﻿334 ME. E. A. NEWELL ARBER ON THE [May I905, 



median nerve. These patches are oval in shape and fairly well- 

 defined, consisting of three or four slight depressions, which recall 

 the sori of certain recent Polypodiums. 



On the other hand, in the following year, Mr. Seward figured 

 a specimen very similar to Prof. Zeiller's leaf, also from South 

 Africa, which showed irregular elliptical or circular holes in the frond 

 instead of sorus-like patches. Mr. Seward has called attention 

 to the 



' danger of attaching any great importance to characters of this kind.' 



In the same year, he 2 also concluded a summary of the present 

 evidence with the following remarks : 



' On the whole, it would seem the safer course to admit that as yet no trust- 

 worthy example of a fertile 67/osso/^ms-frond has been recorded from either 

 India or Australia, and we have certainly no data on which any classification 

 can be legitimately founded.' 



Before leaving this subject, I may point out that a provisional 

 suggestion of my own, 3 put forward in 1903, when describing- 

 some Glossopteris-fionds from Rhodesia, is probably quite mis- 

 leading. Although I am still unable to account satisfactorily for 

 certain features presented by the fronds in question, I do not now 

 regard them as in any way connected with the fructification, as I 

 then suggested. 



Such, so far as I am aware, is practically all the evidence that we 

 have at present on the subject of the fructification of Glossopteris. 

 No organ of the nature of a sporangium has ever been described. 

 The evidence of the position of sori has been gathered from 

 spots, patches, or holes occurring on the lamina of certain of the 

 larger fronds. I have myself observed such occurrences, but in no 

 case have I obtained any real evidence that they are in any way 

 indicative of sori. In order to ascertain whether such holes and 

 patches occur among recent ferns of similar habit, I have looked 

 through the specimens in the Cambridge herbarium, and I have 

 found that holes are far from rare in such fronds. Spots or 

 patches also occur, here and there, on fronds, which are quite with- 

 out any connection with the fructification. A particularly-good 

 instance was found in a sterile frond of Acrostichum lanceolatum, 

 Hook., which corresponds very closely in habit with Glossopteris 

 Browniana, where both holes and well-marked spots, similar in 

 appearance to those mentioned above, were observed : the latter 

 being due to either fungal or insect-agency. 



VII. General Conclusions. 



From a consideration of what has previously been contributed 

 on the subject of the fructification of Glossopteris, we have seen 

 that the evidence that sori are to be found on the larger fronds is 

 not, at present, satisfactory. The suggestion, made here, that the 



1 Seward (97 1 ) p. 320 & pi. xxi, fig. 1. 



2 Seward (97 2 ) p. 181. 3 Arber (03) pp. 288-89. 



