STUDIES ON CHROMOSOMES 409 



part, each of them appears solid and undivided." The most that 

 can be said appears to be that "in the greater number of cases 

 there is to be seen in each geminus at least traces of a clear space 

 which marks the original point of meeting of two univalents" 

 ('11, p. 738). It seems to me that this is hardly a sufficient basis 

 for so important a conclusion. Many similar statements might 

 be cited from other authors. On the other hand some very com- 

 petent observers not only find no evidence of duality in the early 

 post-synaptic bivalents but definitely conclude that the con- 

 jugants completely fuse to form 'zygosomes' or 'mixochromosomes' 

 (e.g., Vejdovsky, '07, for the Enchytraidae, Bonnevie, '08, '11, 

 for Allium and other forms, Winiwarter and Saintmont, '09, for 

 the cat). 



In the case of Batracoseps I can fully confirm Janssens's state- 

 ment that no evidence of longitudinal duality can be seen in the 

 pachytene-loops throughout the greater part of the growth- 

 period, even in the most perfect preparations, and after various 

 modes of fixation, staining and extraction. It is only in the earlier 

 period that the duality appears, and then often only here and there 

 and in a small portion of the thread. It is the same in Tomop- 

 teris. The longitudinal cleft often so clearly seen at the time of 

 synapsis seems soon to disappear, so that for a time nothing can 

 be seen in the pachytene-threads to indicate their bivalent nature. 

 Theoretically it is of course quite possible that this appearance is 

 deceptive, and that the two elements are in reality always dis- 

 tinct; but if we resort to theory an equally strong case can be 

 made out, I think, in favor of partial or complete fusion. It 

 seems at any rate certain that in some of the most favorable 

 material thus far found among animals, synapsis is followed by a 

 union so intimate that no adequate evidence of duality is after- 

 wards seen until the diplotene stage is reached in the prophases 

 of the first division. It is very possible that this may be due in 

 some cases to a close twisting together of the threads (cf. p. 399) 

 but it would hardly be safe to accept such an explanation at 

 present. 



I am myself inclined to accept the evidence at its face value, and 

 to conclude that parasynapsis is followed by at least a partial 



THE JOURNAL OF EXPERIMENTAL ZOdLOOT, VOL. 13, NO. 3 



