136 JAMES O. FOLEV. 



the fishes. In the Dipnoans and Amphibia the primary sperma- 

 tocytes are three times the size of the spermatogonia of the 

 last multiplication division. Turner ('19) finds that a growth 

 period does not occur in the perch. Geiser ('24) shows (Figs, n 

 to 21) a definite increase in size in the primary spermatocytes 

 of Gambnsia. In Umbra there is unquestionably a slight in- 

 crease in nuclear size over that of the preceeding stage. It is 

 probably due to the fact that the primary spermatocyte stage is 

 one in which the cells remain longer than they do in the multipli- 

 cation or secondary spermatocyte phases. All spermatocyte 

 stages may be equally rapid in the perch, and this may perhaps 

 be the reason Turner found no increase in nuclear size. 



Maturation has been infrequently studied in the fishes or the 

 Elasmobranchs. Moore ('95) has given a very good account of 

 the essential points in the maturation divisions of Scyllium 

 canicula, considering the technique employed by the workers of 

 his time. The method of reduction described differs in no 

 essential manner from that recognized in late years as the orderly 

 procedure in most animals and plants. After the conjugation of 

 the leptotene threads and formation of a synaptic knot (synizesis) 

 the chromosomes segment into long thick rods of chromatin 

 which pair and form a closed ring type of tetrad. The subsequent 

 separation of the parts of the tetrad is accomplished by a trans- 

 verse split through the center of the loop. The second division 

 is similar to the first, with the exception that there is no re- 

 duction, namely, the separation of the dyads by a transverse 

 split. The dyads separate and twelve univalent chromosomes 

 are distributed to each daughter cell or spermatic!. It is generally 

 believed now that what Moore describes as a "split" in the 

 ring is the pulling apart of homologous chromosomes at their 

 previous line of fusion. Rawitz ('99) working on Scyllium 

 states that at the time of reduction there is no reduction of the 

 chromatin number, but that there is a reduction of the chromatin 

 material by disintegration. He finds fourteen to sixteen chromo- 

 somes in the primary spermatocyte and fewer and larger chromo- 

 somes in the secondary spermatocyte. It may be that Rawitz 

 has confused the spermatogonial and spermatocyte divisions. 

 Geiser ('24) finds that in the early stages of diakenesis the 



