part 4] JSUMMULOSPEliMCM BOWENEX:SE. 289 



13. 0)1 NuMMULOsPERMUM, gen. nov., ilie phobable Mega- 

 sporangium of Glossopteris, By A. B. Walkom, D.Sc. 

 (Communicated by Prof. A. C. Seward, Sc.D., F.Il.S., 

 F.G.S. Bead May 4th, 1921.) 



[Plate XXI.] 



AmOjS^g the fossil phints of Permo-Carboniferous age in the 

 collections of the Queensland Geological Survey, it was my good 

 fortune to find the specimens which form the subject of this 

 communication. The results of the examination of the whole 

 flora are being publislied by the Queensland Geological Surve}'^ 

 as a continuation of my studies of the Queensland fossil floras. 

 As, however, the specimens above mentioned represent, if my 

 interpretation of them is correct, a very important discovery in 

 fossil botanj^, it was thought that the subject merited a separate 

 communication. In doing this I hope that the attention of 

 palseobotanists may once more be drawn to the subject of the 

 systematic position of Glossopteris, and that the evidence here 

 brought forward may be critically considered in its bearing on 

 this problem. 



From time to time there have been suggestions that specimens 

 of GlossojJteris have been found showing structures regarded as 

 sori, but in no case has anything definite been proved in this 

 direction. As regards Australian specimens, these suggestions 

 date back to 1872, when William Carruthers, in describing 

 G. JBrown.iana collected by Daintree from Queensland, said : 



' one shows some indications of fruit in the form of linear sori running' 



along the veins, and occupying a position somewhat nearer to the margin of 

 the frond than to the midrib.' (Bibliography, No. 3, p. 354.) 



In 1905 perhaps the first defmite step was taken towards the 

 solution of the problem of the systematic position of this plant, in 

 Arber's discovery of small bodies attached to the scale-leaves of 

 Glossopteris, which he described as pi-obable microsporangia 

 (Bibliography, No. 1, p. 324 ; No. 2, p. 41). Then, in 1908, David 

 White noted the intimate association, in Brazil, of seeds of the 

 Samaropsis type with Gangamopteris (Bibliography, No. 8, 

 p. 559), and went so far as to state his belief that the seeds were 

 those of Gangamopteris, referring to them as Gangamopteris 

 (Samaropsis) Seixasi, sp. nov. Prof. Seward (Bibliograph}'', No. 7, 

 p. 354) suggests also that Samaropsis Miller i, a species occurring 

 in the Karharbari Beds in India, may be the seed of a species of 

 Glossopteris, one specimen occurring partly covered by a scale- 

 leaf of a type very similar to that generally recognized as belonging 

 to Glossopteris. 



