368 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY. 



bivalent structures, the primitive univalent chromosomes being repre- 

 sented by the spermatogonial chromomeres. On this assumption we 

 might interpret the constricted form of the somatic chromosomes as evi- 

 dence of a bivalent nature ; yet we must remember that such forms 

 occur in many other animals, where there is no apparent evidence of 

 double nature. Such an explanation does not seem unreasonable in view 

 of the conditions in Artemia, in which, as Brauer ('94) has shown, the 

 chromosomes, under certain conditions, regularly fuse in pairs. 



b. Numerical Reduction of the Chromosomes and Synapsis. 



The process of reduction in Coelenterata has been described by Bovei'i 

 and Downing. Boveri ('90) found that in the hydromedusan Tiara there 

 are fourteen tetrads in the first polar spindle ; fourteen dyads in the sec- 

 ond polar spindle, and fourteen univalent chi-omosomes in the egg nucleus. 

 That this is a reduced number was shown by the fact that there are 

 twenty-eight chromosomes in the first cleavage spindle. Downing 

 (:00, :05) has described a very different condition in Hydra. According 

 to his account there are forty-eight chromomeres in the spermatogonia. 

 These chromomeres unite in fours to form twelve chromosomes. In the 

 last spermatogonial division the daughter chromosomes are composed of 

 only two instead of four chromomeres, so each daughter spermatocyte 

 receives only twenty-four of these bodies, instead of forty-eight like 

 its parent cell. In the first maturation division these chromomeres 

 pair to form twelve chromosomes, which divide longitudinally. In the 

 second maturation mitosis the division of the chromosomes is transverse, 

 each spermatid receiving six chromosomes. According to this account 

 Hydra exhibits an extraordinary condition, for w r hile the number of 

 chromosomes is reduced to one-half, the number of chromomeres is re- 

 duced to one-fourth. But neither Downing's description nor his figures 

 seem altogether conclusive. 



Guenther (:04) has traced the history of the chromatin in the sperma- 

 tocytes of another species of Hydra with very different results. He finds 

 that the chromatin instead of persisting as chromomeres, becomes dif- 

 fused in the resting stage of the primary spermatocyte. The nuclear con- 

 tents then contract into a mass containing all the chromatin of the cell ; 

 when this mass expands once more, the reduced number of chromosomes 

 emerge from it. 



Guenther's figures resemble so closely the contraction phases in the 

 spermatocytes of Gonionemus that I have no doubt he was dealing with 



