NA TURE 



577 



THURSDAY, APRIL m. [906. 



THE GLOSSOPTERIS FLORA. 

 Catalogue o\ the Fossil Plants of the Glossopteris 

 Flora in the Department of Geology, British 

 Museum (Natural History); being a Monograph of 

 the Permo-Carboniferous Flora oj India and the 

 Southern Hemisphere. Pp. Ixxiv + 255. By E. A. 

 Newell Arber, (London: The British Museum 

 [Natural History]. Published by Order of the 

 Trustees, 1905.) Price 12s. 6d. 



IN ibis catalogue the author makes "an attempt 

 towards a complete summarj of what is al pre- 

 sent known on the subject of the Glossopteris Flora," 

 and he may be cordially congratulated on the 

 successful completion of a task both arduous and 

 difficult. Dr. Smith Woodward, the keeper of the 

 department of geology, has very wisely encouraged 

 the production ol catalogues, which are in realit) 

 monographs of the greatest value to both botanical 

 and geological students. The scattered literature 

 dealing with the Glossopteris flora renders the work 

 of the monographer very heavy, and the nature ol 

 much of the material on which he must base his con- 

 clusions necessitates considerable self-control and 

 caution. 



Mr. Arber'- introduction treats of the Glossopteris 

 flora under three heads: — (1) its botanical affinities; 

 (2) its distribution in space; (3) the evidence as to its 

 age and distribution in time. The sediments ol 

 Upper Carboniferous and Permian age have yielded 

 a rich supply of fossil plants in both hemispheres; 

 but in tin case of India, Australia, South Africa, and 

 South America the difficulty of drawing a satisfactory 

 line between the Carboniferous and Permian strata 

 has forced geologists to adopt the term Permo- 

 Carboniferous for the plant-bearing beds of India and 

 mote southern lands. In these Permo-Carboniferous 

 strata the genus Glossopteris is tin- most abundant 

 fossil, and for this reason the southern vegetation 

 which flourished during the epoch between the Lower 

 < arboniferous and Triassic periods has been desig- 

 nated the Glossopteris flora. It has long been 

 recognised that the Glossopteris flora differs in too 

 many respects from the northern flora of the same 

 geological age to justify the belief, which was 

 formerly held, as to the world-wide distribution in 

 the latter part of the Palaeozoic era of the vegetation 

 which is represented by the rich stores of fossils in 

 the American and European Coal-measures. The 

 approximate distribution of the plants of the two 

 provinces is clearly shown in the maps published by 

 Mr. Arber (p. xix.). 



One verj serious difficulty in tin- way ol giving a 

 satisfactory botanical account of the Glossopteris flora 

 i- the lack of petrified material, and, in the case of 

 nearly all the genera, the absence ol fertile leaves or 

 -boots. The genus Catamites, represented bj several 



distinct types in the northern flora, has not SO far 

 been recognised in the Glossopteris flora; in its place 

 occur Phyllotheca and Schizoneura, two represent- 

 atives of the Equisetales about which our information 



NO. I903, VOL. 73] 



is still very incomplete. Both genera are adequately 

 treated by Mr. Arber, but the botanist, however 

 thorough the treatment, cannot help being made- 

 aware of the insufficiency of the material at bis dis- 

 posal. The genus Sphenophyllum affords another 

 example of a characteristic northern type which can 

 hardly be considered a true member "I the <,|o-so- 

 pteris flora. It is true, a- Mr. Arber points out, that 

 Prof. Zeiller has shown good reason lor referring 

 ibi' Indian specimens, for which Royle in 1833 insti- 

 tuted the genus Trizygia, to Sphenophyllum. Mr. 

 Arber figures a fragment from Natal as Spheno- 

 phyllum Sp., but the specimen is loo imperfect to 

 serve as satisfactory evidence of the existence of the 

 genus in South Africa. In (be case ol plants with 

 fern-like fronds we cannot speak with any confidence 

 as to their botanical position. 



The simple tongue-shaped leaves with a distinct 

 mid-rib and anastomosing secondary veins, which 

 Brongniart named Glossopteris and placed among 

 the ferns, have never been found, in spite of their 

 extraordinary abundance, with -<>ii or sporangia. A 

 recent discovery by Mr. Arber affords the first satis- 

 factory clue to the nature "I the sporophylls. The 

 fronds of Glossopteris are occasionally found in 

 association with smaller scale-leaves, and on these 

 -roups of sporangium-like organs have been detected. 

 These bodies are considered, on the whole, to exhibit 

 a greater resemblance to the microspores of recent 

 cycads than to the sporangia ol ferns. The palaso- 

 botahisl who describes specimens of Glossopteris is 

 compelled to face the problem of recognising specific 

 characters among (In- numerous leaf-forms, but if he 

 knows anything of recent ferns he must admit that he 

 has undertaken an impossible task. Mr. Arber, with 

 the conscientious care which characterises his work, 

 has grappled with this difficulty, and his synopsis of 

 species supplies us with the besl working scheme so 

 far devised. 



Neuropteridium is another southern genus founded 

 on simply pinnate and sterile fronds, which, like 

 Glossopteris and Gangamopteris, must be lift as a 

 plant of doubtful position. Among other members ol 

 the Glossopteris flora which it has been customary 

 to pl.ae among the ferns are 'Leniopteris, Spheno- 

 pteris, and Pecopteris ; but the absence of fertile 

 fronds again sets a limit to our knowledge. One 

 interesting conclusion to be gleaned from the occur- 

 rence of certain ferns in the southern Permo-Carbon- 

 iferous vegetation (e.g. Tasniopteris and Clado- 

 phlebis) is that the Glossopteris flora includes types 

 which, in the northern hemisphere, are rather 

 Mesozoic than Palaeozoic. 



Schizoneura, Neuropteridium, and Voltzia (a 

 conifer) also represent other genera which apparently 

 migrated into Europe in earl} Mesozoic times. 

 Another striking fact is the absence of an) members 

 of the lycopodiales in tin Glossopteris flora of 

 Australia and India. On the other hand, Lepido- 

 dendron, Sigillaria, and some rather obscure speci- 

 mens referred to Bothrodi ndron have been found in 

 South Africa and South America. Mr. Arber takes 

 ihe view that the occurreno el these fossils in South 

 Vmerica and South Africa ma> I" unted for by 



