x 74 



BOTANICAL GAZETTE [September 



of sperm cells (it is not clear whether he means body cells or male 

 cells) in Cupressus goweniana. 



It thus seems to be an established fact that more than two pro- 

 thallial cells and more than two sperms occur in widely unrelated 

 gymnosperms. Though we may all be willing, perhaps, to accept 

 this statement of facts, there is no such unanimity of opinion as to 

 the interpretation of them. It is obvious that one may adopt either 

 of two views: (i) One or two prothallial cells were found in the 

 primitive gymnosperms and in some cases these have divided to form 

 a complex. This would be Jeffrey's (5) "coenogenetic prolifera- 

 tion." In support of this view he has urged that a multiplication of 

 prothallial cells is correlated with " protosiphonogamic " fertilization 

 in the Araucarineae, " since the length of the pollen tube, in the 

 absence of any special conductive tissue, such as is found in angio- 

 sperms, calls for a greater development of prothallial tissue." This 

 line of reasoning could hardly be extended to Podocarpus or to 

 any other casenow known. If one could find some physical cause 

 applicable to all cases, he might believe that this multiplication of 

 cells is a thing of recent origin in each case. Furthermore, the idea 

 that a multiplication of cells would be forthcoming just when needed 

 is more teleological than accords with modern physiological teaching. 

 If we do not accept this explanation, as I think we cannot, we may 

 turn to the other possibility. (2) Primitive gymnosperms had a 

 multicellular prothallial tissue, and not unlikely a spermatogenous 

 complex of several cells. The evidence for this view is of two sorts, 

 historical and theoretical. Historically we know that the pollen 

 grains of certain Cordaitales (10) and Cycadofilicales (9, 10, 11) had 

 a considerable tissue of some sort. Whether it was spermatogenous 

 or prothallial is not very material, since this view assumes that both 

 sorts existed somewhere in the ancestry of living gymnosperms and 

 that by a gradual reduction each sort has been reduced to one or 

 two cells. It is of interest to note that all the heterosporous pterido- 

 phytes show the same tendency, each of the living genera having a 

 single prothallial cell and in some cases producing as few as four 

 sperms. This assumption makes it easy to account for the facts 

 already related. We have but to suppose that in these cases the 

 forms showing these peculiarities have merely retained their primitive 



