & Morris). 1850. ? Tceniopteris acuminata, M'Clellaud, Rep. Geol. Surv. India, p. 53, pi. xvi, fig. 2. 1861. ? T.dantBoides (?), Bunburv, Quart. Journ. Geol. Soc, vol. xvii, p. 332, pi. x, fig. 2. 1863. Stangerites McClellandi, Oldham & Morris, Flora Gondw. Syst., vol. i, pt. 1, p. 33, pi. xxiii. 1869. Angiopteridiwm MacOlettandi, Schimper, Traite, vol. i, p. 605. 1877. A. MacClellandi, Feistmantel, Flora Gondw. Syst., vol. i, pt. 2, p. 96, pi. xlvi, figs. 5, 6. 1879. A. MacClellandi, Feistmantel, ibid., vol. i, pt. 4, p. 207, pi. i, figs. 14-16 ; pi. ii, fig. 4. 1880. A. cf. Mc'Clellandi, Feistmantel, ibid., vol. iii, pts. 2, 3, p. 92, pi. xxiA, figs. 4-7. 1882. A. cf. mcClellandi, Feistmantel, ibid., vol. iv, pt. 1, p. 31. Tceniopttris MacClellandi, Zeiller, Ann. Mines for 1882 (ii), p. 302, pi. x. fig. 5. 1902. T. MacCU llandi, Zeiller, Flore Foss. Gites Charb. Tonkin, p. 61, pi. ix, figs. 3-5. Type. ? Mus. Geol. Surv. India, Calcutta. Pinna? (? fronds) entire, with parallel sides, clearly contracted towards the base, ending at the apex in a rounded point. Pinna? '25-55 mm. broad. Rachis 1-3 mm. broad, longitudinally striated. Lateral nerves spreading, clearly decurrent at the base, bifurcating to form two simple or dichotomising nerves, either at their point of origin, or at some little distance from the midrib, clearly arched Feistmantel (89), p. 65, pi. ii, 6gs. 6 11. TiENTOPTERIS. 127 forwards towards their extremity ; at the margin from 20 to 30 in number per centimetre of length. 1 This species is still imperfectly known. It would appear to be essentially a Mesozoic type of frond, but some leaves associated with the Glossopteris flora in India may be compared with those occurring in the Rajmahal Series (Jurassic) of India. Of these, the earliest described was probably the frond which M'Clelland 2 figured as T. acuminata in 1850. His description and figure are, however, too imperfect to feel any confidence as to the identity of the specimen with Oldham & Morris' species. The Taniopteris figured by Bunbury 3 in 1861 (No. R. 10,367 in the Mus. Geol. Soc. London) is probably identical with Feistmantel's Lower Gondwana specimens from the Raniganj Group. They are linear, fairly narrow leaves, the lateral veins being simple, or more often dichotomised close to the point of origin with the midrib. The nerves are not very crowded. They are small imperfect fragments from Kamtlii in the Nagpur district, and are no doubt identical with some specimens in the British Museum described here from the same locality. "While admitting that the fronds under discussion show consider- able resemblance to the Mesozoic frond Tceniopteris IP Clellandi (Oldham & Morris), it does not seem possible to me to correlate them with this species with any confidence, partly on account of the fragmentary nature of specimens so far discovered, and partly because of the great difference in the age of the two deposits. Until better examples have been obtained, they may be referred to provisionally as Tceniopteris cf. J/' Clellandi. reistmantel 4 has also distinguished, undei'thename Ancftopteridium infarctum, some fronds which recall Bunbury's specimens in size and nervation. They differ, however, in the veins being very approximate, and passing out from the midrib at a more acute angle. Both Feistmantel's figures are of apical portions, and this may in part account for the acuteness of the lateral nerves, which in fronds of this type are often less oblique towards the apex. 1 See Zeiller (02 : »), pp. 61, 62. - M'Clelland (o0), p. 53, pi. xvi, fig. 2. :i Bunbury (61), p. 332, pi. x, fig. 2. 1 Feisfcmantel (SO), p. 93, pi. xxxivA, fii 128 PAT.JF.OVITTARIA. This frond, however, appears to me to be hardly worthy of specific distinction. Distribution. — Teeniopteris IV Clellandi occurs in the Rajmahal Series (Jurassic) of the Upper Gondwanas in India, and in beds of Rhaetic age in Tonquin. Fronds somewhat similar to those of the Mesozoic species occur in the Raniganj Group of the Damuda division in the JSagpur district, and in the South Rewah Basin in India. V. 7133. Small fragments of a Teeniopteris of a rather narrow type. The lateral nerves are fine, and fairly distant from each other. These specimens may be compared with T. M' Clellandi, especially with the figures given by Feistmantel (80), pi. xxiA, figs. 4-7, of the species. The veins do not appear to branch quite so close to the midrib as in T. danmoides. Ivamthi, India. Hunter Coll. V. 7132. A very fragmentary portion of a frond, showing only the midrib and a few indistinct nerves on one side. This specimen is probably of the same nature as the last. Ivamthi, India. Hunter Coll. 5. Teeniopteris, sp. (from Victoria). 1898. Teeniopteris Stceeti, McCoy, Proc. Roy. Soc. Victoria, n.s., vol. x, pt. 2, p. 285, and text-figure. Type. National Museum, Melbourne. ]\IcCoy, a few years ago, recorded the first Teeniopteris of Permo-Carboniferous age known from Australia, which he named T. tSweeti. Judging, however, from his text-figure, the specimen is too fragmentary to permit of an accurate specific diagnosis, and I prefer to term it Teeniopteris, sp., for the present. This genus occurs rarely with Gangamopteris in the Bacchus Marsh Sandstones of Victoria. Not represented in the British Museum collection. Genus PAL^OVITTARIA, Feistmantel, 1876. [Journ. Asiat. Soc. Bengal, vol. xlv, pt. 2, p. 368.] Fronds simple, entire, oval-lanceolate or oval-spathulate, con- tracted towards the base. Midrib only present in the basal portion PALJEOVITTABIA. 129 of the frond. Secondary nerves numerous, arising at a very acute angle from the midrib, simple or dichotomising. mil mil ^SSm Fig. ZQ.—Palceovittaria Kurzi, Feist. After Feistmautel. Nat. 130 SPHEXOPTKRIS. A single species only has been ascribed, I believe, to this genus, occurring in India, and also in the Rhsetic beds of Tonquin. A description of the latter specimens has been recently published by Zeiller, 1 and the present account is largely based on his extended diagnosis. Palseovittaria Kurzi, Feistmantel. (Text-fig. 30.) 1876. Paieeovittaria Kurzi, Feistmantel, Journ. Asiat. Soc. Bengal, vol. xlv, pt'. 2, p. 368, pi. xix, figs. 3, 3r?, 4, 4a. 1880. P. Kurzi, Feistmantel, Flora Gondw. Syst., vol. iii, pts. 2, 3, p. 91, pi. xlivA. 1890. P. sp., Schimper, in Zittel, Handb. Palaeont., pt. 2, p. 133. 1902. P. Kurzi, Zeiller, Flore Foss. Gites Charb. Tonkin, p. 81, pi. xvi, fig. 1. Type. Nos. 5325-6, Mus. Geol. Surv. India, Calcutta. Frond oval lanceolate or oval-spathulate, 12-15 cm. long, and 2-5-4-5 cm. broad, rounded at the apex. Median nerve scarcely visible, except in the basal portion of the frond, but indicated by a median fold. Lateral nerves very erect, spreading, a little incurved at the margin, simple or dichotomising once or twice, to the number of eight or ten in a centimetre of length. The fine specimen figured by Feistmantel shows nine very perfect fronds arranged as if springing from a common point of attachment. Nothing, however, is known of the mode of attachment, nor of the stem which bore these fronds. The fructification is also unknown. Tbis species is recorded from the Kaniganj Group of the Damuda division in India, and occurs in rocks of Triassic age in Tonquin. Kot represented in the British Museum collection. Genus SPHENOPTERIS, Brongniart, 1822. [Mem. Mus. d'Hist. Nat., vol. viii, p. 233.] Compound fronds, hi-, tri-, or quadripinnate, deeply or finely divided. Pinnules usually small or, more rarely, of medium size, as a rule more or less deeply divided, lobed, or dentate, rarely entire, conti acted at the base. Median nerve of each pinnule 1 Zeiller (02 3 ), p. si. SPHE-NOPTERIS. 131 usually giving off simple or bifurcating nervules to eacli segment, lobe, or tooth. Sphenopteris is one of the largest, and at the same time most unsatisfactory, of form-genera. In it are placed fern-like fronds of deeply cut or lobed habit, belonging to rocks of many different ages, and presenting no evidence with regard to the fructification. Consequently their true systematic position is unknown. This type of frond occurs not only in the Pala3ozoic, Mesozoic, and Tertiary periods, but can be matched among several genera of recent ferns. 1. Sphenopteris polymorpha, Feistmantel. (Text-fig. 31.) 1850. ? Peroptcris affinis, M'Clelland, Rep. Geol. Suit. India, p. 56, pi. xiii, figs. 11, 11a, 116. 1861. Peeopteris (?), Bunbury, Quart. Journ. Geol. Soc, vol. xvii, p. 331, pi. is, figs. 6-8. ? Cladophlebis, Bunbury, ibid., p. 332, pi. x, fig. 1. 1876. Spheiioptcns polymorpha, Feistmantel, Journ. Asiat. Soc. Bengal, vol. xlv, pt. 2, p. 356, pi. xvi, figs. 5-7 ; pi. xvii, figs. 1-3. 1S80. S. polymorpha, Feistmantel, Flora Gondw. Syst., vol. iii, pts. 2, 3, pp. 76, 77, pis. xv a, xviA, fig. 3 ; pi. xvi a bis, figs. 1-6. ? Cyathea cf. Tchihatcheffi, Feistmantel, ibid., p. 75, pi. xviA, figs. 1, 2, 4. 1882. Sphenopteris polymorpha, Feistmantel, ibid., vol. iv, pt. 1, p. 28. 1886. ? Cyathea cf. Tehihatchefi, Feistmantel, ibid., vol. iv, pt. 2, p. 23. 1893. Sphenopteris polymorpha, Oldham, Man. Geol. India, pi. opp. p. 162. 1902. S. polymorpha, Arber, Quart. Journ. Geol. Soc, vol. lvni, p. 12, pi. i, figs. 4, 5. *S. polymorpha, Arber, Geol. Mag., dec. iv, vol. ix, p. 348. Types. M'Clelland's and Feistmantel' s in (?) Mus. Geol. Surv. India, Calcutta. Frond tripinnate. Eachis broadly winged. Primary pinnae alternate, or sub-opposite, springing from the primary axis at a wide angle, greatly varied in shape, pinnate in the upper portion of the frond, with slightly sinuate or lobed pinnules ; median pinnae longer, also pinnate, with more deeply incised pinnules ; lower pinnae distinctly bipinnate, with denticulate pinnules. Secondary pinna; alternate or sub-opposite. Primary nerve of the ultimate pinnae giving off bifurcating nervules to the lobes of the pinnules. 132 SPHENOrTERIS. In some respects S. polymorpha recalls S. lobi folia and S. alata (see pp. 135, 138), but in the form of the terminal segment or pinnule, and in the regularly bifurcating secondary nervation, it does not agree with either of the Australian species. I have shown recently that this species occurs in the Newcastle Series of New South Wales. Specimens from this locality are in the Sedgwick Museum, Cambridge. Fig. 31. — Sphenopteris polymorpha, Feist. After Feistmantel. a, X 2 ; b, nat. size. I regard M'Clelland's Pecopteris affinis, so far as one can judge from Ids imperfect figure, to be probably identical with S. poly- morpha, a view which Feistmantel 1 at one time shared, although at 1 Feistmantel (80), p. 77. srnENOPTEKis. 133 a later period he concluded that they were on the whole distinct. It would, however, he impossible to adopt M'Clelland's specific name, since, though his plant is undoubtedly a Sphenopteris, the name Sphenopteris affinis had been applied earlier to a Lower Carboniferous frond by Lindley & Hutton. Feistmantel ' has also figured under the name Cyathea cf . TcMhatcheffi, Schnial., some fronds from the Talchir Coalfield, which he originally included with Sphenopteris polymorpha. The characters of these leaves are not described, nor is the nervation very clear from Lis figures. It is therefore difficult to decide how far they differ from the latter species, to which in some cases (cf. Feistmantel (80), pi. xvi a bis, fig. 1) they bear a close resemblance. For the present they may, perhaps, be best regarded as one of the varieties of the fronds included here under this specific name. It may be also pointed out that Schmalhausen's species is now known in all probability to be identical with Pecopteris leptophylla, Bunbury, from the Lower Permian of Portugal. 2 Sphenopteris polymorpha is known from the Barakar and Raniganj Groups of the Damuda division in India, and from the Newcastle Series in New South Wales. Not represented in the British Museum collection. 2. Sphenopteris Hughesi (Feistmantel). (Text-fig. 32.) 1877- Dicksonia cf. eoncinna, Feistmantel, Eec. Geol. Surv. India, vol. x, pt. 4, p. 198, figs. 10, 11 of plate. 1881. I>. Bughexi, Feistmantel, Flora Gondw. Syst., vol. iii, pts. 2, 3, p. 78, pi. xxiiiA, figs. 1, 3, 12, Via, 13. 1882. B. HugJu-si, Feistmantel, ibid., vol. iv, pt. 1, p. 28, pi. xii, figs. 3, 3a, U. 1902. Sphenopteris (Dieksonites) Bughesi, Zeiller, Pal. Indica, n.s., vol. ii, p. 6, pi. iv, figs. 1, 2, 2a-'ld. Type. ? No. 5207, Mus. Geol. Surv. India, Calcutta. Frond tripinnate or tripinnatifid, ultimate pinnse sub-alternate, elongate, somewhat flexuous, set at almost a right augle to the 1 Feistmantel (80), p. 75, pi. xvi a, figs. 1, 2, 4. 2 Zeiller (9G-), pp. 473, 174. 134 SPHENOPTKRIS. axis. Pinmiles sub-opposite, membranous, oblong, deeply lobed in lower portion, less lobed above. Nerves thin, flexuous, giving off branches to each lobe. Feistmantel 1 figured a pinnule which he regarded as showing sori "on the outer margin of the leaf lobes at the end of the veins." The same author also referred this frond to the recent genus Dicksonia on very insecure grounds, as Zeiller has recently pointed out, for there is at present no evidence as to the structure . "r- 11 -oc/ W Fig. 32. — Sphenopteris Hughesi (Feist.). After Feistmantel. A, nat. size; B, a pinnule, X 3. of the sporangia. Zeiller bus suggested Dicksonites as a secondary generic name following Sphenopteris. It appears to me, however, to be better to refer this species to the latter genus alone, until we know more of the fructification. It is very difficult in comparing highly divided, and at the same time fragmentary, portions of leaves of this type to arrive at any conclusion with regard to their identity with other and similar 1 Feistmantel (82 1 ), p. 28, pi. xii, fig. U. SPHENOPTERIS. 135 fragments, especially when, as in this ease, considerable variation in the degree of lobing is known to occur in different portions of the same frond, The Indian species S. Hughesi does, however, appear to be, in certain respects, nearly identical with the Australian fronds which I have grouped together here under S. lobifolia. The enlarged drawings of the pinnules given by Feistmantel, and to some extent also those of Zeiller, may be closely compared in regard to the nervation with McCoy's ' figures of S. hastata and S.flexuom. In the Indian species, however, the rachis apparently is not winged, and in the case of some of the pinnae the habit does not closely agree with that of the Australian fronds. It thus appears somewhat doubtful at present whether the Indian and Australian species are really distinct, but as I have not had any opportunity of seeing the specimens from India, I have maintained Feistmantel's species for the present. S. Hughesi is known only with certainty from the Raniganj Group of the Damuda division in India. Not represented in the British Museum collection. 3. Sphenopteris lobifolia, Morris. (PI. V, Figs. 2, 2a, 3.) 1845. Sphenopteris lobifolia, Morris, in Strzelecki's New South Wales, p. 246, pi. vii, figs. 3, 3«. 1847. S. lobifolia, McCoy, Aim. & Mag. Nat. Hist., vol. xx, p. 149. S. hastata, McCoy, ibid., p. 149, pi. x, figs. 1, la. S.Jlexuosa, McCoy, ibid., p. 150, pi. ix, figs. 4, 4a. S. plnmosa, McCoy, ibid., p. 150, pi. x, figs. 3, 3a. S. ger ma nits, McCoy, ibid., p. 150, pi. x, figs. 2, 2a. 1S49. ? S. lobifolia, Dana, in Wilkes' U.S. Explor. Exped., vol. x, p. 715, pi. xii, fig. 12. 1850. S. lobifolia, Unger, Gen. et Spec. Plant, foss., p. 128. S. hastata, Unger, ibid., p. 127. S.flexuom, Unger, ibid., p. 127. S. germana, Unger, ibid., p. 127. S. plumosa, Unger, ibid., p. 127. 1869. S. (Hymen.) plumosa, Schiraper, Traite, vol. i, p. 411. S. {Hymen.) germana, Schimper, ibid., p. 411. S. (Hymen. ?) flexnosa, Schimper, ibid., p. 411. S. (Hymen.) hastata, Schimper, ibid., p. 410. 1 McCoy (47), pi. x, fig. 1« ; pi. ix, fig. 4a. 136 SPHENOPTEEIS. 1S78. Sphenopteris lobifolia, Feistmantel, Palseontogr., Suppl. iii, p. 87. S. hastata, Feistmantel, ibid., p. 88. S.Jlexuosa, Feistmantel, ibid., p. 88. S. germana, Feistmantel, ibid., p. 88. S. plumosa, Feistmantel, ibid., p. 88. 1883. 6'. lobifolia, Tenison-Woods, Proc. Linn. Soc. New South Wales, vol. viii, p. 88. S. hastata, Tenison- Woods, ibid., p. 90. S.Jlexuosa, Tenison-Woods, ibid., p. 91. S. germana, Tenison-Woods, ibid., p. 91. S. plumosa, Tenison-Woods, ibid., p. 91. 1886. S. lobifolia, Johnston, Papers & Proc. R. Soc. Tasmania for 1885, p. 366. S. hastata, Johnston, ibid., p. 367. S.Jlexuosa, Johnston, ibid., p. 3G8. S. germanus, Johnston, ibid., p. 367. S. plumosa, Johnston, ibid., p. 367. 1887. S. lobfolia, Johnston, Papers & Proc. P. Soc. Tasmania for 1886, p. 173. 1890. S. lobifolia, Feistmantel, Mem. Geol. Surv. New South Wales, Pal., No. 3, p. 93. S. hastata, Feistmantel, ibid., p. 92. S.Jlexuosa, Feistmantel, ibid., p. 91. S. germana, Feistmantel, ibid., p. 92. S. plumosa, Feistmantel, ibid., p. 94. 1892. S. lobifolia, Jack & Etheridge, Geol. & Pal. Queensland, p. 190. S.Jlexuosa, Jack & Etheridge, ibid., p. 190. 1896. S. Morrisiana, Johnston, Papers & Proc. E. Soc. Tasmania for 1891-5, p. 58, figs. 14, 15. 1902. S. lobifolia, Arber, Quart. Journ. Geol. Soc, vol. lviii, p. 11. S.Jlexuosa, Arber, ibid., p. 11. S. hastata, Arber, ibid., p. 12. S. germana, Arber, ibid., p. 13. Mertensia lobifolia, Shirley, Bull. 18, Geol. Surv. Queensland, p. 9, pi. v. Sphenopteris Jlexuosa , Shirley, ibid., p. 10. Types. Morris's, No. 13,530, Mus. Geol. Soc. London; McCoy's, Nos. 2, 3, 5, 7, Foreign Hunt Coll., Sedgwick Mus., Cambridge. Frond probably tripinnate, triangular in contour. Kacbis winged. Pinnae somewhat distant, elongate, broadly or narrowly linear- lanceolate, often slightly falciform, opposite in the upper portion of the frond, alternate below. Pinnules contracted at tbe base, somewhat distant, linear -lanceolate or oval -lanceolate; obtusely lobed. Median nerve of pinnule sinuate, supplying simple or bifurcating branches to the lobes. SPHENOPTERIS. 137 I have included here, as has already been suggested by Shirley ! and Johnston, 2 the very fragmentary fronds described by McCoy in 1847 as Sphenopteris hastata, S.flexuosa, S. germana, and S plumosa, which are probably only different portions, varying somewhat in details, of the same frond or of fronds of the same plant. The specimen on which the species S. lobifolia was founded shows narrow pinnules with four pairs of lobes, each of which is supplied by a bifurcating or doubly bifurcating vein. S. germana, McCoy, is almost indistinguishable from Morris's specimen. S. plumosa appears to be founded on a fragment of a pinna, in which the pinnules are a little larger and the lobes more numerous. In S. flexuosa the lobes are few but larger and broader, while in S. hastata the pinnule is only crenate and hardly lobed. All these so-called species are founded on fragments of pinna?, far too imperfect to obtain any idea of the variation in the form and lobing which may occur on the same frond, and for the present they may best be retained under one specific name. Johnston 3 has given a full description of a more complete specimen from Tasmania, which he terms S. Morrisiana. It would seem best, however, to retain Morris's specific name as the oldest applied to this frond. More recently Shirley 4 has described a fragment similar in form to Morris's figured specimen, which he regards as showing the fructification. It is described as occurring "in masses on the first fork of the primary veins of each pinnule, there being one sorus for each lobe ; and, as in Gleichtnia, their modern allies, the margins of the lobes seem to have been recurved." The details of the structure of the sporangia do not appear, however, to have been made out, nor are the drawings given at all clear, and it would therefore seem premature to insist upon its botanical affinities, or to refer this frond to the recent genus Mertensia as Mr. Shirley has proposed. Sphenopteris lobifolia is known only from the Permo-Carboniferous rocks of Australasia. 1 Shirley (02), p. 10. - Johnston (87-) and (94), pt. 2, p. 58. 3 Johnston (94), pt. 2, p. 58, figs. 14, 15. 4 Shirley (02), p. 10, pi. v. 138 SrnENOI'TEKIS. V. 7288. PI. V, Figs. 2, 2a. A small portion of a frond, consisting- of a bipinnate, slightly winged axis, with three fairly complete pinnae on one side and one on the other. The pinnae are alternate, and lanceolate in contour. The axis of each pinna is grooved. The pinnules are more or less oval, and the margin is slightly lobed ( PL V, Fig. 2a). They recall those figured by McCoy as Sphenopteris hastata. Port Stephens, New South Wales. Odinheimer Coll. V. 6251. PI. V, Fig. 3. A rather larger, and nearly complete fragment of an apical portion of a frond, similar to the last. Other detached pinnae also occur on the same specimen. Port Stephens, New South Wales. Odinheimer Coll. V. 7220. Two small fragments of detached pinnae, with oval, slightly lobed segments, showing the characteristic nervation of k. lubifolia very clearly. Port Stephens, New South Wales. Odinheimer Coll. 4. Sphenopteris alata (Brongniart). 1834-36. Pecopteris (data, Brongniart, Hist. Yeget. foss., p. 361, pi. cxxvii. 1836. Aspidites alatus, Goppert, Die Foss. Farn., p. 3oS. 1838. Sphenopteris alata, Steinberg, Flora Vonvelt, Heft vii, p. 131. 1845. 8. alata, var. exilis, Morris, in Strzelecki's New South Wales, p. 246, pi. vii, figs. 4, 4a. 1847. S. alata, McCoy, Ann. & Mag. Nat. Hist. vol. xx, p. 149. 1S50. S. alata, linger, Gen. et Spec. Plant, foss., p. 124. 1869. S. (Hymen.) alata, Schimper, Traite, vol. i, p. 411. 1878. S. alata, Feistmantel, Pakeontogr. , Suppl. iii, p. 87. S. alata, var. exilis, Feistmantel, ibid., p. 88. 1883. S. alata, Tenison- Woods, Proc. Linn. Soc. New South Wales, vol. viii, p. 89. S. alata, var. exilis, Tenison-Woods, ibid., p. 90. 1S86. S. alata, Johnston, Papers & Proc. E. Soc. Tasmania for 1885, p. 366. 1890. S. alata, Feistmantel, Mem. Geol. Surv. New South Wales, Pal., No. 3, p. 88. S. alata, var. exilis, Feistmantel, ibid., p. 89. 1902. S. alata, Arber, Quart. Journ. Geol. Soc, vol. lviii, p. 9. <$'. alata, Shirley, Bull. 18, Geol. Surv. Queensland, p. 11. Type. ? Museum of the University of Edinburgh. Frond bipinnate above, tripinnate below, sub- triangular in srnENorTERis. 139 contour. Raehis smooth, winged. Pinnules either confluent or contracted at the base, fairly broad, more or less deeply and irregularly pinnatifid, lobes entire or dentate. Nervation of the pinnules pinnate, simple, or bifurcate. This species is very little known. According to Feistmantel only two specimens have been described, one of which, rirongniari's type, was obtained from the coal-mines on the Hawkesbury River, near Port Jackson, and the other, a poor specimen mentioned by McCoy, now in the Sedgwick Museum, Cambridge, is from Newcastle, New South Wales. I have had an opportunity of examining the latter, and it appears to me to be perhaps distinct from S. luhifolia in the habit, and in some details of the nervation, although Shirley 1 has recently expressed the opinion that these two species should be united. The pinnules are broader (I mm.), and more ovate in many instances in the Cambridge specimen. The frond, which is incomplete, is 1 65 cm. long, and more than 20 cm. across. The pinnules are sub-opposite, with three or more bluntly and obliquely cut segments on either side, each of which is supplied by a simple or bifurcating nervule. I have pointed out elsewhere 2 in some detail the confusion which has arisen between this plant and the plant originally described by Brongniart as Sphenopteris alaia, now known as S. Grandini (Gopp.). Sphenopteris alata (Brong.) is known only from the Newcastle Series of New South "Wales. 3 Not represented in the British Museum collection. 5. Sphenopteris, sp. (from India). V. 7193. Badly preserved fragments, probably of a Sphenopteris, in which the nervation is not clearly seen. Nagpur, India. Sankey Coll. V. 7193ff. A branched axis, presumably part of a Sphenopterid frond, but badly preserved, especially as regards the nervation. Nagpur, India. SanJcey Coll. 1 Shirley (02), p. 10. 2 Arber (02 1 ), p. 10. 3 Mr. Etheridge, jun., has expressed the opinion that this species probably also occurs in the Hawkesbury Sandstone ; see Feistmantel (90), p. 89, footnote. 140 PECOPTERIS. 6. Sphenopteris, sp. (from Australia). V. 4205. A portion of a frond, with the nervation preserved in places, but too fragmentary to determine specifically. 7. Sphenopteris, sp. (from S. Africa). Zeiller 1 has figured a fragmentary specimen from the Transvaal, with dichotomising nerves, apparently belonging to this genus. In addition, there are two specimens, also very fragmentary, to be briefly referred to here. V. 3622. Figured by Seward (97 1 ), p. 332, text-fig. la. A very minute fragment, with forking nerves. Probably a pinnule of a Sphenopteris frond. Casey's Township, Transvaal. Prcs. ly B. Draper, Esq., 1897. V. 3260. Imperfect fragments of fronds. A trace of the nervation can be seen here and there. Bedford, Cape Colony. Pres. by B. B. Fraser, Esq., 1893. Genus PECOPTERIS, Brongniart, 1822. [Mem. Mus. d'Hist. Nat., vol. viii, p. 233.] Bi-, tri-, or quadripinnate fronds. Pinnules attached by their whole base to the rachis, not contracted below, usually with parallel sides or margins slightly converging towards the apex, generally entire, with obtuse apex, contiguous, and often set almost at right angles to the rachis. Median nerve clear, giving off pinnately arranged, simple, or dichotomising nervules. The form- genus Pecopteris is now generally restricted to include only certain fern- like fronds of Upper Carboniferous and Permian age, although this type of leaf is met with in the Mesozoic and Tertiary rocks, and can be closely matched among living ferns. Pecopteris phegopteroides (Feistmantel). 1876. Alethopteris phegopteroides, Feistmantel, Journ. Asiat. Soc. Bengal, vol. xlv, pt. 2, p. 362, pi. xviii, figs. 1, la, 2, 2a. 1881. A. phegopteroides, Feistmantel, Flora Goudw. Syst., vol. iii, pt. 3, p. 81, pi. xviii a, figs. ], la, lb. 1 Zeiller (96 1 ), p. 371, pi. xviii, fig. 1. CLAD0PULEI5IS. 141 Type. No. 5183, Mus. Geol. Surv. India, Calcutta. Frond large, bipinnate. Ilachis strong, punctate. Pinnae set almost at right angles to the rachis. Pinnules longest in the median portion of the pinna, shorter towards base and apex, inseiied slightly obliquely, oblong, incurved, contiguous, connate at the base. Median nerve of pinnule distinct, persisting to the apex, giving off simple secondary nerves at an acute angle. The fine frond described b\ T Feistmantel from the Raniganj Coalfield appears to me to belong to the genus Pecopteris rather than to Alethopteris. In habit and nervation it recalls some of the Pecopterids of the Upper Coal Measures and Lower Permian rocks of the Northern Hemisphere. The fructification is unfortunately unknown, and it is therefore not safe to insist, as Feistmantel has done, on its near relationship with the recent genus Phegopteris, which it somewhat closely resembles in the habit of the frond. P. phegopteroides is known only from the Raniganj Group of the Damuda division iu India. Not represented in the British Museum collection. Genus GLADOPHLEBIS, Brongniart, 1849. [Tableau Genr. Veget. foss., p. 25.] "Fronds pinnately divided, pinnae spreading, lobes or pinnules attached by an entire base or slightly contracted towards the place of attachment, rarely somewhat auriculate, acuminate, or obtuse, occasionally dentate, especially at the apex, not rarely sub-falcately curved upwards, midrib strong at the base and towards the summit dissolving into branches, secondary veins given off at a more or less acute angle, dichotomous a little above the base, and repeatedly dichotomous." l Cladophlebis is an artificial form-genus especially characteristic of the Triassic, Jurassic, and Wealden rocks. A full account of fronds of this type will be found in Mr. Seward's Catalogues of the Mesozoic Plants in the British Museum. Seward (94), p. 88. 142 CLADOPHLEBIS. 1. Cladophlebis Roylei, Arber. (Text-fig. 33.) 1833. Pecopteris Lindleyana, Boyle, Illust. Bot. Himal. Mounts., p. xxix*, pi. ii, fig. 4. 1836. Aspidites Lmdleyanus, Goppert, Die Foss. Farn., p. 360. 1845. Pecopteris Lindleyana, Unger, Synops. Plant, foss., p. 96. 1850. P. Lindleyana, M'Clelland, Bep. Geol. Surv. India, p. 56, pi. xiii, figs. 10a-c. P. Lindleyana, Unger, Gen. et Spec. Plant, foss., p. 171. 1861. Alethopteris Lindleyana, Schimper, Traite, vol. i, p. 568. 1876. Pecopteris Lindleyana, Feistmantel, Bee. Geol. Surv. India, vol. ix, pt. 3, p. 76. Alethopteris Lindleyana, Feistmantel, Journ. Asiat. Soc. Bengal, vol. xlv, pt. 2, p. 360, pi. xx, fig. 7. 1880. A. Lindleyana, Feistmantel, Flora Gondw. Syst., vol. iii, pts. 2, 3, p. 80, pi. xviiiA, figs. 2, 2a ; pi. xixA, figs. 3, 4 ; pi. xxiiiA, figs. 11, 11"; pi. xxxixA, figs. 10, 11. 1893. A. Lindleyana, Oldham, Man. Geol. India, 2nd pi. opp. p. 162. 1898. ? A. Lindleyana [?), Shirley, Bull. 7, Geol. Surv. Queensland, p. 20, pi. xiii, fig. 1. 1901. Cladophlebis Jioy/ci, Arber, Geol. Mag., dec. iv, vol. viii, p. 548. Type. V. 4192, Geol. Dept. British Museum (Nat. Hist.). Frond bipinnate; pinna? spreading. Pinnules attached by their whole base, contiguous, oblong-oval, entire, or sinuate. Median nerve slender, extending to the apex ; secondary nerves arising at a sub-acute angle, dichotomising. Fructification of the ?Poly- podiaceous type. The earlier figures of this species given by Royle, and especially by M'Clelland, are somewhat misleading. Feistmantel has, how- ever, shown the nervation fairly accurately (Text-fig. 33). It is not possible, as I pointed out some years ago, 1 to adopt Royle's specific name, as the term Pecopteris Lindleyana had been earl un- applied by Presl to a fern now known as Coniopteris arguta (L. & H.). Nor does this frond seem to fall naturally, as Schimper and Feistmantel have suggested, within the limits of the genus Alethopteris — a term now generally used in a restricted sense. The habit and especially the nervation appear to agree more closely with those of fronds included by Seward and others in the genus Cladophlebis, and although this genus is generally reserved for leaves of Mesozoic age I have ventured to refer this species to it, 1 Arber (01), p. 549. CLADOniLEBXS. 113 for the Indian fronds, as Feistmantel pointed out, closely resemble some of tliose known from the Lower Oolite of Britain and else- where. As a specific name I adopted ' Roylei i in 1901 in honour of the first discoverer of this species. Fertile fronds of this species have been described and figured by Feistmantel. Tbe sporangia are inserted on the lateral veins of the leaflets, midway between the midrib and the margin. There are generally from six to eight sori in each row. Feistmantel has Fig. 33. — Cladophlebis Roylci, Arber. After Feistmantel. A, nat. size; B, a pinnule, X 2. concluded that these fronds belong to the Polypodiacea, although the structure of the sporangia does not appear to have been made out. This frond is known from the llaniganj Group of ihe Damuda division in India. It has also been identified from beds probably of Rhsetic age in Queensland. 1 V. 4192. Figured by Royle (33), pi. ii, fig. 4. Type. Royle's type is a large fragment of a frond about 18 cm. long, and probably more than '20 cm. wide when perfect. The pinna? are badly preserved, and the nervation is very indistinct. The pinnse, of which several are seen, are alternate, and lanceolate in form. The pinnules are attached by a broad base, and are rounded at the apex. They average 1*2 cm. in length, and 8 mm. in breadth. 1 Jack & Etheridge (92), p. 370, pi. xvii, figs. 3,4; Shirley (98), p. 20, pi. xiii, fig-. 1. 144 MEIUANOPTERIS. The nervation can just be made out in places. There is a strong median nerve, giving off dichotomising secondary nerves of the Cludophltlis type. Bardwan Coalfield, India. V. 10,893. A small fragment of a pinna, not unlike that figured on Text-fig. 33a on p. 143. The nervation, however, is rather indistinct. Near Nagpur, India. Sanlcey Coll. 2. Cladophlebis, sp. (from India). 1876. Ahtltopteris cf. whitlyensis, Feistmantel, Journ. Asiat. Soc. Bengal, vol. xlv, pt. 2, p. 362, pi. xxi, figs. 6, 6a. 1881. Asplenium whitbyense, Feistmantel, Flora Gondw. Syst., vol. iii, pts. 2, 3, p. 79, pi. xixA, figs. 2, 2a; pi. xIa, figs. 2, 3. Feistmantel figured some very imperfect fronds from the Raniganj Group, India, which he regarded as identical with Alcthupteris indica, 0. & M., and A. australis, Morr., both of which are of Mesozoic age. It would be interesting if there were trustworthy evidence for the appearance of fronds of this type in rocks as old as the Pernio- Carboniferous. Feistmantel' s specimens, however, are far too fragmentary to warrant any such conclusion. Genus MERIANOPTERIS, Heer, 1877. [Flora Foss. Helvetia?, p. 88.] Heer described this genus as follows : — " Frons dimorpha, pinnulis fertilibus et sterilibus stipite communi affixis ; pinnulas fertiles contractas, angustae, nervo medio valido, nervillis secundariis simplicibus parallelis, soris interpositis, sori rotundato in quavis pinnula biseriales. Frons sterilis tripinnata, speciosa, pinnis secundariis elongatis, segmentis (vel pinnulis) nervo medio arcuato, nervis secundariis dichotomis, infimis, in arcuam acutam anasto- mosatis." The genus was founded for the reception of a frond from the Keuper rocks of Switzerland. Meriauopteris major, Feistmantel. (Text-fig. 34.) 1881. Merianopteris major, Feistmantel, Flora Gondw. Syst., vol. iii, pt. 3, p. 83, pi. xixA, figs. 9-11. Type. Nos. 5192-4, Mus. Geol. Surv. India, Calcutta. Frond tripinnate, large. Secondary pinnae broadly elongate, MERIANOPTERIS. 145 lanceolate, narrowed slightly towards the apex, pinnatisect or pinnatifid. Pinnules or lobes very delicate, rounded at the apex. Median nerve distinct, somewhat curved towards the apex. Secondary nerves arising at an acute angle, dichotomous, and flexuous. The lowest pair of lateral nerves, arising from the base of the median vein, dichotomise, and then arch to join the corresponding nerves of the pinnules on either side. Feistmantel, in his original description, states that the two lower lateral nerves of each pinnule "join with the same of the adjoining Fig. 34. — Merianopteris major, Feist. After Feistmantel. Nat. size. leaflets in a pointed arch." This union is not, however, clearly shown in his figures. This frond presents an unusual type of nervation, and is unfortunately only known from Gondwanaland by a few specimens figured by Peistmantel from the Eaniganj group. Tenison- Woods, 1 however, has recorded this species from the Ballimore Coalfield of New South Wales, apparently on a higher horizon, but I am not convinced that his plant is either specifically or even generically 1 Tenison-Woods (83), p. 114, pi. vi, figs. 2, 3. 146 T5ELEMN0PTERIS. identical with the Indian frond, although Feistinantel l apparently accepted this identification. Merianopteris major is known with certainty only from the Eaniganj Group of the Damuda division, India. Not represented in the British Museum collection. Fig. 35. — Belemnopteris Wood-Masomana, Feist. After Feistmantel. Nat. size. Genus BELEMNOPTERIS, Feistmantel, 1876. [Journ. Asiat. Soc. Bengal, vol. xlv, pt. 2, p. 370.] Fronds simple, broadly sagittate, petiolate. Primary nerves three in number, the median nerve being the stronger. Secondary nerves anastomosing. 1 Feistmantel (90), p. 113. tsaroxius. 147 This genus was founded for the reception of some remarkable Indian fronds, of which only a few specimens are known. So far as I am aware, they are its only representatives. Belemnopteris Wood-Masoniana, Feistmantel. (Text-fig. 35.) 1876. Belemnopteris Wood-Masoniana, Feistmantel, Journ. Asiat. Soc. Bengal, vol. xlv, pt. 2, p. 371, pi. xx, figs. 1, 2. 1881. B. Wood-Masoniana, Feistmantel, Flora Gondw. Syst., vol. iii, pts. 2, 3, p. 112, pi. xliiiA, figs. 3, 4. Frond broadly sagittate, entire. Apex obtuse. Basal lobes obtusely acuminate, the median primary nerve strong, the two lateral primary nerves of the lobes thinner, all attenuated towards their extremities. Secondary nerves arising at a sub-acute angle, anastomosing, forming hexagonal or polygonal meshes. Fructi- fication unknown. The sagittate form of these fronds, and the large open meshes recalling the Glossopterids of the type of G. retifera, give a very distinctive appearance to these fossils, and it is much to be hoped that further specimens will be obtained before long which may give a clue to the affinities of this interesting plant. Belemnopteris Wood-Masoniana is only known from the Raniganj group of the Damuda division in India. Not represented in the British Museum collection. STEMS OF FERNS. Genus PSARONTUS, Gotta, 1832. [Dendrolithen, pp. 27, 28.] Arborescent stems of ferns, in which the anatomical structure is preserved. Central region of the stem consisting of a vascular cylinder of variable dimensions, with a more or less considerable number of steles. Vascular cylinder surrounded by a false cortex, often of great thickness, composed of a dense mass of adventitious roots. Steles of the central cylinder in the form of flattened bands, bent or sinuate, arranged on several concentric circles in a ground- mass of parenchymatous conjunctive tissue, anastomosing here and there, often separated by concentric bands of sclerenchyma, especially towards the periphery of the cylinder. These bands also 148 PSARONIUS. anastomose from time to time, and sometimes the entire vascular cylinder is surrounded by a closed sclerenchymatous band. Steles concentric, with a central strand of scalariform tracbeides, sur- rounded by phloem. The leaf-traces arise from the peripheral steles of the cylinder, and have the form of bands concave towards the centre of the stem. The adventitious roots have a central axis consisting of a polyarch stele, formed of a variable number of radiating woody strands, with a similar number of phloem strands between them. Inner cortex of the root parenchymatous, sometimes solid, sometimes lacunal ; outer cortex sclerenchymatous. The Psaronii form a numerous group, which has been sub- divided by Zeiller into three divisions, based on the phyllotaxis. These are the Pulystichi with numerous vertical rows of leaves, the Tefrastichi with four rows, and the Distichi with two rows. The species described here belongs to the Tefrastichi. Several species of Psaronim are known to have been the stems of plants possessing the Pecopteris type of foliage, especially the Pecopterids bearing the fructifications known as Scolecopteris and Asterotheea. Such plants present a close resemblance to a modern family of ferns, the Marattiacese. Distribution. — Pernio- Carboniferous (Northern Type) : — In the Upper Carboniferous and Permian of Europe, especially the latter. Pernio- Carboniferous (Glossopteris Type): — Brazil. Psaronius brasiliensis, linger. (Plate VII.) 1831-50. Psaronius brasiliensis, linger, in Martius' Hist, natur. Palmarum, i, p. lxx, pi. geol. i, fig. 4. 1872. P. brasiliensis, Brongniart, Bull. Soc. hot. France, ser. v, vol. xix, p. 3. 1889. P. brasiliensis, Bennett & Murray, Handb. Cryptog. Botany, p. 123, text-fig. 95. 1890. P. brasiliensis, Zeiller, Bass, liouill. et perm. d'Autun et d'Epiuac, fasc. ii, pt. 1, p. 246, pi. xxi. fig. 1. 1891. P. sp., Solms-Laubaoh, Fossil Botany, Eng. edit., p. 170. 1895. P. brasiliensis, Zeiller, Bull. Soc. Geol. France, ser. in, vol. xxiii, p. G05. 1900. P. brasiliensis, Scott, Stud. Foss. Botany, p. 271, text-fig. 96. 1904. P. brasiliensis, Sohns-Laubach, Festschr. P. Ascherson's Sicbzigsten Geburtet., p. 18. rsAROxius. 149 Co-types. The earliest specimen, figured by Unger (jS"o. 1446, Museum d'histoire naturelle, Paris), is an imperfect fragment, showing only the pseudo-cortex. Although, strictly speaking, it is the type-specimen, it is impracticable to regard it as such, since it does not include the central vascular cylinder. There is also a suspicion that it was derived from the same stem in Brazil as all the other specimens now in various European museums, but it is not quite certain that this is the case. The real type would appear to be a large stem in the Museum at Bio de Janeiro, Brazil. From this trunk slabs have been cut at various periods, among which are those in the Museum d'histoire naturelle (No. 1445) described by Brongniart, who gave the first adequate description of this species ; those in the British Museum (V. 9002 and V. 5388), all of which, I believe, are now in the Geological Department ; and those at Strassburg. Solms-Laubach 1 has recently traced in great detail the history of these specimens. It would perhaps be best to regard the Bio specimen and its derivatives in Paris, London, and elsewhere as co-types with Unger's specimen. Central vascular cylinder surrounded by a continuous and closed sclerenchymatous sheath. Peripheral steles four in number, separated in part from the central region by sclerenchymatous bands, slightly curved outwards, with inrolled extremities, or some- times folded in the form of a ring. Steles of the central region, lying internal to each of the peripheral steles, two or three in number, strongly curved, arranged radially, often divided into two, sometimes forming a closed ring. Steles lying between the peripheral steles, and often anastomosing with them at their extremities, more or less bent, longer than those nearer the centre of the cylinder. Leaves disposed in four vertical series, probably in opposite or sub-opposite pairs. Leaf-traces on leaving the cylinder having the form of an open curve, with extremities bent in a hook -like manner. Conjunctive tissue of the central cylinder and pseudo-cortex continuous. Inner cortex of roots not lacunal, apparently often containing numerous gum-canals arranged more or less in a circle. Vascular cylinder composed of 6-8 bundles. 1 Solms-Lauback (04). 150 I'SAEONIUS. Professor Zeiller has given a very full description of the detailed anatomy of this stem, from which the above account is in the main derived. The essential features of the complex structure of the vascular cylinder can be made out from the photograph of one of the British Museum slabs reproduced on Plate VII. The central cylinder is in form nearly square, with rounded angles. For the sake of clearness we shall describe the steles with reference to the sides and angles of this square. We may first consider the steles bordering on the sides of the square. At the upper and lower borders there are two well-marked invaginations of the sclerenchymatous band, which is clearly seen completely surrounding the central cylinder. The tissues in these two bays are not preserved. These invagi- nations, however, mark the position at which two outgoing leaf- traces have just become free from the cylinder, and have passed out to supply a pair of opposite or sub-opposite leaves. At the sides of the square two long and slightly curved steles are seen. These will pass out from the cylinder as leaf-traces at a higher level to supply the next pair of leaves alternating with, and situated above the first pair. The third pair of leaf-traces will arise in the same position as the first, i.e. at the upper and lower sides of the square. The steles lying immediately below the invagination at the upper side of the square will anastomose to form one of these leaf-traces. Thus the four sides of the square are, as a rule, occupied by four leaf- traces supplying four vertical rows of leaves. The diagonal corners of the square are occupied by four long, curved steles, lying just inside the sclerenchymatous band, and projecting slightly beyond the leaf-traces. These we may speak of as the peripheral steles. It is from these peripheral steles that the adventitious roots arise. They also contribute to the leaf- traces, with which they anastomose before the latter become free from the central cylinder. The internal portion of the square is occupied by several smaller steles, which anastomose both with the more external steles, and among themselves. Psaronius brasiliensis is one of the most beautiful of fossil plant- remains, the colours, especially of the pseudo-cortex, as seen in polished slabs, being exceptionally striking. CAULOPTERIS. 151 The British Museum specimens were brought to this country by Claussen from Brazil, probably in 1841. The original stem in the Museum at Rio de Janeiro, from which they were derived, has been briefly described and figured by Solms-Laubach. 1 Distribution. — Psaronius Irasiliensis is known only from Brazil. The specimen figured by TJnger was found between Oeiras and Sao Gongala d'Amarante, in the province of Piauhy. The locality from which the stem in the Rio Museum was obtained is unknown. V. 9002. Plate VII. Also figured by Bennett & Murray, Handb. Cryptog. Botany, p. 123, text-fig. 98, 1889. Co-type. Three polished slabs of the original type-specimen are registered under this number, part of one being figured on Plate VII as seen in transverse section. The form of the slab is oval, measuring 22 - 5 cm. along the greater, and 18 - 5 cm. along the smaller axis. The pseudo-cortex, composed of adventitious roots, is considerably larger than the central cylinder. The central cylinder, the structure of which is described above, is nearly square, the diagonals measuring 11 cm. Brazil. Claussen Coll. V. 5388. A small fragment of one of the above slabs, showing part of the pseudo-cortex and vascular cylinder. Brazil. Claussen Coll. Genus CAULOPTERIS, Lindley & Hutton, 1832. [Foss. Flora, vol. i, p. 121 (explan. to pi. xlii).] External impressions or casts of stems of arborescent ferns, possessing the Psaronius type of internal structure, and showing the bark and the scars of the petioles. Petiolar scars large, oval, arranged in longitudinal series, generally fairly contiguous, bearing a large, concentric, closed, elliptical or horseshoe-shaped scar or band, near the apex of which a Q-shaped vascular print occurs. The bark between the leaf-scars shows traces of many adventitious roots, like those which occur on the stems of living tree-ferns. Solms-Laubach (04), text-fig. on p. 23. 152 CAULOPTERIS. Caulopteris and Ptychopteris are impressions of stems, the one showing the true external surface, the other the partially decorti- cated condition of the same plant; the internal structure being known as Psaronius. Distribution. — Permian and Upper Carboniferous (Northern Type) : — Europe and 'N. America. [Caulopteris ?] ' Adamsi, Feistmantel. 1878. Caulopteris (?) Adamsi, Feistmantel, Palteontogr., Suppl. iii, p. 94, pi. xii, figs. 1, 2. 1883. C. (?) Adamsi, Tenison-Woods, Proc. Linn. Soc. New South "Wales, vol. viii, pp. 132-3. 1890. C. Adamsi, Feistmantel, Mem. Geol. Surv. New South "Wales, Pal. No. 3, p. 135, pi. xxi, figs. 1, 2. Type. ? Australian Museum, Sydney. Feistmantel describes this species thus : — " The specimen is hardly sufficiently complete to decide its nature and systematic position with absolute certainty ; but supposing it to be, what it most probably is, a fragment of a fern trunk, and, taking the disposition of the scars to be quincuncial, I thought it would be more correct to place this specimen with the genus Caulopteris, as there are not sufficient characters for placing it anywhere else, or for making it the type of a new genus. I have given the following diagnosis : — Sp. Char. — Trunco arboreo, mediocri, superficie cicatricibus ramorum (foliorum) notato ; cicatricibus in quincunce (spiraliter) dispositis, transverse oblonge-ovalibus, paulum prorninentibus, lateribus linea decurrente notatis ; superficie interna cicatriculis parvulis vasalibus, 7 ad 8, repleta." I am unable to see any resemblance in Feistmantel's figures to the European fern-stems included under Caulopteris, and I have no hesitation in saying that the attribution of these specimens to that genus is unjustified. As Feistmantel admits, they are too fragmentary to allow of exact determination. Without seeing the actual specimens I am unable to refer them to any other genus, and I have, therefore, retained Feistmantel's generic name See footnote, p. 30. LYCOPODIALES. 153 within square brackets, thus implying that, while it is incorrect, there is not sufficient evidence to justify removal at present. This species is known only from the Newcastle Series of New South Wales. Not represented in the British Museum collection. INCERT^E SEDIS. OBSC URE S TEM- STB UC TUBES. There are several fairly large, and sometimes branched stem- structures in the British Museum collection from the Nagpur district, India, some of which may possibly be fern-stems, but they do not offer sufficiently good morphological characters to permit of identification. 46.705. A stout stem, 6 cm. across, and much branched above. Surface smooth, with deep irregular grooves. Silewada, 12 miles north of Nagpur, India. Hunter Coll. 46.706. Branched stems, in one instance, tripinnate. Some of the branches are 15 cm. long, and quite smooth. The main axis of the tripinnate stem is 3 cm. across ; the secondary branch is 3 mm. broad at the base, and bears fine tertiary branches. Silewada, 12 miles north of Nagpur, India. Hunter Coll. V. 7201. A portion of a smooth stem, 13 cm. long and approximately 25 cm. broad. Silewada, 12 miles north of Nagpur, India. Hunter Coll. V. 7149. A badly pi'eserved stem, 17-5 cm. long. Near Nagpur, India. Hunter Coll. V. 9764. Another specimen, 7 cm. long and 6*3 cm. broad. Silewada, 12 miles north of Nagpur, India. Hunter Coll. Class LYCOPODIALES. Stem nearly always well developed, herbaceous, shrubby, or arborescent, dichotomously or monopodially branched, more rarely uubranched. Leaves simple, undivided, as a rule small in comparison with the size of the stem. Sporangia isosporous or heterosporous, simple, exannulate, usually borne singly on the upper surface or 154 LEriDODENDREiE. in the axil of more or less modified leaves, which may he aggre- gated into strobili, or occupy a similar position to the sterile leaves. In some extinct genera the mature sporangium is enveloped by an integument, and thus was of the nature of a seed. The living Lycopods belong to the following genera, of which Lycopodhm (the Club - mosses), Phylloglossum, Psilotnm, and Tmesipteris are isosporous, while Selaginella and Isoetes are heterosporous. In Palaeozoic times this group was a very large and diversified one, and then attained to its maximum development ; the chief geriera being Lepidodendron, LepidopMoios, Sigillaria, and Bothrodendron. All these Palaeozoic types are characterised by an arborescent habit, and the power of secondary growth in thickness. In the Mesozoic and Tertiary rocks the Lycopods are much less abundant, the genus Lycopodites being the chief representative, including plants more closely resembling the recent genera than the arborescent Palaeozoic types. In the Northern Permo-Carboniferous continent the Lycopods formed one of the most important elements in the flora. It is interesting to note that, while this group appears to have been entirely absent from the Southern type of vegetation in India and Australasia, it was well represented in Southern Africa and South America, where Lycopods occur in association with the Glossopteris flora. These Southern Lycopods were, however, all identical generically, and some of them specifically, with those found in Europe, North America, and Northern Asia. All the more important Palaeozoic genera mentioned above have now been recorded from either Southern Africa or South America. Family LEPIDODENDRE.E. Arborescent Lycopods, dichotomously branched. Stem clothed with leaves or with an armour of persistent, spirally arranged, prominent leaf-bases. Leaf-bases approximate, more rarely distant and separated by bands of bark. The leaf-base consists of a leaf- cushion and a leaf - scar. The leaf - cushions are fusiform or rhomboidal. The leaf -scar varies in shape, but is usually more or less transversely elongate, and bears three prints, the central one LEPIDODENDUOX. 155 being the scar of the leaf-trace bundle, while the two lateral together constitute the parichnos. Leaves (Lepidophyttum) simple, acicular, elongate, linear or linear-lanceolate, uninerved. Cones (Lepidostrobus) heterosporous, or possibly isosporous in some cases, borne on the terminations of the finer branches, or arranged spirally, or in two opposite rows on special fructiferous branches (Ualonia and Ulodendron in part). Sporangia large, elongate, borne singly on the upper surface of the spirally arranged sporophylls. Rhizomes (Stigmaria in part) cylindrical, diverging from the base of the ti'unkas four main branches, which repeatedly bifurcate. Rootlets cylindrical, tapering distally, arranged quincuncially on the rhizome. Rootlet-scars circular, with a circular depression between the outer and somewhat raised rim, and the central vascular print. The two more important genera of this family are Lepidodendron and Lepidopfdoios. These types are now known in great detail, both as regards their external morphology and internal anatomy. Genus LEPIDODENDRON, Sternberg, 1820. [Flora Vonvelt, Heft i, pp. 20 and 25.] Stem arborescent, dichotomously branched above. The persistent leaf -bases are vertically elongate, usually longer than broad, approximate, more rarely distant and separated by bands of bark. The leaf-scar, generally placed slightly above the centre of the leaf-cushion, is rhomboidal, and as a rule somewhat broader than long. The leaf-scar bears three prints, the central leaf-trace scar, and the two lateral scars of the parichnos. In some species a small scar, that of the ligule, occurs above the leaf-scar. The leaf-cushion below the leaf-scar in many species possesses a median ridge or keel, with or without transverse notches or grooves. The leaves (Lepidophyttum in part) are simple, acicular or elongately linear, uninerved. The cone (Lepidostrobus in part) consists of an axis bearing crowded sporophylls or bracts, usually spirally arranged, each sporophyll bearing a single, large, radially elongate sporangium on its upper surface. Rhizome (Stigmaria in part), see above. 156 LEPIDODEXDKOX. Anatomy of the stem. Pith, well marked, parenchymatous, or more rarely absent. Primary wood centripetal, composed of scalariform tracheides, with or without conjunctive parenchyma, either occupying the periphery of the pith, or, where a definite pith is absent, forming the central portion of the vascular cylinder. Secondary wood not always present, but, where it occurs, developed centrifugally on the outer margin of the primary wood, composed of scalariform tracheides. Cortex usually differentiated into two or more zones, the inner composed of thin-walled tissue, and the outer, consisting mainly of a periderm of thickened elements. Lepidodendron is one of the most characteristic genera of the Carboniferous rocks of Europe and elsewhere in the Northern Hemi- sphere. It appears first in the Devonian, and extends to the Permian. In addition to casts or impressions of the true external surface, others occur which are derived from more or less decorticated stems. Generic names, such as Aspidiaria, were at one time assigned to such fossils before their true nature had been fully realized. The name Knorria is one of these, which is still sometimes used to denote a particular type of Lepidodendroid cast, formed internally to the bark, which has remained as a hollow cylinder after the decay of the woody tissues. The features of such a cast are consequently the reverse of those exhibited by the inner surface of the ring of bark. 1. Lepidodendron Pedroanum (Carruthers). (PL I, Fig. 2.) 1869. Flemingites Pedroanus, Carruthers, Geol. Mag., vol. vi, p. 151, pi. v. 1895. Lepidodendron Pedroanum, Zeiller, Corapt. Rend., vol. cxxi, p. 962. L. Pedroanum, Zeiller, Bull. Soc. Geol. France, ser. in, vol. xxiii, p. 607, pi. viii, figs. 1-4. 1896. L. Pedroanum, Bodenbeuder, Zeitsclir. deutsch. geol. Gesell., vol. xlviii, table opposite p. 772. 1898. L. Pedroanum, Zeiller, Conrpt. Rend., vol. cxxvii, p. 246. Type. V. 230. Geol. Dept. British Museum (Nat. Hist.). Leaf-cushions and leaf-scars rather small, somewhat varied in length and breadth, in the type-specimen recalling in shape those of Lepidodendron Veltheimianum, Sternb. Leaf-cushions contiguous, lateral angles rounded. Leaf-scar placed at about one-third of the LEPIDODENDRON". 157 length of the cushion from the apex, rhomboidal, upper and lower angles rounded, prints indistinct. Field flat, "without keel or glands, and apparently without well - marked transverse ridges. Leaves short, acicular, persistent, uninerved. Carruthers, in his original description of these fossils, says "fruit a cone?," but he merely deduces this conclusion from the occurrence of numerous detached megaspores which he regards as belonging either to SigiUaria, or to Flemingites, in close association with the stems. It was on the characters of the supposed cone that he referred these specimens to the genus Flemingites rather than to Lepidodendron. There is no specimen of a cone among these South American fossils. The stems themselves, as Zeiller l has already shown, undoubtedly belong to a Lepidodendron, and are closely similar to European species of that genus. It is quite possible that the megaspores (see p. 175) associated with them may eventually prove to belong to L. Pedroanum, as Carruthers believed, but at present there is no positive evidence on this point. Carruthers states that the leaves of L. Pedroanum have a parallel nervation, and figures them as such. Zeiller 2 has, however, made a more detailed study of these organs, and finds that the surface is striated longitudinally by fine and close, parallel striae. In addition, there appear at first sight to be three parallel nerves, of which the two lateral are often better marked than the median. He shows that the central nerve is the single vascular bundle of the leaf, and that the two lateral ridges correspond to two stomati- ferous furrows, such as are found to occur running longitudinally through the leaf of a SigiUaria. Zeiller 2 has already pointed out that the fossils from Argentina, identified by Szajnocha 3 as L. Pedroanum, are not identical with Carruthers' species. Bodenbender 4 has recorded from Argentina another Lepidodendron, L. Sternhergii, Brong., which is a typical and common species in 1 Zeiller (95 2 ), p. 6Q7 ; (95 1 ), p. 962. 2 Zeiller (95 2 ), p. 60S. ;; Szajaocha (91), p. 207, pi. ii, figs. 2, 3. 1 Bodenbeuder (96), table opposite p. 770. 158 LEPIDODENDRON. the British Coal Measures, and elsewhere in Europe. This identi- fication may be very possibly correct, but until a description and illustrations of the specimens have been published, it would perhaps not be wise to lay any stress on the occurrence of this species in association with the Glossopteris flora. L. Pedroanum is known only from the Serra Partida, Candiota, and from Arroyo dos Ratos, both in the province of Rio Grande do Sul, Brazil. V. 230. PI. I, Fig. 2. Also figured by Carruthers (69), pi. v, figs. 10 and 11, and a further specimen on pi. v, fig. 8. A cast of a branch or stem, 65 cm. long and 18 mm. across, showing the contiguous, spirally arranged leaf-cushions. The cushions are 6 mm. long and 2'5 mm. broad. The upper and lower angles of the cushions are elongate, and bent in opposite directions. The lateral angles are very rounded. The leaf-scars are small and indistinct, the lower angles being rounded. The prints are not preserved. The field of the cushion below the scar is flat, without a keel or ridges, or if ridges are present they are very slight. The leaves are small, acicular. In pi. i, fig. 2, leaves are seen at the side of the specimen, apparently still attached to the stem. V. 230 a is a plaster cast of the above, showing the leaf-bases in relief. On the same specimen another branch occurs, densely clothed with acicular leaves. This was figured by Carruthers on pi. v, fig. 8. Serra Partida, Candiota, Brazil. Pres. by N. Plant, Esq., 1869. V. 2305. Figured by Carruthers (69), pi. v, fig. 7. The termination of a leafy branch, clothed with leaves. The leaf-bases are not seen. Serra Partida, Candiota, Brazil. Pres. by N. Plant, Esq., 1869. V. 230 £. Figured by Carruthers (69), pi. v, fig. 1. A fairly broad stem, with a few leaves still attached at the side. The features of the stem are obscure. Serra Partida, Candiota, Brazil. Pres. by N. Plant, Esq., 1869. V. 230^. ? Figured by Carruthers (69), pi. v, fig. 9. A portion of a badly preserved stem, showing the leaf-cushions, and in some instances a triangular pit representing the leaf-scar. Sena Partida, Candiota, Brazil. Pres. by JSf. Plant, Esq., 1869. LEPIDODENDRON. 159 V. 230c. A small portion of a branch, showing the leaf-cushions, for the most part badly preserved. A few leaves are seen at the sides of the specimen. Serra Partida, Candiota, Brazil. Pres. by N. Plant, Esq., 1869. V. 230/1 Part of a branch, covered with persistent leaf -bases. Serra Partida, Candiota, Brazil. Pres. by JV. Plant, Esq., 1869. V. 230,?. A portion of a stem or branch, the leaf-bases being- obscure. Serra Partida, Candiota, Brazil. Pres. by iV. Plant, Esq., 1869. V. 230A. A portion of a stem, and a detached megaspore. The drawing of a leaf given by Carruthers on pi. v, fig. 6 may perhaps have been made from this specimen. Serra Partida, Candiota, Brazil. Pres. by N. Plant, Esq., 1869. 2. Lepidodendron Derbyi (Renault). 1890. Lycnpodiopsis Derbyi, Renault, Compt. Rend., vol. ex, p. 809. L. Derbyi, Renault, Bull. Soc. Hist. nat. d'Autun, vol. iii, p. 109, pi. ix. 1898. Lepiclodmdron Derbyi, Zeiller, Compt. Rend., vol. exxvii, p. 245. Type. Museum d'histoire naturelle, Paris. Renault ' has described the external features of this species as follows : — " Le premier fragment d'ecorce mesure pres de 1 centimetre d'epaisseur ; d'une cote il presente de nombreuses cicatrices disposees tres regulierement, de l'autre une serie de cavites, correspondant au passage des faiseaux vasculaires qui se rendaient aux feuilles. Les mamelons ne sont pas contigus, mais separes par des sillons d'inegale largeur, places legerement en relief sur la surface de l'ecorce, e'est ee relief qui donne l'existence aux sillons dont la forme est deterniinee par celle des mamelons. Les mamelons sont disposes en quinconce sur deux lignes spirales croisees sous un angle de 90°. Quatre mamelons voisins forment une sorte de carre dont la diagonale transversale presque horizontale mesure 11 millimetres et la diagonale verticale 12 millimetres. Suivant sa plus grancle hauteur, chaque mamelon atteint 5 millimetres et i Renault (90-), pp. 110, 111. 1 60 LEPIDODENDRON. 4 rnilliinetres suivant sa largeur. Sa forme est un ovale irregulier a grand axe vertical. En haut cle chaque cote, le contour est presque rectiligne a angle superieur un peu arrondi, les cotes mesurent 3 miilinietres environ. Le contour inferieur est plus arrondi que le contour superieur. On ne distingue aucune trace de carene. Sur la surface du mamelon on remarque l'ernpreinte laissee par la base d'insertion de la feuille indiquee par une sorte de bourrelet de ce dernier qui, a la partie superieure, fait une saillie de pres d'un tiers cle millimetre. La cicatrice foliaire a sensiblement la meme forme que celle du mamelon, elle mesure 4 millimetres cle hauteur et 25 mm. de largeur. Une cicatricule ponctiforme placee un peu au-dessus du milieu, a 2 millimetres et demi a partir du bord inferieur de la cicatrice foliaire, indique le passage du faisceau vasculaire penetrant dans la feuille. Cette cicatricule est entouree d'un bourrelet saillant tres net d'un milli- metre de diametre. W\ le mamelon ni la cicatrice foliaire ne presentent la trace d'autres cicatricules. Sur les rameaux plus jeunes les mamelons sont plus petits, moins oblongs, plus rapproches, mesurant 4 millimetres environ de hauteur et 35 mm. de largeur." The anatomical structure of the stem is thus described by Renault : — l " Sur une coupe transversale on observe les details suivants : au centre, une moelle cylindrique, formee de cellules polyedriques, entouree d'un cylmdre ligneux. Ce dernier est constitue par des bands vasculaires disposees en lames rayonnantes simples ou soudees par leur extremite interue, de maniere a produire la configuration d'un U ou d'un V ; ces lames simples ou doubles sont separees par des rayons cellulaires (continuation du tissu fondamental qui forme la moelle) dont les cellules, au lieu d'etres polyedriques, se sont aplaties ou allongees dans le sens du rayon. Les tracheides qui constituent les b;mdes sont payees ou reticulees, elles sont en diminuant de calibre du centre a la peripherie ; les elements tracheens paraissent etre places a l'extremite externe des bandes. Le liber tres mal conserve forme une assise continue autour de l'ensemble du cylindre ligneux. Des vides places clans l'assise liberienne et dans l'ecorce correspondent sans aucun doute au 1 Renault (90 2 ), p. 116. LEPIDODENDRON. 161 passage des faisceaux vasculaires se rendant aux feuilles. Tl est a reniarquer que ces cordons foliaires prenuent naissance en face de l'intervalle qui separe deux bandes ligneuses voisines ; il est probable que ces deux bandes se soudent chacune a l'extrernite inferieure du cordon et sont en relation conductrice avec lui." The specimens described by the late Dr. Renault were obtained from Piracicaba, San Paulo, Brazil, where they were found in association with Psaionius and Cordaitean wood. They consist of two fragments showing the external or subepidermal features of the stem, and a specimen in which the internal anatomy is preserved. Guided chiefly by the structure of the stem, Renault referred these specimens to a new genus, Lyeopodiopsis, on the ground that the anatomy did not agree with that of any known Palaeozoic member of the Lycopodiales, but presented a closer comparison with the recent genus Lycopodium, from which, however, it appeared to ditfer in certain important respects. The existence of a well-marked pith, surrounded by a discontinuous vascular ring composed of centripetally developed xyletn, made up of single bundles or of bundles united internally in the form of a U or a V, are features unlike those of any Club-moss. Professor Zeiller 1 has, however, re-examined these specimens. He has pointed out that the surface showing the leaf-cushions is somewhat decorticated, and does not represent the true external surface, but that, if allowance be made for the change in appearance due to this fact, the resemblance to a Lepidodendroid stem, such as Lepidodendron selaginoides, is so marked that there can be little hesitation in regarding Renault's species as a true Lepidodendron He finds that the vascular cylinder in reality is perfectly continuous, the apparently broad parenchymatous rays between the bundles being largely composed of badly preserved woody elements, the walls of which appear to be much thinner than the tracheides owing to alteration which has taken place during preservation. He further notices that, in the best preserved specimens, there is evidence that the outer margin of the wood was crenulated, as in Lepidodendron LLarcourtii, and he concludes that 1 Zeiller (98 1 ). 162 LEriDODENDRON'. Renault's species is a true Lepidodendron, near to L. Harcowtti, and from the occurrence of numerous silicified leaves associated with the stem, closely resembling those of L. Pedroanam, he suggests that L. Derby i may perhaps be only a badly preserved stem of that species. L. Derby i is known only from Brazil. Not represented in the British Museum collection. 3. Lepidodendron (Knorria), sp. (from the Orange Biver Colony). Kidston 1 has figured some specimens from Western Australia, includiug Lepidodendron [Knorria) in association with Stigmaria, and Lycopodean leaves. These fossils did not, however, occur in association with the Glossopteris flora, and the precise age of the rocks is doubtful. V. 2900«, V. 2900&. Two sandstone casts presenting great similarity to the Knorria type of Lepidodendroid stem occurring in the European Coal-measures. One of these shows the impersistent longitudinal ridges fairly clearly. Farm Zwartkoppies (at 4,600 feet s.m.), Vredefort, Orange Biver Colony. Pres. by D. Draper, Esq., 1893. Obscure Lepidodendroid Fossils. V. 7596«, 7596/>. Two badly preserved sandstone casts, probably decorticated stems belonging to some Lycopod, such as Lepidodendron or Sigillaria. Sengwe Coalfield, Bhodesia. Pres. by A. J. C. Ifolyneux, Esq., 1901. V. 7594. A badly preserved, ribbed cast which I briefly described in 1903. 2 In appearance it rather recalls a Sigillaria, but it is impossible to determine it with confidence. The ribs are of unequal breadth, and there are no traces of leaf-scars. Sengwe Coalfield, Bhodesia. Pres. by A. J. C. Molymux, Esq., 1901. 1 Kidston (90), p. 102, pi. iv, figs. 4-8. 2 Arber (03), p. 290. LEPIDOrHXOIOS. 163 Genus LEPIDOPHLOIOS, Sternberg, 1826. [Flora Vorwelt, Heft iv, p. xiii.] Arborescent Lycopods, sterns dichotomously branched, branches clothed with imbricated, scale-like leaf-cushions, bearing leaf-scars at or near the summit. Leaf-cushions rhomboidal, transversely elongate, smooth or keeled, upright or reflexed. Leaf-scars oval or rhomboidal, transversely elongate, the median angles usually rounded. Vascular print usually larger than the two lateral prints of the parichnos, sometimes more or less triangular in form. Leaves linear, lanceolate, uninerved. Cones of the Lepidostrobus type, borne on special fructiferous branches (Hafonia), arranged spirally, ? stalked. Internal structure closely similar to Lepido- dendron, from which it differs by a few characters, such as the crenulated outer margin of the wood. Lepidophliiios is a comparatively small genus, occurring in the Lower and Upper Carboniferous rocks of Europe and North America, and with the Glossopteris flora in South America. Lepidophloios laricinus, Sternberg. 1820. Lepidodendron laricinum, Sternberg, Flora Vorwelt, Heft i, p. 23, pi. xi, figs. 2-4. 1826. Lepidqfloyos laricinum, Sternberg, ibid., Heft iv, p. xiii. 1854. Lepidodendron laricinum, Geinitz, Darstell. Flora Hain. -Ebersdorfer, p. 47, pi. xi, figs. 4, 5, 7. 1855. Lepidophloyos laricinum, Goldenberg (pars), Flora Sarsep. foss., Lief, i, p. 22, pi. iii, figs. 14, 14a. Knorria, sp., Goldenberg, ibid., pp. 17, 37, pi. ii, fig. 8b. 1857. ? Sigillaria Menardi, Goldenberg, ibid., Lief, ii, p. 24, pi. vii, fig. 1. 1862. Lepidophloios laricinum, Goldenberg, ibid., Lief, iii, p. 30, pi. xvi, figs. 2-6. 1866. L. acadianus, Dawson, Quart. Journ. Geol. Soc, vol. xxii, p. 163, pi. x, fig. 45. 1870. L. laricinus, Schunper, Traite, vol. ii, p. 51, pi. lix, fig. 4; pi. lxiv, figs. 4, 6, 8. L. intermedins, Scliimper, Traite, vol. ii, p. 51, pi. lxiv, figs. 4-8. 1871. L. laricinus, "Weiss, Foss. Flora jiingst. Steink. u. Kothl., p. 154, pi. xv, figs. 6, 7, 9. L. acuminatus, Weiss, ibid., p. 155, pi. xv, fig. 8. 1873. L. laricinus, Carruthers, Geol. Mag., vol. x, p. 150, pi. vii, fig. 3. 1874. L. acuminatus, Schimper, Traite, vol. iii, p. 537. 164 LEPrDOPHLOIOS. 1874. Lepidodendron laricinum, Feistmantel, Vers. bohm. Kohlenab., pt. 5, p. 191, pi. xxxiii, rigs. 1-4 ; pi. xxxiv, figs. 1-4. 1878. L. acadianus, Dawson, Acadian Geol., 3rd ed., p. 489, text-fig. 171 on p. 457. 1880. Zipidup/iloios laricinus, Zeiller, Veget. loss. terr. houill. France, p. 113, pi. clxxii, figs. 5, 6. 1882. L. laricinus, Renault (pars), Cour. botan. foss., vol. ii, p. 44, pi. ix, figs. 5, 7. 1884. ? L. dilatatus, Lesquereux, Coal Flora, vol. iii, p. 781, pi. cv, fig. 2. 1886. L. laricinus, Zeiller, Flore foss. bass, bouill. Valenciennes, p. 471, pi. lxxii, figs. 1, 2. 1885. L. acadianus, Dawson, Geol. Hist. Plants, p. 166, fig. 44 on p. 121. 1890. L. laricinus, Renault, Flore loss. terr. bouill. Commentry, p. 514, pi. Ixi, fig. 1. 1891. L. acadianus, Dawson, Geol. Nova Scotia (Acad. Geol., 4th Edit.), p. 456, fig. 171 on p. 457. 1893. L. laricinus, Kidston, Trans. Roy. Soc. Edinb., vol. xxxvii, pt. 3, p. 555, pi. i, figs. 4, ia ; pi. ii, figs. 8, 8a, 8b. 1895. L. laricinus, Zeiller, Compt. Rend., vol. cxxi, p. 962. X. laricinus, Zeiller, Bull. Soc. Geol. France, ser. in, vol. xxiii, p. 612, pi. ix, figs. 1-3, la, 2a. 1 898. L. laricinus, Potonie, Lehrb. Pflanzenpal., Lief, iii, p. 240, text-fig. 226. L. laricinus, Zeiller, Mem. Soc. Geol. France, vol. viii, Mem. No. 21, p. 74. 1900. L. laricinus, Scott, Stud. Foss. Bot., p. 160. L. laricinus, Zeiller, Elem. Paleobot, p. 186, text-fig. 128. 1902. L. laricinus, Kidston, Proc. Yorks. Geol. & Polyt. Soc, vol. xiv, p. 34 8, pi. lvi, fig. 2. 1904. L. laricinus, Zalessky, Mem. Com. Geol., n.s., vol. xiii, pp. 30, 99, pi. v, fig. 9 ; pi. vi, figs. 8, 10 ; pi. vii, figs. 1, 2 ; pi. viii, figs. 7, 9. " Leaf -cushions elongated, imbricate, scarcely keeled, directed downwards ; exposed portion of cushions rhomboidal or elongated transversely ; lateral and upper angles acute, lower angle generally rounded. Leaf-scar placed at the summit of the downward directed leaf-cushion, or only slightly below the summit, rhom- boidal, or transversely rhomboidal-elongate, lateral angles very sharp and prominent, upper and lower angles slightly rounded ; within the leaf-scar are three cicati'icules placed centrally, or slightly above or below the centre, the two lateral punctiform, the central punctiform or sub-triangular. When the leaf- scar is placed slightly below the apex of the deflexed cushion two lines run from its lateral angles, which meet the margin of the cushion a short distance below the leaf-scar. The leaf-cushion freque tly BOTHRODEXDREiE. 165 bears, immediately beneath, the leaf-scar, a small tubercle with a circular or sub-triangular depression. Fructification borne on Halonial branches." The above description is that given by Mr. Kidston * in his memoir on the genus Lepidophloios. L. laricinus is an Upper Carboniferous species, found com- paratively rarely in the British Coal Measures, but more abundantly in Continental beds of similar age. Zeiller 2 in 1895 showed that tbis plant also occurs associated with the Glossopteris flora in Brazil ; another instance of the identity of the Lycopodean element of the Permo- Carboniferous flora both in the Northern and Southern Hemispheres. It may be also noticed that Johnston 3 has figured a small and imperfect fragment from Tasmania as being possibly a Lepidophloios, which he also compares with Sigillaria Brardi; but for the present this record must remain doubtful. Distribution. — Permo- Carboniferous (Glossopteris flora) : — Brazil. Upper Carboniferous (Northern Type): — Britain, France, Germany, Austria, Russia, North America, and elsewhere. L. laricinus is represented in the British Museum collection by both British and foreign examples belonging to the Northern flora. 4 There are, however, no specimens from Brazil. Family BOTHEODENDEE^E. Arborescent Lycopods, dichotomously branched. Leaf-cushions absent or feebly developed. Leaf-scars distant, small, oval, bearing- three prints. Bark ornamented with delicate stria?, or smooth. Leaves small, linear - acuminate, or lanceolate. Cones similar in form to Lepidostrobus, borne terminally, or in two opposite longitudinal rows on the stem. The Bothrodendrese, consisting of a single genus Bothrodendron, is a small but interesting family of arborescent Lycopods of great antiquity, appearing first in the Upper Old Bed Sandstone of the 1 Kidston (93), p. 556. 2 Zeiller (95 1 ), p. 962 ; (95-), p. 617, pi. ix, figs. 1-3, la, 2a. ? - Johnston (88), p. Ill, pi. viii, fig. 2. * See Kidston (86), p. 169. 166 BOTHRODENDRON. south of Ireland, and in the Devonian rocks of Australia and the Arctic regions. It recalls certain of the Sigillariae, especially the Stibsigillariae of the Leiodermaria section, and also shows affinities in the type of cone to the Lepidodendreae. It may probably be regarded as occupying an intermediate position between these two families. Genus BOTHRODENDRON, Lindley & Button, 1833. [Foss. Flora, vol. ii, explan. to pis. lxxx, lxxxi.] Leaf-cushions only found on the smaller or younger branches, approximate, elongate, rhomboidal, slightly raised. Leaf-scars small or minute, transversely oval, usually distant, separated by broad bands of bark, more rarely on leaf- cushions. Leaf- scars bearing two lateral, and one central, minute, punctate prints. In some species a small print also occurs immediately above the leaf-scar. Bark ornamented by series of fine wrinkles or striae, straight or flexuous, arranged longitudinally, transversely, or in both directions, sometimes smooth. Leaves small, linear or lanceolate, uninerved. Cones similar in type to Lepidostrolus. Internal structure of the stem for the most part similar to Lepidodendron. The genus extends from the Upper Devonian to the Coal Measures, and possibly to the Permian. Bothrodendron Leslii, Seward. (Text-fig. 36.) 1903. Bothrodendron L/slii, Seward, Ann. S. African Mus., vol. iv, pt. 1, p. 87, pi. xi, figs. 1, la, lb, 4, 5, 6. Type. S. African Mus., Cape Town. Dichotomously branched, cylindrical stems without leaves, 6-2n mm. in breadth. Leaf-cushions absent. Leaf-scars small, transversely oval or subcircular, numerous, not very distant or almost crowded, prominent with a central depression, or occurring as a depression with a central umbo having a small pit in the centre. Bark smooth or somewhat wrinkled. The largest specimen figured by Mr. Seward is 29 cm. long, and is partially decorticated, especially in the lower portion. BOTHKODENDRON. 167 Mr. Seward 1 has concluded that while some of the characters of these specimens do not agree in detail with those of Bothro- dendron, this probably may be explained by taking into account the partially decorticated nature of the fossils, in which state there would naturally be no trace of the parichnos scars or of the ornamentation of the bark. ff4Mi I Ufh>§ II Wa loft mi }.■_ ON > M Fig. 36. — Bothrodendron Leslii, Seward. After Seward, a, nat. size ; b, c, slightly enlarged. The apparently dichotomous branching in these specimens is an argument in favour of referring them to the Lycopodiales rather than to the Gymnosperms. The same author has also concluded that the South African fossils bear so close a superficial resemblance to Bothrodendron kiltorken&e (Haughton), which occurs in the Devonian rocks of the South of Ireland, and of Bear Island in the Arctic Seward (03 1 ), pp. 90, 91. 168 SIGILLAB.IA.. regions, that one might be almost tempted to refer them to that species, but nevertheless he believes them to be distinct. The fragments from India, figured by Feistmantel 1 as Coniferous stems of ? Rhipidopsis, have a superficial resemblance to the South African specimens. Mr. Seward has, however, decided that the Indian fossils are too imperfect to afford any evidence as to their real nature. B-othrodendron Leslii is known only from Yereeniging in the Transvaal, South Africa. Not represented in the British Museum collection. Family SIGILLARLE. Arborescent Lycopods, stems unbranched or very rarely dichotomously branched. Stem longitudinally ribbed or without ribs, bark smooth, often ornamented. Leaf-scars hexagonal, lateral angles prominent, median angles usually rounded, often emarginate above. Prints of the leaf-scar three, the central vascular print punctate, the lateral prints of the parichnos crescentie or straight. Leaves linear-lanceolate or elongate linear, uninerved. Cones (Siyillariostrobus) either stalked and borne in irregular whorls on the stem, or sessile in two opposite vertical rows on a fructiferous branch ( Ulodendron in part). Sporangia heterosporous, large, elongate, borne on the upper surface of the sporophylls or bracts, which are spirally arranged or whorled. Rhizome (Stigmaria in part) similar to that of Lepidodendron. Internal structure similar to Lepidodendron, differing only in details. The family Sigillariae is represented by a single type of stem, Sigdlaria, a genus appearing first in the Lower Carboniferous and probably extending to the Bunter (Trias). Genus SIGILLARIA, Brongniart, 1822. [Class. Veget. foss., pp. 209, 222.] External surface of the stem longitudinally ribbed ( F.udgillarian type) or without ribs (Subsigittarian type). Leaf scars contiguous 1 Feistmantel (80), p. 124, pi. xlviiA, figs. 5-7. SIGILLARIA. 169 or distant, hexagonal, all the angles, except the lateral, being usually more or less rounded. Central vascular print small ; prints of the parichnos larger, crescentic or straight. A small print often occurs above the leaf-scar. Bark smooth or ornamented. Leaves linear -lanceolate or elongate linear, uninerved. Cone (Sigillariostrobus), rhizome {Stigmaria in part). Anatomy of the stem. Pith well marked, parenchymatous. Centripetal primary wood forming a continuous ring or composed of separate bundles, crenulated at the periphery, consisting of scalariform or reticulated tracheides, with small spiral tracheides forming the protoxyleni groups. Leaf- traces arising in the bays or indentations at the periphery of the primary wood. Secondary wood centrifugal, forming a continuous ring, as in Lepidodendron, consisting of scalariform tracheides pitted on both the tangential and radial surfaces. Cortex differentiated into two or more zones, the outer consisting of thick -walled periderm. The Sigittaria are divided into two main groups, the Eusigittarice, with ribbed stems, and the SubsigiUarice, without ribs. Further subdivisions bave been proposed, and while it is now known that these are without taxonomic importance, certain names are still sometimes used to denote a particular type of stem-cast. Thus the SubsigiUarice, the group to which the specimens described here belong, have been divided into two subgroups. The Clathrarice include stems in which the leaf-scars are placed on contiguous, rhomboidal cushions, usuallv somewhat elevated, the cushions being separated by well-marked oblique furrows. In the Leiodermarian type there are no leaf-cushions, the leaf-scars being distant and separated by bark, which is ornamented in various ways by stria?. Both these types may occur on the same specimen, as in the case of Sig Maria Brardi, which is found in South Africa in association with Glossopteris. Sigittaria first appears in the Lower Carboniferous, but is neither very abundant nor represented by many species in these rocks. The genus reaches its maximum in the Upper Carboniferous. In the Permian it is less frequent, and probably dies out at the end of that period with the exception of one species which is known to persist into Triassic times. 170 SIGILLARIA. Sigillaria Brardi, Brougniart. (PL VIII, Fig. 1.) 1822. Clathraria Brardii, Brongniart, Class. Veget. foss., p. 222, pi. xii, fig. 5. 1826. Favularia Brardi, Sternberg, Flora Vorwelt, Heft iv, p. xiv. 1828. Sigillaria Brardii, Brongniart, Prodr. Hist. Veget. foss., p. 65. 1836. S. Brardii, Brongniart, Hist. Veget. foss., p. 430, pi. clviii, fig. 4. S. rhomboidea, Brongniart, ibid., p. 425, pi. clvii, fig. 4. S. Menardi, Brongniart (pars), ibid., p. 430, pi. clviii, fig. 6. Lepidodendron Ottonis, Goppert, Syst. filic. foss., p. 462, pi. xlii, figs. 2, 3. 1839. Sigillaria elegans, Brongniart, Arch. Mus. hist, nat., Vol. i, p. 405, pis. xxv-xxviii. 1845. S. elegans, Corda, Flora Vorwelt, p. 24, pis. vii, viii. S. Brardii, Germar, Vers. Steink. Wettin, Lief, iii, p. 29, pi. xi, figs. 1, 2. 1848. S. spinulosa, Germar, ibid., Heft v, p. 58, pi. xxv, figs. 1, 2. 1855. S. elegans, Goldenberg, Flora Sarsep. foss., Lief, i, p. 26. 1857. S. Brardii, Goldenberg, ibid., Heft ii, p. 25, pi. vii, figs. 7, 8. S. rhomboidea, Goldenberg, ibid., p. 22, pi. vi, fig. 6. S. Menardi, Goldenberg (pars), ibid., p. 24, pi. vii, figs. 1, 2. S. elegans, Goldenberg, ibid., p. 27, pi. v, figs. 6-13. S. spinulosa, Goldenberg, ibid., p. 20, pi. x, fig. 5. 1860. Lepidodendron sexangulare, Eichwald, Lethoea Ross., vol. i, p. 114, pi. v, fig. S. 1864. Sigillaria denudata, Goppert, Foss. Flora Perm. Form., p. 200, pi. xxxiv, fig. 1. S. Brardi, Goppert, ibid., p. 201. 1870. S. Brardii, Schimper, Traite, vol. ii, p. 102, pi. Ixvii, figs. 10, 11. S. Menardi, Schimper, ibid., p. 103. S. spinulosa, Schimper, ibid., p. 102, pi. Ixvii, fig. 12. 1871. S. Brardii, Weiss, Foss. Flora jiingst. Steink. u. Bothl., Heft ii, p. 161, pi. xvi, fig. 1 ; pi. xvii, figs. 7-9. S. denudata, "Weiss, ibid., p. 159, pi. xvi, fig. 3. 1875. S. spinulosa, Renault & Grand'Eury, Mem. Acad. Sci. France, vol. xxii, No. 9, p. 1, pis. i-v ; pi. vi, figs. 33, 34. 1880. S. Brardi, Zeiller, Veget. foss. terr. houill. France, p. 135, pi. clxxiv, fig. I- S. rhomboidea, Zeiller, ibid., p. 137, pi. clxxiv, fig. 2. 1881. S. Brardii, Renault, Cours hot. foss., vol. i, p. 129, pi. xvii, fig. 1. S. elegans, Renault, ibid., p. 143, pi. xviii, figs. 1-10. S. spinulosa, Renault, ibid., p. 130, pi. xvii, fig. 2. S. Brardii, Feistmantel, Archiv. naturwiss. Landes. Bohmen, vol. iv, No. 6, Geol. Abth., p. 88, pi. v, figs. 1, la, 2. S. denudata, Feistmantel, ibid., p. 86, pi. v, figs. 3, Za. 1882. S. Brardi, Weiss, Aus Steinkohlenf., 2nd ed., p. 7, pi. iii, fig. 22. SIGILLARIA. 171 1882. Slgillaria denudata, Weiss, ibid., p. 7, pi. iii, fig. 23. 1886. S. Menardi, Weiss, Sitz.-ber. Gesell. naturforsch. Freunde, Berlin, for 1886, No. 2, p. 8, fig. 2 on p. 9. 8. cf. elegans, Weiss, ibid., p. 8, fig. 1 on p. 9. 1888. S. Brardi, Weiss, Zeitschr. deutsch. geol. Gesell., vol. xl, p. 569, fig. 4. S. Wettinensis, Weiss, ibid., p. 569, fig. 8. 1889. S. Brardi, Zeiller, Bull. Soc. Geol. France, ser. in, vol. xvii, p. 603, pi. xiv. S. Brardi, Weiss, Zeitschr. deutsch. geol. Gesell., vol. xli, p. 377. S. spinulusa, Weiss, ibid., p. 377. 1890. S. Brardi, Seward, Geol. Mag., dec. in, vol. vii, p. 213. S. Brardi, Renault, Flore foss. terr. houill. Comnientry, pt. ii, p. 539, pi. lxiii, fig. 1. S. Brardii, Grand'Eury, Bassin houill. du Gard, p. 250, pi. xi, fig. 1. 1892. S. Brardi, Zeiller, Flore foss. bass, houill. et perm, de Brive, p. 83, pi. xiv, fig. 1. 1893. S. mutans, Weiss & Sterzel, Abhand. k. preuss. geol. Landesanst., n.f., Heftii, p. 88. S. mutans, forma : denudata, Weiss & Sterzel, ibid., p. 92, pi. viii, fig. 39. denudata #, var. carbonica, Weiss & Sterzel, ibid., p. 94. rectestriata, Weiss & Sterzel, ibid., p. 94, pi. ix, fig. 42. subrectestriata ,, ,, p. 96, pi. ix, figs. 44, 45. epulvinata ,, ,, p. 97. subcurvistriata ,, ,, p. 98, pi. ix, fig. 43. undula ,, ,, p. 100, pi. ix, fig. 46. latareolata „ ,, p. 102, text-fig. 6. subspinulosa ,, ,, p 105, pi. xi, figs. 51, 52. spimdosa ,, ,, p. 106, pi. x, figs. 47, 50; pi. xi, fig. 50«. Wettinensis-spinulosa ,, ,, pp. 108, 127. Lardinensis- Brardi ,, ,, p. 110. pseudo-rhomboidea ,, ,, p. 112, pi. x, fig. 48. radicans ,, ,, p. 114, pi. x, fig. 49; pi. xi, figs. 49«, 496. laciniata ,, ,, p. 116, pi. xi, fig. 53. rhomboidea ,, ,, p. 117. subrhomboidea ,, ,, p. 118, pi. xii, fig. 54. mbleiodermaria „ „ p. 120, pi. xix, fig. 72. Wettinensis ,, ,, p. 122, pi. viii, fig. 55rt ; pi. xii, figs. 55, 56 ; pi. xiii, figs. 57, 58. cancellata ,, ,, p. 128, pi. xv, fig. 62. urceolata „ ,, p. 130, pi. xiv, fig. 59. Brardi ,, ,, p. 131. 1 I & SIGILLARIA. 1893. Sigillaria mutatis, forma Brardi, var. : 1ypic.it, Weiss & Sterzel, ibid., p. 133, pi. xv, fig. 60 ; pi. xx, fig. 82. Otfonis ,, ,, p. 138, pi. xvi, fig. 65. catenaria ,, ,, p. 139. sublcevis ,, ,, p. 142, pi. xvi, fig. 63. pnncticulata ,, ,, p. 143, pi. xvii, fig. 67. Otteivlorfensis ,, p. 143, pi. xx. fig. 77. Gtrmari-varians ,, ,, p 145, pi. xv, fig. 61 ; pi. xvii, fig. 66. subcancellata ,, ,, p. 154, pi. xix, fig. 73. S. mutans, forma Menardi, Weiss & Sterzel, ibid., p. 156. S. mutans, forma Menardi, var. : Cisti, Weiss & Sterzel, ibid., p. 157. sab-Brardi ,, ,, p. 158. Antunensis ,, ,, p. 159. varians ,, ,, p. 160, pi. xviii, figs. 68, 69, 71. subquadrata ,, ,, p. 163, pi. xix, fig. 74. Ahenziensis ,, ,, p. 164, pi. xx, fig. 78. minima ,, ,, p. 165, pi. xx, fig. 80. approximata ,, ,, p. 166, pi. xix, fig. 76. S. mutans, forma favulina, Weiss & Sterzel, ibid., p. 168, pi. xviii, fig. 70. ,, ,, Heeri, Weiss & Sterzel, ibid., p. 170, pi. xix, fig. 75. S. ambigua, Weiss & Sterzel, ibid., p. 172, pi. xx, fig. 79. S. Brardii, Potonie, Flora Rotbl. Thuriugen, p. 190, pi. xxvii, fig. 2. 1896. S. Brardii, Kidston, Proc. R. Soc. Edinb., vol. xiii, p. 233, pi. vii. 1897. S. Brardi, Seward, Quart. Journ. Geol. Soc, vol. liii, p. 326, pi. xxiii, fig. 2 ; pi. xxii, fig. 3 ; text-figs. 2a-p on p. 327, and text-fig. 3 on p. 329. 1900. 8. Brardi, Scott, Studies Foss. Bot., p. 193, text-fig. 75 on p. 190. 8. Menardi, Scott, ibid., p. 196, text-fig. 76 on p. 197. S. spinulosa, Scott, ibid., p. 200, text-figs. 77-79 on pp. 201, 205. S. Brardi, Zeiller, Elem. Paleobot., p. 193, text-fig. 136 on p. 193. 1901. S. Brardi, Kidston, Trans. Nat. Hist. Soc. Glasgow, n.s., vol. vi, pt. 1, p. 93, text-fig. 17 on p. 91. 1902. S. Brardi, Kidston, Proc. Yorks. Geol. Polyt. Soc, vol. xiv, pt. 3, p). lviii, fig. 2; pi. lix, fig. 1. Type. Museum d'histoire naturelle, Paris. Leaf-scar rhomboidal, as long as broad, but often broader than long; upper margin rounded, flattened, or notched, lower margin rounded, sides generally convex with sharp lateral angles. Prints three, placed slightly above the centre of the leaf-scar, the central print transversely linear, generally concave, the two lateral oblique, frequently lunate, embracing the central print ; above the leaf-scar there is frequently a small circular print. SIGILLAKIA. 173 A definite limit of the leaf-cushion often absent (Leioderman'a), occasionally incomplete, or distinct ( Clathraria). Leaf-cushion, when present, more or less elevated, generally rounded above and below, or subquadrate or spathulate with truncated base, and generally without any surface ornamentation. Leaf-scar usually on the upper portion of the cushion, central only in the young condition. In the Leiodermarian forms the outer surface of the bark, between the leaf-scars, is generally ornamented with a shagreen-like sculpturing of fine longitudinal ridges and delicate transverse stria?. The longitudinal lines are longer and coarser than the fine transverse striae ; under the leaf-scar the bark is frequently smooth — above the leaf-scar it is generally smooth. Immediately beneath the leaf-scar on certain Leiodermarian forms there sometimes occur one or two small circular scars. Cortex thin— decorticated stem longitudinally striated. The cone- scars form an irregular girdle round the stem. They are small, circular, oval, or subtriangular, and in the Clathrarian forms are placed on small cushions inserted between the leaf-scars, which, like the cone-scars, are usually more or less deformed by mutual pressure. In the Leiodermarian forms they possess no cushion, and are circular from the absence of all pressure. The internal anatomy of the stem has the characters described above under the genus. Mr. Kidston 1 published some years ago a full account of this species and its variations, from which the above description, with the exception of a few verbal alterations and additions, is taken. He has confirmed the conclusion that S. Menardi, Lrong., S. spinulosa, Germar, and S. denudata, Giipp., are merely varieties of S. firardi, Brong. The internal anatomy of at least two varieties of S. Brardi is known, but as the South African specimens described below are not petrified, the structure of the stem need not be dealt with in detail here. 2 1 Kidston (96) ; see also the papers by Zeiller (89), Weiss & Sterzel (93), Seward (90) mentioned above. • For a full account of the anatomy of S. Brardi see the memoirs by Brongniart (39), Renault & Grand'Eury (75), Corda ^45), & Scott (00) given hi the list of synonyms. 1 74 SIGTLf.ARIA. Distribution. — Upper Carboniferous and Permian (Northern Type) : — Europe : — Britain, France, Germany, Austria, Russia, and elsewhere ; ? North America : — Pennsylvania. Pernio- Carboniferous (Glossopteris Type) : — South Africa : — Transvaal. V. 3623. PI. VIII, Fig. 1. Also figured hy Seward (97 1 ), p. 326, pi. xxiii, fig. 2, and text-figs. 2a-e on -p. 327. A part of this fossil is figured on PI. VIII, Fig. 1, showing the spirally arranged leaf-scars, which in the upper portion are fairly well preserved. The entire specimeD, which is a sandstone cast, measures 18 cm. in length and 12 cm. broad. The leaf-scars are 9 mm. broad and 7 mm. in height. The stem had become decorticated to some extent before preservation took place, especially in the lower portion, consequently the characteristic features of a Sigillarian leaf- scar — the central vascular print and the two lateral crescentic prints of the parichnos — are not preserved. But on one or two of the scars in the upper portion the vascular print may be seen and also the parichnos, which is usually represented by a V-shaped projection. Enlarged drawings of several of these scars, accompanied by a full description, have been given by Mr. Seward. 1 The leaf-scars are closely set, and between them occur narrow sloping ridges of sandstone, which represent the grooves separating the leaf- scars in the living state. Vereeniging, Transvaal. Pres. by D. Draper, Esq., 1897. V. 3617. Figured by Seward (97 1 ), p. 330, text-fig. 3 on p. 329. A small piece of a stem, partially decorticated, showing a cast of the internal tissues (Syringodendron type). On one side, however, the leaf-scars are plainly seen, and they resemble those of the previous specimen. Vereeniging, Transvaal. Pres. by D. Draper, Esq., 1897. Sigillaria, sp., cf. Sigillaria Brardi. V. 3616. Figured by Seward (97 1 ), p. 330, pi. xxii, fig. 3. A faint sandstone impression of a badly preserved stem witli 1 Seward (97 1 ), text-fig. 2 on p. 327, and pp. 328, 329. LYCOrODEAN SPORES. 175 leaf-scars rather smaller than in the previous specimens. I agree with Mr. Seward in regarding this as probably a smaller form of S. Brardi. Yereeniging. Pres. hj D. Draper, Esq., 1897. V. 3618. Figured by Seward (97 1 ), p. 330, pi. xxii, fig. 4a; pi. xxiv, fig. 3, and text-fig. 2/ on p. 327. A fairly well-preserved portion of a broad stem occurring in association with Glossopleris and other genera. Some of the best preserved of the leaf- scars have been studied in detail by Mr. Seward, who has discussed the nature of this specimen, and has expressed some uncertainty as to its affinities. Vereeniging. Pres. by D. Draper, Esq., 1897. V. 8324. Described by Seward (98 2 ), pp. 92, 93. A sandstone cast of a fairly broad stem showing the leaf -scars, which for the most part are badly preserved. Vereeniging. Pres. by Dr. F. B. Hatch, 1898. [V. 233. Sigillaria (Syringodendron), sp. A decorticated, ribbed Sig Maria bearing the record " Coal shale from the Stormberg with impression of Calamites." There is every reason to believe that this specimen was not derived from South Africa. See Historical Sketch under heading Cape Colony.] LYCOPODEAjS" SPORES. Carruthers, 1 in 1869, figured a number of megaspores, some of which were associated with the type-specimens of Lepidodi-ndron Pedronnum. These bodies he speaks of as sporangia, but there can hardly be any doubt that they are megaspores. They are of fairly large size, about 2 mm. in diameter, and are flattened discs, yellow in colour. Many of the specimens are crushed arid broken, but some of them appear to have a narrow rim surrounding a somewhat raised central disc, and, on one surface, a triradiate ridge. Zeiller 2 has also noticed and described these bodies, pointing out that they very closely resemble those of Sigillariostrobus Tieghemi, 1 Ciirruthers (69), pi. v, tigs. 2-5. - Zeiller (95 1 ), p. 962, aud (95-), p. 609, text-fig. 2. 176 LYCOPODEAN SPORES. and since the megaspores of cones of Lepidodendron are not known to attain to such a size, he does not agree with Carrnthers in regarding them as probably referable to Lepidodendron Pedroanum. Zeiller has also described other spores from Brazil, some of which he figures as microspores, 1 and others as possibly pollen- grains. 2 Lycopodean Megaspores from Brazil. V. 231. A piece of brownish shale showing a few detached megaspores. Candiota, Rio Grande do Sul. Pres. by JSf. Plant, Esq., 1869. V. 231rt. Figured by Carruthers (69), pi. v, fig. 4. A detached megaspore. Candiota, Rio Grande do Sul. Pres. by N. Plant, Esq., 1869. V. 231c ? Figured by Carruthers (69), pi. v, fig. 2. A detached megaspore, probably one of those figured by Carruthers. Candiota, Rio Grande do Sul. Pres. by JSf. Plant, Esq., 1869. V. 231^, V. 231 d-i. Several specimens showing megaspores similar to those figured by Carruthers. Candiota, Rio Grande do Sul. Pres. by iV. Plant, Esq., 1869. Spores of Unknown Affinity. [Tasmnnites punctatus, Newton.] The nature of the minute discs occurring abundantly in the Tasmanite and Australian White Coal of the Mersey River Beds in Tasmania (Permo-Carboniferous), to which Newton 3 gave the name Tamianites punctatus, has been much discussed by several authors. 4 They have been regarded as sporangia, algae, spores, and scales, but there is no doubt that they are of the nature of spores, although of what particular type of plant there is no evidence to show. They can, however, hardly be of Lycopodean 1 Zeiller (9.5 2 ), p. 611, text-figs. 3-5. 2 Zeiller (95 2 ), p. 612, text-figs. 6, 7. :i Newton (75), p. 337, pi. x. 4 For a lull bibliography on this subject see Etheridge (78), pp. 199, 200 ; see also Feistmantel (90), p. 143. COKDAITALES. 177 origin, since Lycopods are unknown at present from Australasia in association with the Glossopteris flora. The discs vary in diameter from '3-'5 mm. Newton describes them as "more or less circular bodies somewhat thickened towards the circumference, many of them having their surfaces raised into irregular folds." The walls of the sacs are dotted, the dots being "minute lines (tubes?) passing from the outer to the inner surface." The spores are brownish in colour, and usually compressed. The examination of the two specimens in the collection showing these spores has not thrown any fresh light on their structure or nature. It appears to me to be hardly necessary to retain the name Tasmanites punctatus. "V. 4010. A piece of sandy shale showing a large number of minute, rounded spores. Mersey River, Tasmania. Pres. by G. Sweet, Esq., 1900. V. 3778. A similar specimen showing numerous yellowish spores. Mersey River, Tasmania. Pres. by T. Stephens, Esq., 1896. Group GYMNOSPERMEiE. Trees or shrubs with horaoxylous wood (except Gnetacece). Flowers always unisexual, and as a rule without perianth (except Gnetaeeae). Ovules and seeds naked, not enclosed in carpels. Seed endospermous, endosperm being formed before fertilisation. Anemophilous plants. There are five great classes of Gymnosperms, four of -which, viz., the Cycadophyta, Ginkgoales, Gnetaceae, and Coniferales, are repre- sented in the flora of the present day, and one, the Cordaitales, has been long extinct. All these classes, except Gnetales, contribute to the Permo-Carboniferous floras both of the Northern and Southern Hemispheres. Class CORDAITALES. An extinct race of seed-bearing, arborescent plants, with large, simple, entire leaves, traversed by simple, parallel or sub parallel nerves, with occasional dichotomy. The female fructification is a seed of the Gymnosperrnous type. 178 NOEGGERA.THIOPSIS. The best known genus of this group, Cordaites, is a characteristic member of the Pernio -Carboniferous flora of the Northern Hemi- sphere. In the Glossopteris flora, the Cordaitales were probably represented by JVbeggerathiopsis, but our knowledge of the habit, structure, and fructification of this plant is still very imperfect, and consequently this conclusion is to some extent provisional. Genus NOEGGERATHIOPSIS, Feistmantel, 1879. [Flora Gondw. Syst., vol. iii, pt. 1, p. 23.] Leaves simple, spathulate, lanceolate or linear-lanceolate, or ovate. No midrib. Nervation sub-parallel, several parallel nerves entering the base of the leaf, dividing by dichotomy at a very acute angle, and without anastomosis. Considerable difference of opinion has existed both with regard to the affinities of this genus, and also in respect to the identity of some of its members. The first Indian specimens were described by Buubury in 1 8 '"> 1 , who referred them provisionally to the Gyrnnospermous genus JVbeggerathia. This generic name was also at one time adopted by Feistmantel, but in 1879 he proposed a new genus Noeggerathiopsis for their reception, on the grounds that they were closely related to the Cycads. The same author, in 1881, pointed out the close similarity presented by the leaves described as Rhiptozamites by Schmalhausen l with the Indian and Australian species of Noeggerathiopsis, and the probable affinity of both these genera to the Mesozoic Cycads of the family Zamire. In more recent times, Zeiller, Seward, Solms-Laubach, 2 and others have regarded this genus as in all probability a member of the Cordaitales, closely allied to Cordaites. Mr. Seward, 3 in fact, has recently expressed the view that " had the leaves referred to this genus been found in European Palaeozoic rocks there can be little doubt that they would have been described under the name Cordaites." Zeiller 4 has made a careful study of the leaves described by Schmalhausen as Rhiptozamites. He regards them as belonging 1 Schmalliauseii (79). 2 Solms-Laubach (91), p. 110. 3 Seward (03 1 ), p. 9.3. 4 Zeiller (9C-), p. 475 ; (02>), p. 32 ; (02 2 ). NOEGGEKATniOrSrS. 179 to a true Cordaites, and as distinct from the members of the Glossopteris flora described here, which, however, they closely resemble in several respects. Kurtz ' has recently drawn attention to the more spreading type of nervation in the leaves occurring in the rocks of Gondwanaland. Zeiller 2 has pointed out that the constant occurrence of Cordaitean seeds of the genera Cardiocarpus or Cordaiearpus in association with N, JJislopi is an additional argument in favour of referring Noeggeraihiopsis to the Cordaitales. On the whole, there seems to be a consensus of opinion at the present time that Noeggeraihiopsis is probably best regarded as distinct from Cordaites, although closely allied to that genus. It is quite possible that the future may somewhat modify this view, as our knowledge of the genus is still very imperfect, little or nothing being known as to the habit of the plant, and there is an entire absence of any evidence as to the fructification. Distribution. — Pernio- Carboniferous (Glossopteris flora) : — India, in the Talchir and Damuda divisions ; New South "Wales, in the " Lower Coal Measures " and Newcastle Series; Tasmania, Cape Colony, Transvaal, Argentina. Triasso-Rhatic : — ? Tonquin, ?China. 1. Noeggerathiopsis Hislopi (Bunbmy). [PL VI, Figs. 2, 3 ; PL VIII, Fig. 2 ; Text-fig. 37.] 1847. Zeugophyllites elongatus, McCoy, Ann. & Mag. Nat. Hist., vol. xx, p. 152. 18-19. Noeggerathia spatulata, Dana, in Wilkes' U.S. Explor. Exped., vol. x, p. 715, pi. xii, fig. 9. N. media, Dana, ibid., p. 715, pi. xii, fig. 10. 2V. elongata, Dana, ibid., p. 715. 1S50. Zeugophyllites elongatus, Unger, Gen. et Spec. Plant, foss., p. 332. 18G1. Noeggerathia? (Cgclopteris?) Hislopii, Bunbury, Quart. Jo urn. Geol. Soc, vol. xvii, p. 334, pi. x, fig. 5. 1872. N. Goepperti, Schimper, Traite, vol. ii, p. 130. 1878. X. spathulata, Feistmantel, Palseontogr., Suppl. iii, p. 97. 3*. media, Feistmantel, ibid., p. 97. ? 2V. prisca, Feistmantel, ibid., p. 158, pi. viii (xxvi), fig. 3. 1 Kurtz (03), p. 25. - Zeiller (02 1 ), p. 32. 180 XOEGGERATHIOPSIS. 1879. Noeggerathiopsia Hislopi, Feistmantel, Flora Gondw. Syst., vol. iii, pt. 1, p. 23, pi. xix, figs. 1-6 ; pi. xx, figs. 1, la. 2V. Eislopi, var. subrhomboidalis, Feistmantel, ibid., p. 24, pi. xx, fig. 2. 1S81. iV. Hislopi, Feistmantel, ibid., vol. iii, pts. 2, 3, p. 118, pi. xIva, figs. 1-11 ; pi. xlviA, figs. 3, 4. N. Hislopi, Feistmantel, ibid., vol. iii, pt. 1, Suppl., p. 58, pi. xxviii, figs. 1-4, 6-7; pi. xxix, figs. 1-4; pi. xxx, figs. 5-9. 1882. N. Hislopi, Feistmantel, ibid., vol. iv, pt. 1, p. 41, pi. ix, figs. 1-3 ; pi. xiii, figs. 2-4 ; pi. xiv, figs. 1-3, 6, 9 ; pi. xv, fig. 4# ; pi. xvii, fig. 4 ; pi. xviii, fig. 1 ; pi. xx, fig. 10 ; pi. xxi, figs. 6, 8, 10. 1883. N. media, Tenison- Woods, Proc. Linn. Soc. New South. Wales, vol. viii, p. 154. N. spathulata, Tenison-Woods, ibid., p. 153. JV. elongata, Tenison-Woods, ibid., p. 154. ? N. prisca, Teuison-Woods, ibid., p. 154. 1886. N. Hislopi, Feistmantel, Flora Gondw. Syst., vol. iv, pt. 2, p. 40, pi. xiiA, fig. 5a; pi. xiiiA, fig. 5. iV. spathulata, Johnston, Papers and Proc. E. Soc. Tasmania for 1885, p. 386. N. media, Johnston, ibid., p. 386. N. elongata, Johnston, ibid., p. 386. ? N. prisca, Johnston, ibid., p. 386. 1888. N. spatulata, Johnston, Geol. Tasmania, p. Ill, pi. x, figs. 2, 3. N. media, Johnston, ibid., p. Ill, pi. ix, figs. 2-4. ? N. prisca, Johnston, ibid., p. 111. 1S89. N. Hislopi, Feistmantel, Abhaud. bohm. Gesell. Wiss. Prag, ser. vn, vol. iii, p. 38, pi. iv, fig. 1. 1890. N. Eislopi, Feistmantel, Mem. Geol. Surv. New South Wales, Pal., No. 3, p. 153. N. spathulata, Feistmantel, ibid., p. 154. N. media, Feistmantel, ibid., p. 154, pi. xxi, figs. 3-5. ? N. prisca, Feistmantel, ibid., p. 155, pi. xiii, fig. 2. 1893. N. elongata, Etheridge, jun., Pec. Geol. Surv. New South Wales, vol. iii, pt. 3, p. 75. N. Hislopi, Oldham, Man. Geol. India, p. 162, pi. opp. p. 158. 1894. N. Hislopi, Kurtz, Eev. Mus. La Plata, vol. vi, p. 131, pi. iii, figs. 3, 4 ; pi. iv, fig. 1. N. Hislopi, var. subrhomboidalis, Kurtz, ibid., p. 132, pi. iv, fig. 2. 1895. N. Hislopi, Bodenbender, ibid., vol. vii, table opposite p. 148. 1896. N. Hislopi, Bodenbender, Zcitschr. deutsch. geol. Gesell., vol. xlviii, table opposite p. 772. X. Hislopi, Zeiller, Bull. Soc. Geol. France, ser. in, vol. xxiv, p. 372, text-figs. 16, 17; pi. xviii, figs. 6-9. 1897. N. Eislopi, Seward, Quart. Journ. Geol. Soc, vol. liii, p. 322, pi. xxi, fig. 4b. 1898. K Hislopi, Seward, ibid., vol. liv, p. 93. 1900. ? iV. Hislopi, Potonie, in Deutsch- Ost-Afrika, vol. vii, p. 501. NOEGGERATHIOrSIS. 181 1902. Noeggerathiopsis Gocpperti, Arber, Quart. Journ. Geol. Soc, vol. lviii, p. 17, pi. i, figs. 1, 2. ? N. Hislopi, Zeiller, Flore Foss. Gites Charb. Tonkin, p. 149, pi. xl, figs. 1-6, ?7-9. 1903. A'. Hislopi, Seward, Ann. S. African Mus., vol. iv, pt. 1, p. 9G, pi. x, fig. 5 ; pi. xiii, figs. 2-4. A 7 ". Hislopi, Kurtz, Quart. Journ. Geol. Soc, vol. lix, p. 25. Type. No. It. 10,370, Mus. Geol. Soc. London. Leaves considerably varied both in shape and size, broadly linear, oval-linear, sub-linear, spathulate, or obovate. Apex obtusely rounded. Lamina contracted at the base. Nerves numerous, somewhat strong, fairly distant, sub-parallel, branching dichotomously at a very acute angle, especially in the basal portion and occasionally in other parts of the leaf. Fine and close longi- tudinal striations occur between the nerves on the upper surface of the leaf. The epidermal cells have rectilinear walls, and fine punctures, indicating stomata, occur on the lower surface. The size of the leaf varies greatly, from 8-23 cm. or more in length, and from 1-5 cm. across at the broadest part. The nerves are strong (•2-"3 mm. broad), and are distant about '25-*75 mm. In some cases, as Feistmantel pointed out, the leaf is slightly unsymmetrical. Feistrnantel l has described a variety which he named N. Hislopi, var. subrhomboidalis (Text-fig. 37b), in which the leaf is rather wider than usual in the upper portion and somewhat narrower towards the base ; but in view of the great variety in shape and size presented by these leaves, this form seems to me to be hardly worthy of distinction. The leaf named by Kurtz, 2 Splienozamites multinervis, appears to be probably a young or small leaf of a Noeggerathiopsis (cf. Feistmantel (79 1 ), pi. xix, figs. 3-5), and I do not see any good evidence for assuming that it represents a Cycadean frond. A leaf from the older Glossopteris-he&ring rocks of New South Wales, at Greta, described by Feistmantel 3 as Noeggerathiopsis prisca, hardly appears to me to be distinct from N. Hislopi. In nervation it may be compared with the specimen figured here on PI. VIII, Fig. 2, from India. 1 Feistmantel (79 1 ), p. 24, pi. xx, fig. 2. - Kurtz (94 2 ), p. 130, pi. iii, fig. 2. 3 Feistmantel (90), p. 155, pi. xiii, fig. 2. 182 NOEGGERATmor.SIS. A good deal of confusion has arisen with regard to some Australian leaves, similar to the Indian species Noeggerathiop&is A B Fig. 37. — Xoeygcrathiojisis Hislopi (liunb.). After ETeistmantel. Nat. size. NOEGGERATniOPSIS. 183 Hislopi, which were first described by McCoy in 1847. McCoy 1 referred them to a Triassic species instituted by Morris as Zeugo- pihyllites ehngatus, which has been transferred recently by Mr. Seward 2 to the genus Pheenicopsis of Heer. It was not until 1892 that Mr. Etheridge, jun., pointed out that McCoy's determination was incorrect. McCoy's specimens were redescribed 3 in 1902, when I referred the leaves in question to JVbeggeratMopsis Goepperti, a species described by Schmalhausen 4 from the Permian rocks of Russia. A full account of the confusion with regard to McCoy's plant will be found in that paper. At that time, as I stated, I was unable to arrive at a definite conclusion with regard to the identity of N~. Goepperti with N. Hislopi, despite their great similarity in habit and detail. I am now, however, of opinion that the Australia fronds should be referred to N. Hislopi, since such differences as exist seem to be insufficient to warrant separation. Zeiller has shown that Schmalhausen's species must now be trans- ferred to Cordaites, and that Noeggerathiopsis is probably not identical with that genus. It appears to me that all the three species instituted by Dana 5 in 1849 are identical with N. Hislopi. Feistmantel 6 has already united two of the leaves in question. The plant recently figured by Shirley 7 as Noeggerathia (?), sp., is of very uncertain affinity. The leaf doubtfully ascribed to the genus Noeggeratlnopsis by Seward 8 is rather like some of the Indian and Australian species described by Feistmantel. The same author 9 has recently expressed doubt whether the Tonquin specimens referred to this species by Zeiller may not be identical with Pheenicopsis elongata (Morris). Distribution. — Per mo- Carboniferous (Glossopteris flora) : — India, in the Talchir and Damuda divisions ; New South Wales, in the "Lower Coal Measures" and Newcastle Series; Tasmania, Cape Colony, Transvaal, Argentina. Triasso - Rhcetic : — ? Tonquin, ? China. 1 McCoy (47), p. 152. - Seward (03 1 ), p. 67. 3 Arber (02 1 ), p. 17. 4 Schmalhausen (79). 5 Dana (49), p. 715, and pi. xii, figs. 9, 10. Feistmantel (90), p. 154. ' Shirley (02), p. 8, pi. iv. 8 Seward (04), p. 100, pi. iv, fig. 8. ° Seward (03 1 ), p. 96. 184 NOEGGEBATHIOPSIS. Specimens of Noeggerathiopsis Hislopi from India. V. 7233. PI. VIII, Fig. 2. This specimen shows two leaves, of which the one figured on PI. VIII, Fig. 2, is nearly perfect. It is obovate, contracted at the base, and broadly rounded at the apex. It measures 4'6crn. long, and 3'2 cm. broad at its greatest width. At the base the nerves are few in number, but they dichotomise above. The veins are rather coarse. The other specimen is a rather imperfect basal portion, showing the nervation clearly, Silewada, near Nagpur. V. 7131. A median portion of a leaf, without base or apex, measuring 5-3 cm. long and 2'3 cm. broad. The parallel nerves are well marked, distant, slightly spreading from the base, and forking in the distal portion of the leaf. Kamthi. Hunter Coll. V. 4207. Among a number of fronds on this specimen, most of which belong to the genus Glossoptcris, there are one or two fragments of narrow leaves with sub-parallel veins, and a much broader leaf, in which the veins are strong and forked. ? India. Specimens of Noeggerathiopsis Hislopi from Tasmania. V. 3776^. PI. VI, Fig. 2. A portion of a broadly oval or spathulate leaf, 12*5 cm. long, and more than 4 - o cm. across. The nerves are rather close, for the most part sub-parallel, but rather spreading and arched towards the margin, and dichotomising occasionally, especially near the base of the leaf. Mersey River. Pres. bg T. Stephens, Esq., 1898. V. 3776y. PI. VI, Fig. 3. A basal portion of a strap-shaped leaf, 8 cm. long. At the narrow base there are only about six or eight veins, which, however, dichotomise at a higher level. This leaf is apparently indistinguishable, except in size and shape, from the much broader leaves, such as that figured on PI. VI, Fig. 2, or from the other fragmentary portions also occui-ring on this specimen. Several fronds of Glossopteris, and a small seed are also associated. Mersey River. Pres. by T. Stephens, Esq., 1898. NOEGGERATHIOrslS. 185 V. 3776 v. A strap-shaped leaf, almost linear in the upper portion, but contracting slightly at the base. It measures 6 - 7 cm. long, and 14 cm. across at its greatest breadth. It is associated with fronds of Glossopteris. Mersey River. Pres. by T. Stephens, Esq., 1898. V. 3776 ah. A felted mass of leaves, among which are a few fragments belonging to iV. Hislopi, showing the sub -parallel nervation clearly. Mersey River. Pres. by T. Stephens, Esq., 1898. V. 3776aa. A small portion of an oblong leaf with sub-parallel veins, associated with Glossopteris and Cardiocarpiis. Mersey River. Pres. by T. Stephens, Esq., 1898. V. 3776 x. Several fronds of Glossopteris, with which occurs a fragment of an oblong leaf, in which the veins are fairly distant from one another. Mersey River. Pros, by T. Stephens, Esq., 1898. V. 3776 ad. A felted mass of fragments of leaves belonging to this species, in some of which the nervation is fairly clear. Mersey River. Pres. by T. Stephens, Esq., 1898. V. 3776 ae. This specimen shows portions of two leaves, associated with Glossopteris Broicniana and Glossopteris ampla. One of these measures 8 cm. in length, and has fairly distant nerves, but the nervation is not very well seen. Mersey River. Pres. by T. Stephens, Esq., 1898. Other specimens :— V. 3776 z, V. 3776 ac, V. 3776 af, V. 3776 ay, V. 3776 ah (Mersey River ; pres. by T. Stephens, Esq., 1898). Noeyyerathiopsis, sp., 40,942 (Newcastle Coalfield, New South Wales; pres. by Sir E. Home, 1857). Specimens of Noeyyerathiopsis Hislopi from South Africa. ? V. 3625. Figured by Seward (97 1 ), p. 322, pi. xxi, fig. 6. A small fragment 2 '5 cm. long, with rather indistinct, almost parallel veins. The surface is strongly convex. Boschmans Fontein, Middelburg, Transvaal. Pres. by I). Draper, Esq., 1897. V. 3813. Figured by Seward (97 1 ), p. 322, pi. xxi, fig. 4b. A basal portion of a leaf, about 5 cm. long, associated with a frond of Glossopteris anyustifolia. Casey's Township, Transvaal. Pres. by D. Draper, Esq., 1897. 186 NOEGGERATHIOPSIS. V. 8323. Described by Seward (92 2 ), p. 93. A piece of white sandstone, with a basal portion of a long leaf, 20 cm. in length, and nearly 5 cm. across. The nervation is not very clear, but the nerves do not appear to anastomose, and are nearly parallel. The leaf is markedly constricted towards the base. Fragments of fronds of Glossopteris are also associated. Vereeniging, Transvaal. Pies, by Br. F. II. Hatch, 1898. 2. Noeggerathiopsis Whittiana (Feistmantel). 1879. Euryphyllum Whittianum, Feistmantel, Flora Gonchv. Syst., vol. iii, pt. 1, p. 26, pi. xxi, figs. 1, Iff. 1881. E. Whittianum, Feistmantel, ibid., vol. iii, pt. 3, p. 120. 1894. Noeggerathiopsis Hmhpi, var. curyphylloides, Kurtz, Rev. Mus. La Plata, vol. vi, p. 132, pi. iv, fig. 3. 1896. Euryphyllum Whittianum (?), Boden bender, Zeitscbr. deutscb. geol. Gesell., vol. xlviii, table opposite p. 772. Type. ~No. 5036, Mus. Geol. Surv. India, Calcutta. Leaves ovate - spathulate, rounded at the apex, gradually contracted at the base, unsymmetrical. Nerves strong, radiating from the base at a very acute angle, with frequent dichotomy, those of the median portion of the leaf sub-parallel, those in the lateral portions arched towards the margin. The figures given by Feistmantel of these leaves from the Karharbari Group in India recall some of the broader forms of Noeggerathiopsis Rislopi (cf. with Feistmantel (79 1 ), pi. xx, fig. 2) in shape and general nervation. They apparently differ only in the larger size, broader outline, and in the nervation being less nearly parallel, more arched, and more subdivided. These differ- ences in the nervation, however, may be very well accounted for by the larger size of the leaf. Feistmantel placed them in a new genus Euryphyllum on the grounds that they were borne spirally on a broad axis, which he figures. He, however, admits (explanation to his figure (79 1 ), pi. xxi, fig. 1) that there is no real evidence of continuity between the leaves and the axis. It seems to me safer for the present to conclude that these leaves belong to the genus Noeggerathiopsis, and to compare them with N. Ilislopi, which they so closely resemble in habit. NOEGGERATHIOrSIS. 187 Zeiller 1 has recently pointed out that this species is also, in all probability, a member of the group Cordaitales. The leaf from Argentina, named by Kurtz 2 Noeggerathiopsis Hislopi, var. eurgphylloides, appears to be a basal portion of a frond of this type. iV. Whittiana is known from the Karharbari Series in India, and from Argentina. Not represented in the British Museum collection. 3. Noeggerathiopsis (?) Stoliczkana (Feistmantel). (Text-fig. 38.) 1876. Glossozamites Stoliczkanus, Feistmantel, Rec. Geol. Surv. India, vol. ix, pt. 4, p. 142. 1879. G. Stoliczkanus , Feistmantel, Flora Gondw. Syst., vol. iii, pt. 1, p. 19, pi. xx, figs. 4, 5. 18S1. G. Stoliczkanus, Feistmantel, ibid., vol. iii, pt. 3, p. 117. Type. Nos. 5034-5, Mus. Geol. Surv. India, Calcutta. Leaf simple (?), oblong, slightly unsymmetrical. Apex broadly rounded. Base truncated, broad, lateral angles slightly rounded. Nerves somewhat radiating in the basal portion, becoming very erect, sub-parallel, and dichotomising once or twice at a very acute angle. Feistmantel has figured as Glossozamites Stoliczkanus a single specimen of a leaf from the Karharbari Beds in India, which he regarded as a pinnule of a Cycadean frond, and not as a member of the genus Noeggerathiopsis . This conclusion was founded on the nature of the broad, truncated base, which he interpreted as evidence of insertion on a rachis in a similar manner to the pinnules of the Cycadean fronds of Mesozoic age included in Schimper's genus Glossozamites. There is no reliable evidence, however, as to the mode of attachment, and it seems to me im- possible to overlook the great similarity which this leaf presents to Noeggerathiopsis Hislopi, both in its nervation and the general habit other than the characters of the basal portion. In the case of N. Hislopi also the mode of attachment is unknown. Thus, on the present evidence, it would seem to me safer provisionally to 1 Zeiller (02), p. 38. • Kurtz (94°-), p. 132, pi. iv, fig. 3. NOEGGERATHIOPSIS. refer Feistinantel's specimen to that genus, on the grounds that it possesses several characters in common with iV. Hislopi. It is possible that the discovery of further examples may prove that Feistmantel was right in his conclusion, but at present there seems to be little or no trustworthy evidence as to the Cycadean nature of this species. Not represented in the British Museum collection. Fig. 38. — Noeggerathiopsis (?) Stoliczkana (Feist.). After Fcistmantel. Nat. size. 4. [Noeggerathiopsis ?] lacerata, Feistmantel. (Text-fig. 39.) 1882. Noeggerathiopsis lacerata, Feistmantel, Flora Gondw. Syst., vol. iv, pt. 1, p. 42, pi. xv, figs. 1-3, 4a; pi. xvii, figs. 2, 3. 1902. N. (?) lacerata, Zeiller, Pal. Indica, N.S., vol. ii, p. 32, pi. vii, figs. 2, 3. Type. Nos. 5455-7 and 5472, Mus. Geol. Surv. India, Calcutta. Leaf rather small, broadly spathulate, deeply lobed at the apex. NOEGGERATHIOPSIS. 18'J Lobes linear, acuminate. Nerves strong, radiating from the base, dichotomous. Leaf apparently pleated. Zeiller has recently pointed out that it is very doubtful if this fossil should be included in the genus JYoeggerathiopsis, and in this conclusion I agree. I have, therefore, enclosed the generic name in square brackets to indicate that this species should be transferred to some other genus, but, not having seen any specimens of this plant, I am unable to suggest any change at present. Feistmantel regarded these leaves as pinnules of a pinnate frond, but there is apparently no real evidence in favour of this conclusion. They differ markedly from N. Uidopi, both in the divided apical Fig. 39. — [Noeggerathiopxis ?] laeerata, Feist. After Feistmantel. Nat. size. portion and also in the nervation. Zeiller has suggested that they may be more closely compared with certain members of the class Ginkgoales, such as Ginkgo, Ginkgodium, and Whittlesey a. This species is known only from the Karharbari and Raniganj Groups in India. Not represented in the British Museum collection. Leaf of Doubtful Affinity from S. Africa. V. 3614. Figured by Seward (97 1 ), p. 333, pi. xxi, fig. 5. This specimen shows a strongly convex leaf, 3*8 cm. long, and 7 mm. broad. The apex ends in a sharp, narrow point. The base 190 DAD0XYL0N. is not well preserved. It is traversed by a number of parallel veins. It is impossible to refer this fragment to any genus. The closest comparison is with the Gymnosperms, as Mr. Seward has pointed out. Casey's Township, Transvaal. Pres. by D. Draper, Esq., 1897. Genus DADOXYLON, Endlicher, 1847. (Syuop. Conifer., p. 298.) Endlicher defined this genus thus : — " Truncus cylindricus, e medulla centrali, et ligni stratis concentricis obsoletis aut rarius distinctis et e cortice compositus. Vasa (ligni cellulae prosenchymatosae) porosa, poris in series 1-4 spiraliter dispositis, quam maxime appi'oximatis, demum ob mutuant pressionem sex- angularibus, pleTumque nonnisi in parietibus radiis medullaribus parallelis, et invicem oppositis obviis. Radii medullares e cellu- larum parenchymatosarum serie unica v. pluribus formati, simplices v. compositi." This genus is now, however, used in a restricted sense, and the following may be regarded as its chief characters: — Petrified wood of Palaeozoic age of the Coniferous type, closely similar to that of the modern Araucariae. Pings of growth usually well marked. Xylem centrifugally developed, tracheides as a rule pitted only on the radial walls. Bordered pits, hexagonal in outline, usually contiguous, pluriseriate, more rarely distant and uni- or biseriate. Pith usually large, fistular or solid. Cortex as a rule without gum-canals or resin-cells, but the latter are sometimes found in the cortex, wood, and pith. In both Palaeozoic and Mesozoic rocks petrified woods of the Araucarian type occur, whose precise affinities it is often difficult or impossible to determine. The term Araucarioxylon, Kraus, was at one time ' applied to such specimens, whether of Palaeozoic or Mesozoic age, but it has been recently decided that this name should be restricted to woods of the latter age. 2 The old genus Dadoxylon has been revived, and to it the Palaeozoic specimens are 1 Schimper (69), vol. ii, p. 380. 2 Zeiller (00 1 ), p. 280; Knowlton (90), p. 006 ; Scott (02), p. 331. DADOXYLON. 191 now referred. Some woods of this type have proved to belong to Cordaites, while in the great majority of cases there is no evidence as to the precise affinity of such specimens. It would seem probable that some of the Australian woods described here may belong to Noeggerathiopsis, but at present there is no definite evidence that this is the case. 1. Dadoxylon australe, sp. nov. (Text-figs. 40-43.) Nicol 1 figured, in 1833, some transverse sections of petrified woods from Australia. There are several similar sections in the British Museum collection, most of which are obviously of the same tj'pe, and these I propose to describe as Dadoxylon australe, since Nicol did not give any name to his specimens. Co-types. — Transverse sections, 51,479 (Australia), 51,434 (Newcastle). Kadial sections, 51,616 (Australia), 51,476 (New- castle). Tangential sections, 51,503 (Australia), 51,461 (New- castle). Trans., Bad., and Tang, sections, V. 8299 (Australia). Pith not preserved. Xylcm with well-marked rings of growth, each as a rule between 3"5 and S'5 mm. in radius. Some- times the growth is irregular. The ' spring wood ' graduates imperceptibly into the denser 'autumn wood.' Xylem com- paratively dense, apparently entirely centrifugal. Tracheides, in transverse section, very small, narrow, square, or slightly oblong, with rounded angles, about 25 in a millimetre of length. Tracheides short, with hexagonal bordered pits only on the radial walls. Usually these pits are multiseriate and crowded, but not infrequently they are found to be uni- or biseriate, distant or few. Medullary rays extremely numerous, uuiseriate, very rarely two cells thick, usually 6-12 or more cells in height. The medullary ray cells communicate with the tracheides by 2-6 simple, oblique pits. Cortex absent. It has not been found possible to identify the actual specimens figured by Nicol, but they are probably among those mentioned here. In describing this species, difficulty has arisen from the fact 1 Nicol (33), pi. iii, fig's. 1-3. 192 D.VDOXYLOX. that not only is there no complete section showing the cortex, wood, and pith in the collection, but it is impossible to say which transverse and longitudinal sections of the wood were originally cut from the same specimen. However, the great majority are undoubtedly identical specifically, and I have therefore chosen two of the best sections cut in each direction as co-types together witli Fig. 40. — Dadoxylon australe, sp. nov. Transverse section. 5 1,479. x 25. No. V. 8299, which has three sections on one slide, and thus shows that the longitudinal sections described here belong to the same species of stem as those cut in a transverse direction. The special characteristics of this species are the small and short tracheides, and the numerous medullary rays. DAD0XYL0N. 193 The great majority of Nicol's sections are from Newcastle. In a collection which Mr. A. G. Hamilton, of Willoughby, New South Wales, very kindly placed in my hands for examination, I have recognised several sections from Avondale, Mount Kembla, Mount Keira, and Bellambi, all in the Illawarra district, as probably identical with this species. Fig. 41. — Dadoxy Ion australe, sp. nov. Transverse section. V. 8299. x 30. Mr. Hamilton informs me that petrified wood is very common in the Newcastle Series of New South Wales, especially in the Illawarra district, in association with Glossopteris. Large trunks of trees have been found at Mount Kembla, and elsewhere stumps and trunks, as well as smaller fragments, are common. As far back 194 DADOXYLON. as 1843 the Rev. TV. B. Clarke 1 gave some account of the fossil wood occurring- in abundance at Lake Macquarie, New South Wales, and some of Nicol's sections, figured in 1833, came from the same locality. Schenk 2 has also described, under the name Araucarioxylcn Felixianum, some woods from Queensland and Kiami in New Fig. 42. — Badoxylon australe, sp. nov Radial longitudinal section. 51,616. x 35. South Wales, in the botanical collection at Leipzig. From his brief description these do not appear to be identical with the present species. 1 Clarke (43). Schenk in Zittel, p. 870. DADOXYLON. 195 Carruthers 1 mentioned petrified wood from Jack's Creek, Queens- land, but did not give any description. His name Araucarioxylon Nicoli is a nomen nudum. Jack & Etheridge 2 also pointed out the common occurrence of fossil wood on certain horizons in the Bowen Series. 51,479. Text-fig. 40. Co-type. Trans. Sect. A large and good section, showing several rings of growth, of which a part is reproduced from a microphotograph in Text-fig. 40, enlarged 25 times. This figure shows the small size of the tracheides, and the numerous medullary rays ; also the gradual transition from the larger elements of the ' spring wood ' to the denser and smaller ' autumn ' tracheides. Australia. Nicol Coll. 51,434. Co-type. Trans. Sect. A rather smaller section than the previous specimen, but identical in every respect. The rings of growth are clearly seen. They are sometimes irregular, as Nicol noticed in his description of similar sections ; false rings of varying outline occur, which are due to bad preservation. Newcastle, New South Wales. Nicol Coll. 51,616. Text-fig. 42. Co-type. Had. Long. Sect. A well-preserved section, a small portion of which is figured enlarged 35 times in Text-fig. 42. It shows the numerous medullary rays and the pitted tracheides. Australia. Nicol Coll. 51,476. Co-type. Had. Long. Sect. A section similar to the last, but not quite so thin. The large number of the medullary rays and the short tracheides are well seen. The bordered pits, often only uni- or biseriate, can here and there be distinguished. Bottom of the cliff, Newcastle, New South Wales. Nicol Coll. 51,503. Text-fig. 43. Co-type. Tang. Long. Sect. A small portion of this section is reproduced from a microphotograph enlarged 35 times in Text-fig. 43. The numerous medullary ray cells, only one cell thick, are seen. The absence of bordered pits on the tangential walls of the tracheides is noticeable. Australia. Nicol Coll. ' Carruthers (SO), p. 32S. 2 j ac t & Etheridge (92), p. 198. 196 DAD0XYL0N. 51,461. Co-type. Tang. Long. Sect. A tangential section similar to 51,503, showing the medullary rays and tracheides. Bottom of the cliff, Newcastle, New South Wales. JSficol Coll. V. 8299. Text-fig. 41. Co-type. Trans., Bad., and Tang. Long. Sections. These sections, all mounted on one slide, are precisely identical in structure with Fig. 43. — Dadoxylon australe, sp. now Tangential longitudinal section. 51,503. x 35. the specimens above described. There is every reason to believe that all three were cut from the same specimen, but in different directions. The transverse section is very thin, and is particularly well preserved, a portion of three rings of growth being seen. Part of DADOXYLON. 197 this section is figured in Text-fig. 41, enlarged 30 times. Among the xyleni elements a number of cells with brown contents are found, sometimes forming a well - marked ring, at other times more scattered. These are possibly secretory cells. In the radial section, the short tracheides with bordered pits, and the numerous medullary rays are very clear. This is also the case in the tangential section, where, however, pits are absent from the walls of the tracheides. Australia (?near Newcastle). 51,475. Trans. Sect. A section precisely similar "to sections 51,479 and 51,434. The rings of growth are apparently irregular, owing to bad preservation along certain lines. Bottom of the cliff, Newcastle, New South Wales. Nicol Coll. 51,455. Tang. Long. Sect. A section similar to 51,503 and 51,461, but not so well preserved. The section in part is nearly radial. Newcastle, New South Wales. Nicol Coll. 51.445. Had. Long. Sect. A radial section similar to 51,616 and 51,476, but not so thin. Newcastle, New South Wales. Nicol Coll. 51,423. Trails. Sect. A good section showing the medullary rays and well-marked rings of growth. Australia. Nicol Coll. 51,432. Trans. Sect. A section, probably of this species, 4 3 cm. long, but only showing one ring of growth. The medullary rays are well marked. Scaffold Hill Quarry, near Newcastle, New South Wales. Nicol Coll. 51.446. Trans. Sect. The square tracheides, the medullary rays, and the rings of growth are seen in this section. Australia. Nicol Coll. 51,459. Trans. Sect. A badly preserved, crushed, and distorted stem, possibly of this species. Lake Macquarie, New South Wales. Nicol Coll. 51,462. Had. Long. Sect. A well-preserved section, showing the tracheides with multiseriate bordered pits, and the medullary rays, usually 6-9 cells in height, very numerous, or almost crowded. Australia. Nicol Coll. 198 DAD0XYL0N. 51,467. Had. Long. Sect. Not very well preserved, but showing the numerous medullary rays. Scaffold Hill, near Newcastle, New South Wales. Nicol Coll. 51.470. Had. Long. Sect. A good section, in which the multi- seriate bordered pits are seen here and there. New South Wales. Nicol Coll. 51,474. Trans. Sect. A fairly well-preserved section, showing the small, almost square tracheides, numerous medullary rays, and well-marked rings of growth. Australia. Nicol Coll. 51,478. Had. Long. Sect. This section may possibly have been derived from the same petrifaction as section 51,475. It shows the bordered pits and medullary rays clearly. Bottom of cliff, near Newcastle, New South Wales. Nicol Coll. 51,502. Trans. Sect. A well-preserved section, possibly derived from the same petrifaction as section 51,479 (co-type). Australia. Nicol Coll. 51,505. Trans. Sect. A fairly good section, showing the medullary rays and rings of growth. Australia. Nicol Coll. 51,617. Tang. Long. Sect. A good tangential section, possibly derived from the same petrifaction as section 51,616 (co-type). Australia. Nicol Coll. V. 8238. Trans., Bad., and Tang. Long. Sects. A good series of sections in three directions, showing the structure of the tracheides and medullary rays. The radial section resembles section 51,616 (co-type). Australia. V. 8287. Trans. Sect. A fairly good section, showing the medullary rays and rings of growth. New South Wales. 51.471. Trans. Sect. A rather crushed stem, but the structure is fairly clear in places. Australia. Nicol Coll. 51,504. Had. Long. Sect. Not well preserved, but the medullary rays are fairly clear, and the pits on the tracheides can be seen here and there. Australia. Nicol Coll. DADOXYLON. 190 51,729. Trans. Sect. A good section, showing the rings of growth, medullary rays, and small tracheides. Australia. Nicol Coll. 51,794. Had. Long. Sect. The medullary rays and multiseriate bordered pits are well seen in this section. Bottom of cliff, near Newcastle, New South Wales. Nicol Coll. Some of the following indifferently preserved sections may belong to Badoxylon australe, while others perhaps to a different species. Other sections : — Transverse. 51,419 (Australia), 51,435 (New- castle, New South Wales), 51,436 (Newcastle), 51,457 (Newcastle), 51,460 (bed of Hunter River, New South Wales), 51,459 and 51,464 (Lake Macquarie, New South Wales), 51,468 (Newcastle), 51,469 (New South Wales), 51,477 (bottom of cliff, near New- castle), 51,501 (Newcastle), 51,545 (26 miles north-west of Newcastle), 51,615 (Lake Macquarie), 51,655 and 51,657 (Tele- graph Hill, Newcastle), 51,683 (bottom of cliff, near Newcastle), 51,686, 51,692, and 51,708 (Newcastle), 51,755 and 51,768 (Wollongong). Had. Long. Sects. 51,472 (New South Wales), 51,525 (Australia). Tang. Long. Sects. 51,918 and 51,919 (Sydney, New South Wales). Ail the above in the Nicol Coll. Trans. Sects. V. 8286 (Lake Macquarie), V. 8289 (Hunter River), V. 8291 (Botany Bay), V. 8300 (Australia). Tang. Long. Sects. V. 8288 (Hunter River), V. 8290 (New South Wales). All the above in the General Collection of Sections. 2. Dadoxyloxi Pedroi, Zeiller. 1895. Badoxylon Pedroi, Zeiller, Bull. Soc. Geol. France, ser. in, vol. xxiii, p. 619, pi. ix, fig. 4, aud text-figs. 8-19. D. Pedroi, Zeiller, Compt. Read., vol. exxi, p. 964. 1902. I). Pedroi, Scott, Trans. Eoy. Soc. Edinb., vol. xl, pt. 2, p. 359. Type. Ecole superieure des Mines, Paris. Stem 6 cm. long. Pith large, 37-38 mm. in diameter, with three prominent angles, marking the points of origin of lateral 200 DADOXYLOX. appendages -which were borne at equal distances from one another throughout the length of the specimen. Pith solid, parenchymatous, but sometimes the tissues are only imperfectly preserved. The cells are larger in the centre of the pith, and smaller bordering on the wood, where they are 4-5 times higher than broad, and arranged in vertical rows. At the centre, the cells are more irregularly arranged, and are 1-25 times broader than high. Secretory canals, surrounded by a pseudo-sheath of parenchymatous tissue, also occur in this region. They form groups of from 2 to 15 or 20 elements. Xylem elements exclusively centrifugal, consisting of spiral, annular, and pitted tracheides, as in Cordaites, but the pits are usually less numerous and less crowded, and the central pore is exactly circular and not elliptical. The pits are sometimes uni- seriate, sometimes biseriate, but not pi uri seriate, generally contiguous, rarely distant. The radial rows of tracheides do not form definite bundles, but are grouped together in a variable number of rows, separated by medullary rays. The medullary ray consists of large cells, and is usually only one cell broad, except on the border of the pith. The height of the ray is considerable, and may be as much as fifty superposed cells. The cells are polygonal, elongated in the direction of the ray, generally 5-6 times longer than broad. A few oblique small pits occur in the wall of the cells adjacent to the tracheides. Cortex absent. Zeiller has figured bacteria found in some of the more altered or decomposed medullary ray cells. He points out that Dadoxylon Pedroi resembles Cordaites in having a large pith. It differs from that genus in the pith being solid and without diaphragms. In the occurrence of secretory canals in the pith, as well as in the size of the pith, this species recalls the structure of the Cycadege. He concludes, however, that this stem has more or less affinity with the Cordaitales, such as perhaps Noeyyerathiopsis and Puryphyllum, but that this suggestion is little more than a conjecture. Dadoxylon Pedroi is known only from the valley of the Jaguarao, Brazil. Not represented in the British Museum collection. DADOXYLON. 201 3. Dadoxylon Maitlandi (Shirley). 1898. Araucarioxylon Maitlandi, Shirley, Bull. 7, Geol. Surv. Queensland, p. 14. Type. Brisbane Museum, Geol. Surv. Coll., Cease 53, No. 73. "Annual zones of wood distinctly marked, and of equal transverse diameter ; medullary rays of one or two rows of cells, showing in a tangential section 4-26 cellules, the central few often flanked by a short second row ; pits spirally arranged in 2-3 rows. Near the annual zones the tracheides are peculiarly compressed, and the ranks distorted " (from Shirley). D. Maitlandi is known only from Yam Creek, near Cooktown, Upper Bowen Formation, Queensland. Not represented in the British Museum collection. 4. Dadoxylon Binneyi (Shirley). 1898. Araucarioxylon Binneyi, Shirley, Bull. 7, Geol. Surv. Queensland, p. 14, pi. xxv. Type. Brisbane Museum, Geol. Surv. Coll., No. 77. "Annual zones of wood distinct, varying much in transverse diameter; medullary rays of a single chain of cellules, in a tangential section numbering 2-9 ; tracheides usually with three rows of pits, spirally arranged and polygonal by compression " (from Shirley). D. Binneyi is known only from Jack's Creek, Bowen River, Upper Bowen Formation, Queensland. Not represented in the British Museum collection. 5. Dadoxylon Williamsoni (Shirley). 1898. Araucarioxylon Williamsoni, Shirley, Bull. 7, Geol. Surv. Queensland, p. 15. Type. Brisbane Museum, Geol. Surv. Coll., Case 53, No. 74. " Annual zones moderately distinct in transverse section, and of equal diameter ; medullary rays of one or two rows of cells ; tracheides unusually broad, and very unequal ; pits spirally arranged in two or three rows, densely packed and polygonal. Near the boundaries of the annual zones the tracheides are peculiarly complicated, and the ranks distorted " (from Shirley). 202 DAD0XYL0X. D. WilUamsoni is known only from Jack's Creek, Bowen River, Upper Bowen Formation, Queensland. Not represented in the British Museum collection. 6. Dadoxylon brisbanense (Shirley). 1898. Araucarioxylon brisbanense, Shirley, Bull. 7, Geol. Surv. Queensland, p. 15, pi. xxvi. Type. Brisbane Museum, Geol. Surv. Coll., No. 91. " Annual zones of wood well marked, the different zones varying considerably in radial length ; tracheides of medium diameter, fusiform or cylindrical, terminations overlapping or ending bluntly ; pits small, rounded, usually in two rows spirally arranged; in transverse section the cut ends of the tracheides are sub-quadrate ; in tangential section the medullary rays show a single series of 2-10 cells, the average number being five" (from Shirley). Mr. Shirley has also described some other woods as new species, but in these cases either the precise age is unknown or the preservation does not appear to be very good. D. brisbanense is known only from Ann Street, North Brisbane. Not represented in the British Museum collection. 7. Dadoxylon, sp. (from Australia). The following sections may belong to an undescribed species, but the preservation is not sufficiently good to determine them specifically : — 39,144. Trans. Sect. Two sections of a stem, somewhat im- perfect in places, showing large rings of growth, the medullary rays, and the tracheides. The petrifaction (three pieces) is also registered under this number. Port Stephens, New South Wales. Odinheimer Coll. V. 7224. Trans. Sect. A section similar to the preceding. The petrifaction (two pieces) is also registered under this number. Port Stephens, New South Wales. Odinheimer Coll. V. 7226. Trans. Sect. A section similar to the last. The petrifaction (two pieces) is registered under this number. Port Stephens, Kew South Wales. Odinheimer Coll. rimuFiED woods. 203 V. 7225. Trans. Sect. A badly-preserved section, but similar in type to the preceding. The petrifaction (two pieces) is registered under this number. Tort Stephens, JN'ew South Wales. Odinheimer Coll. 51,458. Trans. Sect. A section of an almost complete stem showing the pith-cavity, which is small owing to compression. This stem may be identical with D. australe, but without longi- tudinal sections it is impossible to determine it. The appearance of the transverse section, however, is not quite similar to those described under that species. Australia. Nicol Coll. 51,456. Trans. Sect. A similar specimen, cut from the same petrifaction. Australia. JSficol Coll. V. 8292. Tang. Long. Sect. A section showing the tracheides without pits on the tangential walls, and the uniseriate medullary rays. Queensland. V. 8293, V. 8294. Trans. Sects. Sections showing thick-walled tracheides, with well-marked rings of growth. Queensland. V. 8295. Had. Long. Sect. An indifferent section, but showing here and there the multiseriate bordered pits on the walls of the tracheides. Queensland. V. 8296. Tang. Long. Sect. In this section the walls are falsely pitted owing to imperfect mineral crystallization. Queensland. V. 8298. Had. Long. Sect. A poor section. Queensland. IN CERT M SEDIS. PETRIFIED WOOLS {WITHOUT SECTIONS). 39,143. Port Stephens, New South Wales. 32,477. Newcastle, New South Wales. Pres. hg Sir E. Rome. V. 7289. Port Stephens, New South Wales. Odinheimer Coll. V. 10,650. Over "Dirty Seam," Newcastle, New South Wales. Kcene Coll. 204 CARDIOCARPUS. V. 10,651. Cylindrical piece of wood from a bed of chert. Wollongong, New South Wales. Keene Coll. V. 10,652. A wedge-shaped piece of wood. Newcastle Series, New South Wales. Keene Coll. V. 10,654. Stoney Creek, West Maitlaud, New South Wales. Keene Coll. V. 2394. Six pieces of silicified wood and one obscure specimen. Beaufort or ? Ecca Series, Korn-koin River (a branch of the Kaga River \ south of the Winterburg, Cape Colony. V. 2395, V. 2395«, Two pieces of wood. Beaufort Series, Zwart-kei River, Cape Colony. V. 2420. Fossil wood. Beaufort (?) Series, Rietfontein, near Prince Albert (The Gouph), Cape Colony. V. 489. ? Ecca Series, New Colesburg Kop (at a depth of 22 feet), De Beers, near Kimberley, South Africa. Pres. by Prof. T. R. Jones, 1884. V. 3254. Bedford, Cape Colony. Pres. 01/ D. Draper, Esq., 1893. ? V. 360. Main street, town of Tette. Livingstone Coll. GYMNOSPERMOUS SEEDS (Possibly of Cordaitean affinity). Detached and isolated Gymnospermous seeds occasionally occur in the rocks of Gondwanaland. Some of these probably belong to the Cordaitales or the Coniferales, while others may eventually be referred to the Cycadales or the Ginkgoales. Genus CARDIOCARPUS, Brongniart, 1828. [Prodr. Hist. Veget. foss., p. 87.] Platyspermic (bilateral) seeds, lenticular, reniform or cordate. The outer, fleshy portion of the testa (sarcotesta) has often the appearance of being winged. The central body is in most cases the hard sclerotesta, surrounding a cast of the nucellus. Some seeds of this type are known to belong to members of the Cordaitales. The genus is essentially of Palaeozoic age. CARDIOCARPUS. 205 1. Cardiocarpus indicus, Zeiller. (Text-fig. 44.) 1902. Cardiocarpus indicus, Zeiller, Pal. Indica, n.r., vol. ii, p. 37, pi. vii, figs. 7, 8. Type. ISTos. 7311-12, Mus. Geol. Surv. India, Calcutta, Sclerotesta orbicular-cordate, 30-35 mm. in diameter, surrounded by an oval sarcotesta, 50-55 mm. in beigbt and 40-45 mm. broad, at the apex deeply eniarginate. ~7\ ^,. v<*^ Fig. 44. — Cardiocarpus indicus, Zeiller. Alter Zeiller. Nat. si/.e. This fine seed (Text-fig. 44), recently described by Zeiller from the Karharbari Group in India, is regarded by him as most probably belonging to a member of the Cordaitales. This species is possibly the largest known member of the genus. Not represented in the British Museum collection. 2. Cardiocarpus (?) Milleri (Feistmantel). 1879. CarpolitJics Milleri, Feistmantel, Flora Gondw. Syst., vol. iii, pt. 1, p. 30. 1881. C. Milleri, Feistmantel, ibid., vol. iii, pt. 1, Suppl., p. 59, pi. xxx, fig. 14. 1882. C. Milleri, Feistmantel, ibid., vol. iv, pt. 1, p. 43, pi. xv, figs. 5-12. 206 CAKDIOCARPUS. Type. ? Mus. Geol. Surv. India, Calcutta. Sclerotesta ovate, base rounded, ? emarginate, apex obtuse, 2-3 cm. in height, and 1*4-2 cm. in breadth. Sarcotesta ovate, base emarginate, apex acute, 2-3-3 cm. in height, and 1-8-2- 1 cm. in breadth. In the type-specimen figured by Feistmantel the apex is divided into two lobes, and the sclerotesta has a median groove. The sarcotesta in this species is not much larger than the sclerotesta, and it is possible that this seed may have been of the radiospermic type. For the present, however, it may be referred to the genus Cardiocarpus. Cardiocarpus Mitteri occurs only in the Karharbari Beds of India. Not represented in the British Museum collection. 3. Cardiocarpus, sp. (from India). Zeiller ' has figured some seeds from India with a cordate sclerotesta, 7-8 mm. in height and about 7 mm. broad, pointed at the apex, surrounded by a rather narrow sarcotesta, fleshy or perhaps fibrous. This fossil is only known from the Barakar Group, India. It appears to be somewhat similar to the specimen from the Newcastle Series, Australia (see below), which I figured in 1902. 2 Eeistmantel 3 has also described several seeds of this type, which he named Cardiocarpum(T), sp., Carpolithes, sp., and. Samaropsis, sp. V. 7110. A nearly perfect specimen of a seed showing the sclerotesta and sarcotesta, the latter having the appearance of being narrowly winged. The sclerotesta measures 6 mm. in length and 6-5 mm. in breadth. Bharatwada, India. Hunter Coll. V. 7135. A few small seeds, not very well preserved, in association with fronds of Glosmpteris Broivniana. The seeds are possibly referable to Cardiocarpus. Kamthi, India. Hunter Coll. 1 Zeiller (02 1 ), p. 38, pi. vii, fig. 11. 2 Arber (02 1 ), p. 20, text-tig-. 3 Feistmantel (79 1 ), p. 30, pi. xxiv, fig. ~> ; p. (10. pi. xxx, fig. 15 ; p. 59, pi. xxviii, fig. 8; pi. xxx, figs. 11-13; (80), p. 123, pi. xIyua, figs. 9-15; 82'), p. 50, pi. si, fig. 7 ; pi. xvi. fig. Ac ; pi. \\i, fig. 2; (SO), p. 45, pi. xiiA, figs. 7 15; pi. xivA, figs, s, '.) ; pi. va, figs. 8« <-. CARDIOCARrUS. '207 4. Cardiocarpus, sp. (from Australasia). I figured in 1902 l some subcircular seeds of medium size, from Newcastle in New South Wales, with an oval sclerotesta 5-6 mm. long, and a sarcotesta 6-5-7 - 5mm. in height. Shirley 2 has figured a seed from Queensland as Cycadospermum Dawsoni, Shirley, which somewhat recalls some of the Indian seeds noticed above. Johnston 3 has also figured a seed under the name Carpolithes tasmanicus, but without description, and others which he compares with Rhipidopsis ! 32,477. This specimen shows two fragments of large fronds of Glossopteris Brown/ana with broad nets, and a portion of the rhizome, Vertebraria. There is also a seed-like body, not very well preserved, of pear-like form, measuring 12 mm. long and 11mm. across at the widest part. There does not appear to be any wing, and the seed perhaps should be referred to some other genus, but the absence of a wing in this case may be due to imperfect preservation. Hunter River, New South Wales. Pies, by Sir E. Home, 1853. V. 3776«rt. A winged seed occurs on this specimen in association with fronds of Glossopteris and Noeggerathiopsis. The sclerotesta measures 9"5 mm. in length, and the wing is fairly broad. Mersey River, Tasmania. Pres. by T. Stephens, Esq., 1898. V. 3776(?/?. Possibly one or two fragments of winged seeds occur on this specimen. Mersey River, Tasmania. Pres. by T. Stephens, Esq., 1898. V. 3776//. This specimen, in addition to the leaf of Noeggera- thiopsis Hislopi figured ou PI. VI, Fig. 3, shows a small seed with a fairly broad wing. Mersey River, Tasmania. Pres. by T. Stephens, Esq., 1898. V. 3776«/. Two or three badly-preserved and rather small seeds occur on this specimen, showing an oval sclerotesta, 5*5 mm. in length, and traces of a narrow wing. Mersey River, Tasmania. Pres. by T. Stephens, Esq., 1898. Arber (02 1 ), p. 20, text-fig. Shirley (02), p. 9, pi. iii. Johnston (88), pp. Ill, 134, pi. viii, fig. 7; p. 134, pi. viii. figs. 1 (i. 208 I'TEKOPHYLLVM. 5. Cardiocarpus, sp. (from South Africa). V. 3614. Figured by Seward (97'), p. 332, text-fig. Id on p. 324. A small seed, 9 mm. long and 65 mm. in breadth. Mr. Seward has pointed out that the broad margin surrounding the oval central region is interrupted at the base by what appears to be a canal leading to the base of the seed. This seed occurs with another specimen, described by Mr. Seward as a leaf of doubtful affiuity. Casey's Township, Transvaal. Pres. by B. Draper, Esq., 1897. Class CYCADOPHYTA. The term Cycadophyta has been recently proposed by Nathorst to include fossil fronds similar in habit to those of the recent Cycads, whose affinities, whether to that family or to other extinct races of the same stock, are at present unknown. Such fossils appear first in the Permo-Carboniferous rocks, and are especially characteristic of the Mesozoic period. A few Cycads, remnants of what was once a great and complex group, still survive at the present day. Genus PTEROPHYLLUM, Brongniart, 1825. [Ann. Sci. Xat., ser. i, vol. iv, p. 211.] Fronds simply pinnate, pinna? attached to the rachis by their whole base, usually linear, with parallel sides, generally con- siderably longer than broad, truncated, rounded, or, more rarely, pointed at the apex, more or less spreading, sometimes curved. Nerves parallel, simple or dichotomously divided. This genus is one of the earliest of Cycadean fronds, appearing first in the Carboniferous, and reaching its maximum in the Triasso-Rhaetic. In the Jurassic and Cretaceous rocks it is less abundant. Pterophyllum (Anomozamites) Balli (Feistmantel). (Text-fig. 45.) 1881. Anomozamites (Pterophyllum) Balli, Feistmantel, Rec. Geol. Surv. India, vol. xiv, pt. 3, p. 2.36, pi. ii, figs. 3, 4. ITEUOPIIYLLUM. 209 1886. Platypterygium Balli, Feistmantel, Flora Gondw. Syst., vol. iv, pt. 2, p. 37, pi. iiA, figs. 4-8 ; pi. iii a, fig. 2 (pars). 1902. Pterophyllum (Anomozamites) Balli, Zeiller, Flore Foss. Gites Charb. Tonkin, p. 182. Type. Nos. 5504-8, Mus. Geol. Surv. India, Calcutta. Frond of medium size, elliptical, pinnately divided. Segments linear, of unequal breadth, decurrent at the base and truncated at the apex, apparently attached laterally to the rachis. Nerves 4-8, sub-parallel, simple or more often dichotomising, the first dichotomy taking place near the base of the segments. Fig. 4.3. -Pterophyllum (Anomozamites) Balli (Feist.). Alter Feistmantel. Nat. size. The name Anomozamites was originally instituted by Schimper to include a particular type of frund belonging to the large genus Pterophyllum, in which the leaflets are rectangular, truncated at the apex, with rounded angles, the distal margin being parallel to the rachis. Zeiller '' has recently described some fronds of this type from the 1 Zeiller (02 s ), p. 177, pi. xliii, fig. 8; pi. xliv, figs. 1-5. 210 CYCADITES. Rhaetic of Tonquin, one of which, Pteropliylhim (AnomozamiUs) inconstans (F. Braun), is not unlike Feistmantel's species. There can be little doubt that these fronds are of the Cycadean type, according to the at present accepted views as to the nature of such leaves. PteropJiyllum (Anomozamites) Balli is known only from the Barakar Group of the Daniuda division in India. Not represented in the British Museum collection. Genus CYCADITES, Sternberg, 1826. [Flora Vorwelt, Heft iv, p. xxxii.] Frond pinnate, pinna? alternate or opposite, linear, lanceolate, entire, with a single median vein ; attached to the rachis by the entire base, the lower margin of which may be slightly decurrent on the frond axis, or slightly narrowed towards the point of attachment. 1 Cycadites (?), sp. 1902. Cycadites (?), sp., Zeiller, Pal. Indica, N.8., vol. ii, p. 33, pi. vii, figs. 4, 5. Zeiller has recently ascribed to this genus, provisionally, a single, linear, uninerved leaf, 6 cm. long and 5 mm. broad, from the Karharbari Group of India, which is the only specimen of its kind known from Gondwanaland. Not represented in the British Museum collection. Class ? GINKGOALES. The class Ginkgoales includes fossil remains, chiefly foliage leaves and flowers, which may be compared more or less closely with Ginkgo (Salisburia), the Maidenhair-tree of Japan and China, now the sole survivor of what was once a great group of Gymnosperms. Such remains appear first in Permo-Carboniferous times, and are especially characteristic of the Mesozoic period. 1 Seward (95), p. 24. r.niPiDopsrs. 211 Genus EHIPIDOPSIS, Schmalhausen, 1879. [Mem. Acad. Imp. Sci. St. Petersb., ser. vir, vol. xxvii, p. 50.] Leaves large, usually orbicular or oval, with a long petiole, palraisect, coriaceous. Segments 5-10, cuneate, entire, obtuse, flabellate, lateral segments usually smaller than the median. Nerves numerous, radiating, several times dichotomised. This genus was instituted by Schmalhausen for the reception of some large leaves from rocks of Permian age in Russia not dis- similar in habit to those of Ginlcgo. Similar leaves have since been found in the Permo-Carboniferous of Tndia and Argentina. Fig. 46. — Rhipidopsis ginglcoides, Schmal. After Feistmantel. Nat. size. 1. Rhipidopsis g-inkgoid.es, Schmalhausen. (Text-fig. 46.) 1879. Rhipidopsis ginglcoides, Schmalhausen, Mem. Acad. Imp. Sci. St. Petersb., ser. vn, vol. xxvii, p. 50, pi. viii, figs. 3-12 ; pi. vi, fig. 1. 1881. R. ginglcoides, Feistmantel, Rec. Geol. Surv. India, vol. xiv, pt. 3, p. 257, pi. ii, fig. 2. R. ginglcoides, Feistmantel, Flora Gondw. Syst., vol. iii, pt. 3, p. 122. 1885. R. ginlcgoides, Renault, Cours Bot. foss., 4 e ann., p. 67, pi. v, figs. 3, 4, 7, 8. 1886. R. ginglcoides, Feistmantel, Flora Gondw. Syst., vol. iv, pt. 2, p. 43,