The spermatogenesis of Peripatus (Peripatopsis) balfouri. 301 



be seen in their entirety, partly because some of the chromosomes lie 

 over and hide others, partly because on such views only a few of the 

 chromosomes lie exactly in the plane of the section. Accordingly in 

 none of these views are the whole number, 14, of bivalent chromo- 

 somes shown. On cross sections of such nuclei (Figs. 85—87) each 

 of the 28 univalent chromosomes is more or less transversely cut, 

 but on such sections necessarily their union into bivalent chromo- 

 somes cannot be seen. My method of determining the number of 

 bivalent chromosomes, accordingly, was to count on cross section the 

 number of univalent chromosomes, and these counts averaged 28 uni- 

 valent chromosomes for each cell; this number, 28, is also assured 

 by the fact that in the last spermatogonic metakinesis each daughter 

 cell receives one half of each of the 28 chromosomes of the mother 

 cell; then taking into consideration the fact that on lateral views of 

 the nuclei each U- or V-shaped chromosome, when it is seen in its 

 entirety from the side, is seen to be composed of a right and left 

 half joined by a central linin band, the conclusion is practically 

 rendered certain that in the synapsis there are formed 14 pairs of 

 univalent chromosomes or 14 bivalent chromosomes. And this is cor- 

 roborated by the number of bivalent chromosomes found in the equator 

 of the 1st maturation monaster, where exactly 14 are found. 



I have searched carefully to find evidence of a discharge of 

 chromatin substance from the nucleus in the synapsis stage, but have 

 found none. The nucleus and cytoplasm are in this stage rapidly in- 

 creasing in size (compare Figs. 60 — 63, Plate 19, with Figs. 65—83, 

 Plate 20), so that there must be an active metabolism, and therefore 

 there may be an expulsion of bye-products of the chromatin from the 

 nucleus, but it can not be in the form of granules, but must be in 

 the form of thin fluid substances which are not differentiated by the 

 stains employed. Necessarily with any construction of new substances 

 in organic growth there must be katalysis of other substances \). 



1) Here I would like to call attention to the probability, that the 

 disappearance of the nuclear membrane in the mitoses of Metazoa 

 may be correlated with an act of excretion or discharge of substance 

 from the nucleus. There are bye-products formed in the nucleus, such 

 perhaps as the true nucleoli, which are in the form of bodies too large 

 or too dense to pass out through the nuclear membrane ; for such 

 bodies the periodical disappearance of the nuclear membrane, in mitosis, 

 would offer the possibility of discharge. Unless there were some such 

 provision as this the nucleus in the course of long-continued metabolism 

 would become choked up with the denser waste products. If this be 



