24 ARTHUR BOLLES LEE 
They ultimately all fuse, apparently, into a single homogeneous 
ring, as shown in figs. 49, 50, and others. 
As soon as these globules have attaimed a certain size, 
figs. 48, 44, 45, 49, and less clearly yet still indubitably in 
figs. 86, 87, 38, the chromosomes, which in the clump appear 
as straight staves, now appear as more or less sharply 
curved staves, set on the surface of the globules or ring, 
that is, outside them and not embedded in them, see 
particularly the profile views figs. 48, 44, 45. Their outer 
surface is irregularly convex; but their inner surface is 
flattened on to the curvature of the globule or rmg. They are 
at the stage we are considering—of a length equal to about 
that of one of the limbs of the V-shaped chromosomes of the 
anaphase (see figs. 8, 4, 17, 61). They do not form complete 
hoops round the ring, but ares that embrace about half a 
meridian of it. They thus show two ends, a polar end and an 
antipolar end. The polar ends, abutting on the lumen of the 
ring, are generally closely huddled together and sharply curved 
downwards, so that it is impossible to get clear images of them. 
But their antipolar ends are generally widely spaced (figs. 48, 44, 
45), and here their two component threads may frequently be 
seen, with certainty, to be widely divaricated, figs. 43 (in the 
middle), 44, 45, which is not the case with the polar ends. 
As soon as the process of expansion has set in, the images of 
the clump become less indistinct, and the chromosome staves 
appear as shown in figs. 80, 38, 35, 36, 87; that is, they are seen 
with certainty to contain or consist of the thin chromatic 
threads running in pairs, which in our study of the clump in 
Part I we recognized by their structure as shrunken chromo- 
some axes, without discussing the fact of their collocation im 
pairs. The members of these pairs run very close together and 
in the main parallel to one another, as shown in figs. 80 to 85. 
Images such as these may suggest, strongly, that durmg the 
earlier stages of the clump the chromosomes have contracted 
into short staves, each of which has undergone a longitudinal 
division ; so that the threads would be the cleavage products 
of such a division. Now there is no sign of any such division 
