SUMMARY AND CONCLUSION 2O/ 



case the primary halving in number is a pseudo-reduction, and each 

 tetrad arises by one longitudinal and one transverse division of 

 a bivalent chromosome, representing two different regions of the 

 spireme-thread (Hacker, vom Rath, Riickert, Weismann). In the 

 other case* the primary halving appears to be an actual reduction, 

 and if tetrads are formed, they arise (Ascaris) by a double longitudi- 

 nal splitting of the primary rod, and all of its four derivatives repre- 

 sent the same region of the spireme-thread. Since the latter consists 

 primarily of a single series of granules (" ids " of Weismann, or 

 chromomeres), by the fission of which the splitting takes place, the 

 difference between the two views comes to this : that in the second 

 case the four chromosomes of each tetrad must represent identical 

 groups of granules, while in the first case they represent two differ- 

 ent groups (Fig. 102). In the second case the maturation-divisions 

 cannot cause a reduction in the number of different kinds of ids. 

 In the first case the number of ids is reduced to one-half by the 

 second division by which the second polar body is formed, or 

 by which two spermatids arise from the daughter-spermatocyte 

 (Riickert, Hacker, vom Rath). 



The first view must obviously stand or fall with the conception of 

 the primary chromatin-rods as bivalent chromosomes. That this is 

 a valid conception is in my judgment demonstrated by Brauer's 

 remarkable observations on Artemia ; for in this case it is impossi- 

 ble to escape the conclusion that the "chromosomes" of those 

 parthenogenetic embryos in which the number is halved are bivalent, 

 i.e. have the value of two chromosomes united by their ends, 

 and they lend the strongest support to vom Rath's and Hacker's 

 hypothesis. For if the number of chromosomes be merely the 

 expression of a formative tendency, like the power of crystalliza- 

 tion, inherent in each specific kind of chromatin, why should the 

 chromatin of the same animal differ in the two cases though derived 

 from tJie same source in both ? Yet if the cleavage-nucleus arises 

 from eighty-four dyads the same number of chromatin-rods appears 

 in all later stages ; whereas if the dyads break each into two separate 

 chromosomes before their union, the number is thenceforward one 

 hundred and sixty-eight. So great is the force of this evidence that 

 I think we must still hesitate to accept the results thus far attained 

 in Ascaris and the plants, and must await further research in this 

 direction. Until the contradiction is cleared up the problem of 

 reduction remains unsolved. 



