The Spermatogenesis in Pentatoma up to the Formation of the Spermatid. 43 



number of figures are given in illustration. They vary in size, and 

 are frequently so thick as to include only a small space. Most 

 frequently they do not appear as quite closed rings but either show 

 a constriction at one point (Figs. 130, 136, 143), or show a break 

 at one point (Figs. 129, 136, 139). Not infrequently the rings are 

 elongated (Figs. 125, 145). Now between such an incompletely closed 

 ring and the more usual V- or U-shaped elements all intermediate 

 forms are to be found, which speaks decidedly for the conclusion that 

 the rings are chromosomes which differ from the others in having 

 their ends in contact. The strict proof for this view is brought out 

 by counting all the chromosomes in a nucleus: the number of the 

 rings plus the number of the other elements is almost always 7, so 

 that each of them has the value of any of the other chromosomes 

 (Figs. 125, 126, 132, 139, 143; of the Figs. 124—149 all represent 

 entire nuclei except 135, 142 and 146). Thus the facts show that in 

 almost all cases each ring has the value of a single chromosome, 

 comparable to the other chromosomes of the spermatocytes, and hence 

 (as will be shown more fully) are bivalent ; they differ from the others 

 only in having their ends joined together. The difference in form 

 between them and the rod-shaped chromosomes is striking, but inter- 

 mediate in form are the U- and V-shaped elements. In process of 

 formation they are undoubtedly derived from those chromosomes of 

 the dense spirem stage, which had a circular shape, with the ends 

 close together or apposed (Fig. 122). In only very few cases have 

 the rings the value of more than one chromosome each. In one 

 nucleus (Fig. 146) we find a ring formed by two, small, transversely- 

 constricted chromosomes, lying parallel to one another; since there 

 were only 5 other chromosomes in this nucleus, this ring has the 

 value of 2 chromosomes. A similar relation was found also in two 

 other nuclei (Figs. 136, 146). In a single case (Fig. 147) a ring was 

 formed of 4 well- demarcated chromosomes, and there were only 3 

 others in the nucleus. These were the only cases observed by me, 

 where a ring had the value of more than 1 chromosome. Even in 

 these atypical cases the chromosomes are only temporarily in contact 

 with one another, since in the equatorial plate of the 1st spermato- 

 cytic division (PL 4, Figs. 170, 171, 176 — 189) 7 chromosomes always 

 occur without exception, and never unions of two or more. At the 

 commencement of the loose spirem, before the definitive number of 

 chromosomes is produced, the sum total of a chromatin ring plus the 

 other elements may be less than 7, so that by observation of such 



