OBSERVATIONS ON THE MATURATION PROCESSES. 39 



demonstrating the fibers and circumpolar bodies clearly. As a rule, 

 the individual fibers and their attachment to the chromosomes are not 

 easily distinguishable. Although one can not be absolutely certain that 

 there are fibers which are continuous from pole to pole without being 

 connected to any of the chromosomes, it is perhaps reasonable to assume 

 that such is the case, because of the general similarity of the second 

 spindle to the first one, where such a condition is fairly evident. The 

 daughter chromosomes, after their migration toward the poles of the 

 spindle, are connected by interzonal filaments (figs. 29, 30). " Zwi- 

 schenkorperchen " form midway between the ends of the filaments, as 

 described for the first spindle (p. 32), and later fuse into a cell plate. 



The second spindles do not differ from one another much in size, 

 nor do their dimensions, on the average, change appreciably with pro- 

 longed existence due to the absence of semination. This constancy in 

 size is shown by a comparison of spindles from two groups of eggs : one 

 group composed of eggs which have been but a short time in the oviduct 

 (taken not later than i6j hours after parturition), the other of eggs taken 

 from the oviduct 29 or more hours after parturition. Because of the unfa- 

 vorable position of many spindles, measurements of only 30 young and 26 

 old ones could be used. The average dimensions for the young spindles 

 are: length 17.9 micra, diameter 7.2 micra; for the old spindles: length 

 17.5, diameter 7.3 micra. A comparison of these averages with those of 

 the mature, or nearly mature, first spindles in Stages IVa and IVb (viz., 

 19.2 X 10.8 micra, and 22.4 X 9.9 micra, respectively) proves that the 

 second maturation spindle is somewhat smaller than the first. 



4. CENTROSOMES, CIRCUMPOLAR BODIES, AND CLEAR REGION. 



For the second spindle, as for the first, the existence of typical 

 centrosomes is highly improbable. However, there are at certain times 

 structures which to some extent resemble centrosomes. 



The circumpolar bodies (figs. 22, 24) correspond exactly in posi- 

 tion, abundance, and general appearance to those of the first spindle. 

 When the spindle is first fully formed they are already present, and per- 

 sist for some time. But with spindles which, in the absence of semina- 

 tion, persist for a long time they have a tendency to dwindle away, 

 sometimes, however, leaving a few granules at the poles where centro- 

 somes might be expected (figs. 25, 26, 27). These statements are based 

 upon a comparison of the eggs used in calculating the size of the spin- 

 dle. The eggs in the oviduct (and a few in the periovarial space and 

 in the ovary) taken from mice killed 16^ hours or less p.p. show in most 

 instances well-developed circumpolar bodies, whereas most of the eggs 

 from animals killed 29 or more hours p.p. show very few or none of them. 

 These bodies also disappear following normal metakinesis induced by 

 semination after the chromosomes have migrated and become confluent 

 (figs. 29a, 30), as in the case of the first spindle. 



