Thorn — Fertilization in Aspidium and Adianticm. 291 



appears on the opposite edge of the band in the third turn of 

 the spiral. 



In considering the function of the sperraatozoid and its 

 various parts, it is necessary to go back and review to some 

 extent at least its development. The antheridium contains a 

 number of " spermatozoid mother cells " (<* spermatogenous 

 cells " or spermatids, if we are to use the same term as the 

 zoologist), varying from 32 to probably 128. The 

 prothallia differ in the number produced, and often antheridia 

 on the same prothallium seem to contain different 

 numbers. The number probably varies inside these lim- 

 its, however. Belajeff observes that the spermatogenous 

 cell contains a large nucleus which comes to lie upon one 

 side of the cell (fig. 28) close under the wall. At 

 this time, according to his statement which he makes general 

 for the Filicineae, there appears in the cytoplasm a round 

 granule or body staining more deeply than ordinary cyto- 

 plasm and sharply marked off from it. He describes this 

 body under the name of "Nebenkern." Shaw observes and 

 describes the same body under the name of " blepharoplast " 

 and finds it present in the mother cells of the spermatids in 

 Marsilia and Onoclea. That is, according to Shaw, the 

 blepharoplast in these forms must divide with the division 

 of the cells at least once before it assumes its final 

 form. This is unquestionably the same body which was de- 

 scribed by Webber for Zamia as " blepharoplast " and has 

 since been decribed by Ikeno for Cycas as a centrosome. In 

 my material, at a stage just preceding this, the nucleus con- 

 tains a network bearing large chromatin granules which are 

 disposed, for the most part at least, close under the nuclear 

 membrane. There are from one to three larg-e nucleoli which 

 stain a clear red with safranin while the chromatin takes a 

 deep blue from the gentian violet (fig. 27). Comparing 

 this figure, where changes have already begun, with earlier 

 stages, there seems to be a rapid increase in nucleolar mate- 

 rial. In many cases the nucleoli appear to have just divided. 

 All of the cells of the antheridium appear to be in nearly the 

 same stage, and there is a great regularity in the successive 

 changes as they occur in cells of the same group. This regu- 



