100 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY. 
the spermatogonial chromosomes had failed to pair. He describes 
two methods of rmg-formation. The correctness of his conclusions 
as to the succession of stages in some of his series might be questioned 
on the ground that they are not different stages of the same chromo- 
some. The series shown in his figures 2 to 11 probably represents a 
normal method of ring-formation, viz., by the opening out of a para- 
synaptic spireme segment along one of the longitudinal splits, with 
the ends remaining in contact. The series in figures 12 to 16 might 
also easily be derived from a parasynaptic segment. It is extremely 
questionable whether the figures in the series 18-20 are arranged by 
the author in their natural sequence; the reverse order is more likely 
to be the correct one. His figures 19 to 27 (Plate 2) doubtless repre- 
sent different shapes of the same chromosome-pair, the so-called 
“middle granule”’ serving to identify the element. I would suggest, 
however, that the stage that he represents in figure 19 may have 
resulted from an opening out of a parasynaptic segment in the same 
way that I have described for chromosome-pair A, in which case the 
“middle granules”’ would be polar granules instead of “middle” ones. 
In view of the rather far-reaching conclusions that Robertson (’15) 
has drawn from his work on the Tettigidae, I would call attention to 
some differences, as well as similarities, between his work and mine. 
In the first place, he describes parasynapsis in the early stages of the 
growth-period, as I have done, but in the postspireme stages he as- 
sumes that the conjugants separate along the plane of conjugation 
(primary longitudinal split), the separation beginning at the proximal 
end. He shows in the metaphase of the first spermatocyte most of 
the chromosomes as elongated rods with appearance and orientation 
similar to that seen for my tetrad A. It will be remembered that in 
the latter case the separation from the proximal end of the spireme 
segment toward the distal end is not along the plane of the primary 
split, but along the plane of the secondary split. I believe that 
Robertson may have overlooked a similar behavior in the chromo- 
somes of his material. Curiously enough the unequal elements that 
he describes are very similar to the unequal type of chromosome B 
and of chromosome C, in Phrynotettix. I have shown that the 
behavior of the chromatids in B and C is similar to that in A. And 
that in the cases in which C divides reductionally in the first division 
the spindle-fiber attachment is necessarily shifted to the distal ends. 
I believe that such a condition probably occurs in the unequal tetrads 
described by Robertson, and if so, theoretical explanations, as to 
how the elements came to be related to each other in the way they are, 
