XX] 



SORI 



117 



with those of the Coenopterids, and especially of Etaptcris (Fig. 333), and 

 Corynepteris (Fig. 405). 



The facts derived from the fossils and their comparison with the living 

 Ophioglossaceae and Marattiaceae suggest in the first instance a marginal 

 origin of the sporangia upon the leaf or segment. A slide of the sorus from 

 the margin to the surface of a broadening leaf-area probably took place : also 



A B 





'CK-'C '- '' * '^/C 'CCV/ij ji' M 



'; 1-1 ' 



'Ouo'jKj 



^j:)i\'j.j 



Fig. 409. Danaea elliptica Smith. Drawings illustrating the partial septations of the sporangia. 

 A, tangential section through three sori, showing loculi in ground plan: the septa are often thin, 

 so that pairs of loculi are in close relation : the loculi marked ( x ) are large and show one or more 

 partial septa. ( x ^o.) B, C, D, E, show individual loculi with partial septa in greater detail. In 

 D and E it is difficult to decide w^hether the cells marked (?) will develop as tapetum or as spore- 

 mother-cells. ( X 150.) 



an extension of the sorus from a circular to an elongated form, so as finally 

 to occupy the whole space between the margin and the midrib: and some- 

 times fissions of the elongated sori into rosette-like secondary sori. The 

 synangial fusion of the closely related sporangia appears to have taken place as 

 a step not dependent upon anyone of these changes, and probably it happened 

 repeatedly, appearing as a secondary and biologically intelligible amendment 



