236 GLOSSOPTERIS AND ITS ALLIES, 



extending but a short distance into the substance of the frond* 

 No doubt Feistmantel's comparison of the fructification of the 

 two ferns was a very appropriate one. 



Feistmantel also believed | he had discovered traces of a third 

 kind of fructification iu the Indian G. angustifolia, Broug. On 

 the frond referred to there existed a clear space along each 

 margin, " which perhaps shows that in fertile fronds the fructifi- 

 cation is a marginal one." We would here suggest a comparison 

 with the genus Schizoloma amongst recent ferns. He generally 

 confirmed Brongniart and Bunbury's observations on the fruiting 

 of G. indica, and gave some excellent additional figures of similar 

 sorus scars in G. communis, Feist. 



In lending provisional support to these views, Zeiller remarked 

 on the absence of any hitherto discovered sporangia, and says that 

 provided the foregoing evidences of fructification are sound, it is 

 difticult to doubt the presence of the three groups proposed by 

 Feistmantel. Their conjoint opinion may be briefly expressed 

 thus :— 



Group 1. — Type of G. browniana, Carruthers, allied to genera 

 of living ferns, such as Anthrophyum, with linear sori on the 

 nervules. 



Group 2. — Type of G. indica, Schimper (G. browniana, var. b. 

 indica, Brong.), allied to genera of living ferns, such as Polypo. 

 dium, with round sori in longitudinal rows parallel to the frond 

 margin. 



Group 3. — Type of G. angustifolia, Feistmantel (? G. angusti- 

 folia, Brong.), with a supposed marginal fructification along the 

 edge of the frond, as in the living genus Pteris. 



Now, there is one weak point in this classification. We 

 have already pointed out that, according to strict priority, should 

 any division of Glossopteris be necessary, it is to Brongniart's 



* See Beddome's Ferns of Southern India, 1863, t. 52. 

 + Loc. cit. p. 97, t. 39a, f. 1 and 2. 



