XII.] CONJUGATION AND SEXUAL REPRODUCTION. 153 



But this latter process renders two ' reducing divisions ' neces- 

 sary, — that is if the normal number of idants must be reduced 

 to one half. 



That there is no such reduction in regular parthenogenesis 

 may be inferred from the large number of idants present in the 

 parthenogenetic eggs of Artemia salina, viz. twenty-four or 

 twenty-six. If a diminution to one half of the original number 

 of idants normal for the species took place at each maturation, 

 it is obvious that in each successive generation the idants would 

 be reduced to half, and we should at the present day find only a 

 single one left in Artemia. Either this polar division is not a 

 ' reducing division,' or it is preceded by a doubling of the number 

 of idants, just as in ova which require fertilization. 



If this latter be true, it follows that in parthenogenesis we 

 meet with a simple retention of the first of the two polar divi- 

 sions which occur in other ova. 



It is unfortunate that direct observation has not hitherto led 

 to an entirely certain decision upon the point. Dr. Otto vom 

 Rath has had the great kindness to examine, with this object 

 in view, many of my old sections ^ of the parthenogenetic ova 

 of Artemia salina, in order to find out those parts of them 

 which were most important in this respect. From my earlier 

 researches, conducted upon the same material, I was already 

 aware that the germinal vesicle, after having approached the 

 surface, contains a large number of chromatin granules, which 

 are distributed with almost complete regularity. It was evident 

 that these granules had not yet become the definite chromato- 

 somes or idants, but that they were smaller and more numerous 

 (Fig. IX. i). In one germinal vesicle I counted 115 of them ; 

 in another, which was already changing into a spindle, I also 

 found 115, all lying in the equatorial plane (Fig. IX. 2) ; in a 

 third, 77 ; in a fourth, 70 ; and in a fifth, 57. Now in the equa- 

 torial plate of the polar spindle, from 48 to 52 spherical idants 

 are always arranged in a double wreath (Fig. IX. 3 a). These 

 must therefore have arisen from the fusion of several of the 

 primary chromatin granules, and the great variation in the 

 number of the latter must depend on the fact that the fusion was 



^ Weismann und Ischikawa, ' Weitere Untersuchungen zum Zahlen- 

 gesetz der RichtungskOrper ; ' Zoologische Jahrbucher, Bd. III. p. 575, 

 1888. 



