514 Wisccnsin Academy of Sciences, Arts, and Letters. 



cleavaffe bv means of surface constrictions. In c-ell division 

 in Cladoplwra, conidia formation in the mildews, and the 

 cutting off of the gametes in Sporodinia Ave have simple con- 

 striction, as Harper lias pointed out. All of these cases may 

 be c^rouped together as examples of cell division l\v constric- 

 tion and thereby be clearly distinguished from the case of tbe 

 cormophytes where cell division is accomplished by means of 

 the formation of a cell plate. 



These two types of cell division are most sharply differ- 

 entiated when compared as to their relation to tlie nuclei. In 

 the case of division bv constriction there is no visible relation 

 of tlie nuclei to the process so far as the mechanism involved 

 is concerned, while in the case of cell plate formation the nu- 

 clei are apparently the active controlling centei's for the pro- 

 cess bv meaiis of tbe spindle fibres, manv of which, and some- 

 times all, may be formed for the express pui'j:)ose of cell 

 division. 



The relation of the processes of cell division in certain 

 brown Algae t-o the two types al)ove described is not at all 

 clear. In all of these forms so far described division is 

 said to take place by means of a cell plate that is formeil A\dth- 

 out being directly connected with the nuclei. (See 9, 17, 24, 

 26.) Owing to the difficulty of getting numerous stages of 

 division in these forms it is quite possible that some of tlie 

 most important phases have been missed. 



The structure of the swarm spore of Ilydrodictyoii com- 

 pared t/O the structure of other ciliated cells presents many 

 interesting problems. The most thoroughly investigated 

 cases of the development of such cells is that of the spermato- 

 zoa of the higher animals and the antherozoids of some Pteri- 

 dophytes and Gymxnosperms. While there are some differ- 

 ences as to detail, zoologists seem to be in accord as to the main 

 facts concerning tbe development of tlie spermatozoon (See, 

 Wilson, 36.) The axial filament of the tail grows out from a 

 spermatid centrosome tliat may persist as a distinct body ("end 

 knob") at the base of the fibre or may enlarge to form a part of 

 the middle piece. This fact of the growth of the axial filament 

 from tlie centrosome forms a possible basis of comparison be- 



