1905.] XATURAL SCIENCES OF PHILADELPHIA. 177 



and its only perceptible change is a longitudinal split of each univalent 

 component. This is shown in fig. 446ts, A. and B, which show merely 

 the bivalent heterochromosome and an arc of the contiguous nuclear 

 membrane. In most cases the space of this split is widest at the inner 

 ends of the univalent components of this bivalent heterochromosome, 

 as shown in fig. A. This longitudinal split cannot be seen from every 

 point of vision, but only when the heterochromosome lies in particular 

 directions, as is quite understandable. 



Following the synapsis is a post-synapsis stage, with the bivalent 

 chromosomes more evenly distributed through the nucleus (figs. 45, 

 46). The longitudinal split is wide and very evident, but does not 

 extend through the distal ends of the still generally V-shaped loops. 

 In each bivalent chromosome the angle of the V (a; of the figures) is 

 the point of junction of two univalent parts. Here also the longitudi- 

 nal split of the bivalent heterochromosome (A^. 2) can sometimes be 

 seen. No nucleolus is formed in any part of the growth period. 



Immediately succeeding are the prophases of the maturation mitoses 

 (figs. 47-52). There occurs in them a gradual shortening and conden- 

 sation of the chromosomes, leading to narrowing or even complete 

 temporary obliteration of the longitudinal split. This split in the early 

 prophases (figs. 47-49) in the case of some of the chromosomes becomes 

 a little wider than during the post-synapsis (figs. 45, 46) ; so with the 

 chromosomes marked H in these figures. But this happens with only 

 a minority of the chromosomes in any nucleus. And it is not a definite 

 stage in the structural change of every chromosome, for the reason of 

 its relative infrequency. Most of the chromosomes, on the contrary, 

 are straight or bent rods, and the angle or middle point of these marks 

 the point of conjunction of two univalent chromosomes (x of the figs. 

 47-52). Such chromosomes as those marked H in figs. 47 and 49 are 

 ones where the longitudinal split has become very wide at the point of 

 union of the two univalent chromosomes ; but even in such chromosomes 

 one axis always remains longer than the other, so that there is no evi- 

 dence of a bivalent chromosome becoming extended out in a hne at 

 right angles to its previous long axis. And even for these chromo- 

 somes, as is clearly the case with the others where there is no extensive 

 widening of the longitudinal split, the successive prophases lead toward 

 a narrowing or closure of this split. Regular rings appear not to be 

 formed. But chromosomes in the form of an X are not infrequent. 

 There is no difficulty in the interpretation of the form of these. For in 

 the one shown in fig. 51, marked D., the X is seen to be two univalent 

 chromosomes, each longitudinally split, joined by their middle points; 

 12 



