Tlic (li'lcrniiiuil idii nï sex in miiiiKil (l('vi'lii|iiiii'iil,. 753 



Though the (lot(>rniiii:ition of sex be not clVected directly by cell-division, 

 there must be in every developmentiil history some point :it which the 

 sei)iirnti()n of the forerunneis of unlilu; «^nnietes is c.irried out by 

 mitosis. The phenomena of the reduction only complete the process 

 initiated at this. 



In the formation or dill'erentiation of two sorts of gametes from 

 the starting point of a prinniry gcM-m-cell, it may be insisted, it is pro- 

 bably mainly a matter of the inimber of mitoses from the start to the 

 linish. 'j\) put the i)oint dilferently, ap;irt from its dilferent destination, 

 l)ut i)robal)ly in some way correlated with this, in the divisions leading 

 to the formation of a male-egg (I*'ig. A), as already pointed out, theie 

 is obviously one more mitotic division than along the line resulting in 

 a female-egg. 'J1iis in the skate is evidenced by the lesser number 

 of primary germ-cells, formed by a male-egg, than by a female one. 

 But, as the primary germ-cells are of the like sizes in the two cases, 

 as obviously from this fact and from the number 51^^ of primary 

 germ-cells formed by the female-egg the missing division has been 

 made good in the cleavage of the latter, and th(>. like stiarting point 

 has been reached in both cases in the primary germ-cells. 



In addition, were an oocyte, destined to give rise to a male-egg, 

 to be turned aside; from the final phenomena of the rij)ening, and 

 were it instead thereof to undergo some additional mitoses, perhaps 

 only two, the result would be the production of spermatozoa, instead 

 of a male-egg. 



Similarly, it is probable, that in spermatogenesis with two (Cate- 

 gories of si)ermatozoa, as in Paludina^ the number of mitoses in some 

 way causally related to the kind of sperm resulting, is a dilferent one 

 along the line of the hairlike sperm (])r()bably a great.er one), than 

 along that of the wonidike one. And these two numbers again 

 may or may not be identical with those in the parts of the cycle 

 leading to the differentiation of two sorts of eggs. 



Sumniing up, the differentiation of a particular form 

 of gamete is, apart from other possible factors, pro- 

 bably a function of the number of mitotic divisions 

 from the starting ])oint of the primary germ -eel I s to 

 the ending at the oocytes and sp er ma to cy tes respec- 

 tively. 



