﻿Manchester Memoirs, Vol. xlix. (1905), No. 12. 21 



That no great weight can be attached to the difference 

 in symmetry, bilateral in Cordaiteae, radial, so far as we 

 know, in Pteridospermeae, is shown by comparison with 

 recent Cycadaceae, among which the seed is bilateral in 

 Cycas, but radially symmetrical as regards internal 

 structure in all the other genera of the order (Stopes, '04). 



In other respects there is a general agreement in the 

 organization of the seed between Cordaiteae and Pterido- 

 spermeae, as regards the vascular supply, the relation of 

 integument to nucellus, the formation of a well- 

 differentiated pollen-chamber containing multicellular 

 pollen-grains, and the presence of archegonia, instead of a 

 developed embryo in the ripe seed. It is difficult to 

 imagine so near a correspondence having risen indepen- 

 dently of relationship (Oliver, '03). 



The question whether the testa of the Cordaitean seed 

 represents a single or a double integument may best 

 be left open for the present. Precisely the same doubt 

 exists in the case of living Cycads, where Miss Stopes, who 

 has found reason to believe that the integument is double, 

 compares its outer layer to the cupule of Lagenostoma 

 Lomaxi (Stopes, '04). 



In the species of Cordaites investigated by M. Renault, 

 there is little direct evidence from the anatomical structure 

 indicating any connection with the Pteridospermeae, 

 though it must be borne in mind that the foliar structure, 

 as well as the organization of the seeds, in their close 

 agreement with corresponding features of Cycadaceae, 

 afford indirect evidence in that direction, for whatever 

 tends to connect the Cordaiteae with Cycads tends also to 

 connect them with the group from which we believe the 

 Cycads to have been derived. Comparative anatomy, 

 however, furnishes a further line of evidence. One of the 

 Palaeozoic genera best known from an anatomical point 



