Some problems of reproduction. 599 



whose filaments are composed of a single row of cells^ multi- 

 plying by transverse fission, at a certain moment tlie cells 

 conjugate two and two : the cells that unite are apparently 

 sister-cells. But this statement needs to be modified : in either 

 cell the nucleus divides by mitosis, and two cousin-nuclei 

 fuse, while the other two are cut off with a trace of cytoplasm 

 as "polar bodies.^' In most Desmids two adult cells approach 

 to pair, but either divides into two, which fuse respectively 

 with those formed from the other original mate (progamete); 

 so that the actual pairing-cells are not those that approached 

 one another in actual cellular life, but their daughter cells 

 In the Conjugate Spyrogyra the cells fuse by the outgrowth 

 of tubes that meet and anastomose ; but the nuclei long 

 remain merely approximated without fusion in the restiug- 

 zygospore. It would seem from the results of Chmielewsky 

 that either nucleus at the approach of germination undergoes 

 fission to form a pairing-nucleus and one that aborts, and 

 that it is these daughters of the distinct nuclei of the original 

 pairing-cells that actually fuse at last as gameto-nuclei. 



In Diatoms the cells that approach may either divide so as 

 to form two pairs of gametes, as in Desmids, or they may 

 themselves apparently pair; but the process of pairing is 

 only comj^eted after the nuclei have divided once or twice, 

 only one of the daughter-nuclei in either mate being func- 

 tional, and the rest abortive. Similar divisions produce the 

 pairing-nuclei of the Infusoria, in which the mates (progo- 

 metes) are also adult. 



In other Fungi than Basidiobolus, whose isogamous 

 syngamy has been best studied, the Uredinege and the Ustila- 

 ginese (the Husts and Smuts) the gameto-nuclei are sister- 

 nuclei of the same undivided cell; but the Basidioinycetes 

 afford so far no support for my present thesis, for the origin 

 of the several ( — 7) nuclei that fuse to form the nucleus of 

 the ba^idium is not known to be from recent mitosis. 



In cases of unequal fusion the sperms have usually to be 

 produced loiig in advance and strongly differentiated, so 

 that they are indeed incapable of fission, and consequently 



