XXI v] 



GAMETOPHYTE 



213 



Ferns does this come out in so interesting a way as in Gleiclienia. It has 

 already been seen that the genus is variable in spore-output, which is larger 

 in Dicranoptcris than in Eii-Gleichenia. The spore-number in G. flabellata 

 is higher than in any other recorded Leptosporangiate Fern. For the pur- 

 pose of comparing this with the number of spermatocytes, a rough measure 

 may be obtained by counting the number of these traversed in a single 

 section, though this naturally falls far below the total : but it will serve for 

 a rough comparison with similar sections through the mass of spore-mother- 

 cells. The result is shown in the followino- table: 



From these data, which should be extended by other observers, it appears 

 that, speaking generally, Dicranopteris has a larger spore-output per spor- 

 angium than Eii-Gleichenia. So far as the available evidence goes it has 

 also a larger output of sperms per antheridium. These facts point directly 

 to a primitive relationship for Dicranopteris. On the other hand, the spore- 

 output for En-GleicJienia, though considerable, rarely exceeds 256 per 

 sporangium. The only record to hand gives a very low figure for the 

 spermatocytes per antheridium, in En-Gl. polypodioides. Both these features 

 suggest for En-Gleichenia a derivative position compared with Dicranopteris. 

 It need not be expected that the numbers should run exactly parallel: it 

 may be held as remarkable that the parallel should come out as clearly as 

 it does. 



Gleichenia pectinata is thus seen to possess relatively large antheridia and 

 many sperms, together with a relatively large spore-output: but G. linearis 

 has fewer sperms per antheridium, while its spore-mother-cells (20, 15 in sec.) 

 and spores actually counted (120-228) are fewer than in other species of 

 Dicranopteris. These two species have been seen to stand apart in other 

 respects from their Section, and this is borne out b}' the figures, especially 

 for G. linearis. They confirm these species as holding a peculiar and an aloof 

 position: they have hitherto been ranked with Dicranoptcris, but are not 

 typical of that sub-genus. 



