296 BOTANICAL GAZETTE [aprii 



2. " Prothallial cells" 



Several species have been described in which the '"prothallial" 

 cell has the power of division. Among the number are Ginkgo 

 biloba, by Strasburger (i); Picea excelsa, by Miyake (2) and 

 Pollock (3); Abies balsamea (9); Agathis, by Jeffrey and 

 Chrysler (14) ; Podocarpns, by Coker (13a) ; and Dacrydium 

 by Miss Young (16). The similarity of the generative cell and 

 the prothallial cells is pointed out by Miss Young: "In Dacrydium, 

 as in Podocarpns and Abietineae, a third cell is cut off from the 

 main body of the spore. It overlies the others and is so similar 

 to them that, but for its subsequent behavior, one might think it a 

 prothallial cell. It is the generative cell, generative in the sense 

 that it is the ancestor of the spermatogenous cell. This and the 

 second prothallial cell now divide." The first prothallial cell may 

 also divide. Again: "at shedding the pollen grain contains the 

 body cell and five free nuclei. The nucleus of the body cell is in- 

 distinguishable from those of the prothallial cells and the tube 

 nucleus." Pollock (3) states that in a large proportion of the 

 gametophytes of Picea excelsa there is only one prothallial cell. 

 Burlingame (20) says : "in Podocarpns one primary prothallial 

 cell may be cut off, after which the free nucleus divides to form the 

 free spermatogenous cell and the tube nucleus; or two primary 

 cells may be cut off before the tube nucleus is separated from the 

 primary spermatogenous cell." In these two species, as well as in 

 Picea canadensis, the antheridial function is not limited to a defi- 

 nite primary derivative. It has been established that "prothallial 

 cells" and generative cells may be similar in appearance; that 

 frequently they are similar in their power of division; the simi- 

 larity is further emphasized by the presence of "prothallial cells" 

 in the pollen tube. The present account has emphasized the 

 similarity in the origin, and has shown that potentially there is 

 a similarity in function; that any one or sometimes two of the 

 primary derivatives may be antheridial. To what extent we 

 are justified in suggesting that these phenomena are indicative 

 of a multi-antheridial ancestral form only further research can 

 determine. 



