WENRICH: SPERMATOGENESIS OF PHRYNOTETTIX MAGNUS. td 
not a simultaneous process for all the chromosome-pairs, but that 
it is a gradual process, some conjugating earlier than others. Just 
how the members of the different pairs are enabled to select their 
mates is a very puzzling question, but probably the stretching and 
orientation of the threads as shown in figure 29 might facilitate this 
process. That some of the pairs conjugate quite early, is shown in 
figure 30, where it may be seen that in selected pair B conjugation is 
complete. In figure 27, which is of a very much earlier stage, there 
are to be seen two of the still hazily defined threads lying side-by-side. 
They are similar enough in their constitution to be regarded as the 
two members of a pair, and it would not be surprising if conjugation 
should begin at a stage as early as this. 
As an additional detail it should be pointed out that the bead-like 
granules which are strung along the threads of the leptotene and 
zygotene stages are not always of exactly equal size in the two con- 
jugating elements. In figure 58 (Plate 5) the example of chromo- 
some-pair B well illustrates the disparity in size between the two 
members of some of the pairs of granules. This condition may well 
answer the criticisms of those who hold that the accuracy with which 
the granules are paired could be accounted for only on the assump- 
tion that they arose by a splitting of single granules into equal parts. 
I am able to show in this case that the members of each pair of granules 
are not always of equal size. 
3. The pachytene stages.— Figures 32 and 33 (Plate 3) are of early 
pachytene stages. It sometimes happens that even at such stages 
there may remain one or two pairs of threads that are not fully conju- 
gated, though I have not added a drawing of such a condition. In the 
case of some of the pachytene threads of figure 32, complete loops have 
been formed, both ends being attached at the polar region. The 
formation of such loops is not necessarily the rule, however, as has 
been indicated already in connection with the spireme loops of the 
selected chromosome-pairs (fig. 56-61, Plate 5). In figure 33 a scat- 
tered arrangement of the polar granules is to be seen, though they have 
coalesced to form several composite granules. Figure 34, of a later 
_ pachytene, exhibits one of the large composite granules. Figures 
35-37 indicate how the composite granules break up into their com- 
ponent polar granules. A comparison of the examples of chromo- 
some-pair B in figures 30 and 35 will indicate the extent of the process 
of gradual shortening which takes place during the pachytene stages. 
It will be noticed that the line of separation between the threads 
which have conjugated (7. e., the primary longitudinal split) remains 
visible throughout the pachytene stages. 
