152 T. H. MONTGOMERY, JR. 



tion to generation. And I could also prove that in all cases the 

 two components of each bivalent chromosome always become 

 separated from each other in the first maturation mitosis. 



Following this came the paper of Sutton (/. c.\ proving con- 

 clusively that in the spermatogonium of Bracliystola the chromo- 

 somes occur regularly in pairs of graduated sizes, the two of a 

 pair being always of the same length ; that in the synapsis stage 

 bivalent chromosomes are produced by the conjugation of every 

 two chromosomes of the same length ; and that corresponding 

 chromosomes became separated from each other in the reduction 

 division (here the second maturation mitosis). So he concluded 

 quite rightly that there are two series of chromosomes in the sper- 

 matogonium, a paternal series, A, B, C . . . n, and a maternal 

 series, a,b,c . . . n, in which A corresponds to a in size and hered- 

 itary value, B to b, and so through the series. By A joining with a 

 in the synapsis, like chromosomes conjugate ; and by these sep- 

 arating from each other in the reduction divisions, it results that 

 two chromosomes of like size are not found in the spermatid. 

 Sutton was the first to demonstrate this for the whole series of 

 chromosomes, and to argue that such a conjugation, together 

 with the following reduction division, would operate so that no 

 spermatid could receive two chromosomes of like hereditary 

 value, but only one chromosome representing a particular value. 



In the present paper I have shown that in Plctliodon and Dcs- 

 mognathus also one may recognize the two corresponding series 

 of chromosomes in the spermatogonium. An examination of 

 Ascaris megalocepliala bivalens shows the same relation. Pole 

 views of the first cleavage spindle (Figs. 28-30) show each two 

 larger (A, a) and two smaller chromosomes (B, b}. The differ- 

 ences in size of the two pairs is not very great, but always recog- 

 nizable when the chromosomes can be seen in their entirety. 

 This is then evidently a case parallel to those described above : 

 that of the larger pair (A, a] one is paternal and the other ma- 

 ternal, and that of the smaller pair (B, b] the same relation holds. 

 Now the formation of the tetrads in the ovogenesis of this species 

 has been described by Boveri as two equation (longitudinal) di- 

 visions of each bivalent chromosome ; and Braur has reached the 

 same result for the formation of the tetrads in the spermatogen- 



