134 JAMES O. FOLEY. 



about three times the size of the final spermatogonia and in 

 Gambusia, eight times the size of the final spermatogonia. It 

 was found in Umbra that they are from three to six times the 

 size of the final spermatogonia. Agar ('11) differentiates be- 

 tween primary and secondary spermatogonia in the Dipnoan, 

 Lepidosiren. He finds that there is a decrease in nuclear size 

 from thirty micra to seventeen micra during the transition from 

 primary 7 to secondary spermatogonia. 



In the Dipnoi the final spermatogonia are only about one third 

 the size of the primary spermatocytes. In this respect the 

 Dipnoi appear to resemble the Amphibia more than they do 

 the Teleostomi. It would be interesting to know what the con- 

 dition is in the Elasmobranchs and Cyclostomes. The writer 

 believes that this early growth phenomenon in the early spermato- 

 gonia in Umbra and other teleosts is bound up with the seasonal 

 cycle in these forms; namely, the formation of spermatozoa 

 from all the spermatogonia, the absence of reserve germ-cells, 

 the destruction of the cysts of the lobule at the end of spermato- 

 genesis together with their reformation the following season, and 

 the formation of the early spermatogonia from migrating cells. 



The number of generations these early spermatogonia go 

 through before the final spermatogonia are produced varies in 

 different fishes. Turner ('19) finds, in the perch, that there are 

 five or six generations, Geiser ('24), in Gambusia ten to twelve. 

 In Umbra, at least six generations are passed through before the 

 final spermatogonia are formed. 



Chromosome numbers have been given for the following species: 



Mak 1 . Female. 



Perca flavascens (Turner, '19) 27 



Gambusia holbrooki (Geiser, '24) 35 or 36 



Lebistes reliculatus (Winge, '22) 46 46 



Lepidosiren paradoxa (Agar, "n) 38 



Scylliiim canicula (Moore, '95) 24 



Scyllium canicula (Rawitz, '99) 20 to 24 



Umbra limi 22 



The behavior of the spermatogonial chromosome^ in the fishes 

 has been stressed little. Agar ('11) ('12) gors into the- matter 

 in Lepidosiren in some detail. lie finds, as I do, that the 

 chromosomes of the newly formed equatorial plates are long and 



