82 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY. 
greater or less degree. The tendency for adjacent granules to fuse 
probably accounts for some of the variations to be noticed. If one 
will compare in order the examples J, m, a, and c, (fig. 98) the different 
steps in the fusion of granule no. 4 with the smaller, yet prominent, 
granule close to it will be seen. As the threads shorten during the 
later pachytene and postspireme stages, this coalescence of adjacent 
granules becomes more noticeable and the individual granules all 
finally lose their visible identity in the compact metaphase chromo- 
somes. 
It will be observed that the members of a pair of granules may also 
appear to be fused together into a single mass. An example of this is 
seen in figure 98, n, granules 4 and 5. This fusion must be very 
temporary in character, since it is not the general rule, and since the 
granules separate again in the postspireme stages, as shown in figure 
63 (Plate 6); yet so close an association of these granules apparently 
offers opportunity for the exchange of chemical substances between 
them. In the case of the proximal granules (no. 1), the members may 
not only fuse with each other but, as previously noted, characteristi- 
cally unite with the polar granules of other chromosome-pairs to form 
the composite granules. The association is fully as close as in that of 
any single pair, for frequently all traces of the outline of individual 
granules is entirely lost, as, for example, in figure 34 (Plate 3). Al- 
though the individual granules separate out again in the postspireme 
stages, if we admit that there is an exchange of chemical substances 
between members of a single pair of granules, I think we must also 
assume it for the polar granules of the different chromosome-pairs. 
At j (fig. 98, Plate 8) may be seen another example of an expanded 
polar granule, such as has already been mentioned. The possible 
significance of this peculiarity will be discussed on page 112. 
It will be instructive to compare the members of particular pairs 
of granules. Figure 58 (Plate 5), as already mentioned, represents a 
zygotene stage. The paired chromatic threads near the middle have 
just begun to conjugate, while in the case of chromosome-pair B, in 
the left half of the figure, the two conjugating threads have only 
recently come to lie side-by-side, for the members of the different pairs 
of granules are yet distinct. This condition fortunately gives us an 
opportunity to compare the relative sizes of the members of each pair. 
On examination it will be seen that the members of the pair numbered 
4 are not equal in size. This is also true for the pair numbered 3. In 
the case of number 4, the disparity in size between the two granules 
is considerable, and it is interesting to observe that this difference in 
