1905.] NATURAL SCIENCES OF PHILADELPHIA. 165 



Jcorper of the German writers) persists for a long while after the division 

 of the penultimate generation of spermatogonia, even up to the pro- 

 phases of the last spermatogonic mitosis. Accordingly the rest stage 

 of the last generation of spermatogonia (Plate IX, fig. 1) shows a dis- 

 tinct polarity of the cell-body, with a distal pole at which is the persisting 

 cell-plate, and a dark mass of idiozome substance, which appears to be 

 in part, at least, derived from the connective fibrils of the preceding 

 mitosis; and an opposite or central pole containing the nucleus. The 

 nucleus shows minute chromatin globules distributed in bead-like 

 chains along the linin fibrils, and also accumulated in larger masses. 

 With great regularity there is found also in each nucleus two or three 

 larger, somewhat irregular, deep-staining bodies ; whether they are 

 nucleoli or heterochromosomes cOuld not be decided by the use of the 

 iron hsematoxyline stain. 



The prophases of the last and penultimate spermatogonic mitoses 

 appear similar in character. The chromatin seems to arrange itself 

 into a continuous spirem, or, if not into one thread, certainly into but 

 a small number of very long threads. Plate IX, figs. 2-6 illustrate a 

 succession of the later prophases, and all show stages of segmentation of 

 the spirem. Fig. 2 shows a pair of minute centrosomes just external to 

 the idiozome body, and figs. 3 and 5 successive stages of the central 

 spindle ; the nuclear membrane commences to dissolve first in the vicin- 

 ity of the central spindle. The only point deserving particular comment 

 in the stages is a chromatin element, marked A". 2 in the figures, that is 

 found in every cell ; it is a portion of the chromatin spirem of smaller 

 diameter than the other segments, much more convoluted and in such a 

 manner as to represent a small corkscrew, and frequently appearing to 

 be enclosed within its peculiar membrane. It resembles in this respect 

 the accessory chromosome described by Sutton (1900) for the sper- 

 matogonia of Brachystola. It is an element that appears to be retarded 

 in its stages on comparison with the others— not condensing nor seg- 

 menting as rapidly as they do. When the nuclear membrane has com- 

 pletely dissolved away this single loop segments into two, which are 

 still to be distinguished from other chromosomes of the same length by 

 narrower diameter and more spiral form. These two chromosomes 

 resulting from the division of the single convoluted element are prob- 

 ably the heterochromosomes which become much better demonstrable 

 in the spermatocytes ; for the heterochromosomes of the spermatocytes 

 differ from the other chromosomes in their behavior, as will be shown, 

 and this pair in the spermatogonia behave at first differently from the 

 others. Because these heterochromosomes are demonstrable in such 



