Studies on Chromosomes. 527 



much larger chromosome beside which it lies is to be identified 

 as the larger idiochromosome. Besides these fifteen undoubted 

 chromosomes one or more paler rounded bodies are often present, 

 lying outside the chromosome-group, sometimes close to it, that 

 are undoubtedly the remains of the plasmosome of the growth- 

 period. 



In side views of the metaphase-figure all of these chromosomes, 

 with one exception, have a symmetrical bipartite (rarely a quad- 

 ripartite) shape; and in the ensuing division these are equally 

 divided. One of the small chromosomes (heterotropic) never 

 shows a bipartite shape, but is simply elongate and more or less 

 fusiform (Fig. 4, c, d, e). As the division proceeds, this chro- 

 mosome at first remains near the equator of the spindle and then 

 passes over bodily toward one pole where it enters the daughter 

 group (Fig. 4, /, -), finally shortening again so as to assume 

 a spheroidal form. One of the secondary spermatocytes there- 

 fore receives fifteen chromosomes, the other fourteen. 



The failure of this small chromosome to divide in the first 

 mitosis at first seemed to me so anomalous (I had not then observed 

 the similar phenomenon in Archimerus, described in the foregoing 

 section) that for a time I thought that this body must be one of the 

 fragments of the plasmosome; and this suspicion was strengthened 

 by the fact that other plasmosome-fragments are often found 

 lying near or in the spindle (Fig. 4, g~). Further study, however, 

 conclusively showed that this suspicion was not well-founded. 

 The plasmosome-fragments are always rounded, paler, wholly 

 inconstant in position and never lie in the equatorial plate. The 

 heterotropic chromosome, on the other hand, is always present 

 (many division-figures in all stages have been studied) and 

 every stage of its asymmetrical distribution has been repeatedly 

 observed. All doubt is, moreover, removed by a study of the 

 metaphase-figures of the second division. Great numbers of 

 these, showing the relations with schematic clearness, are avail- 

 able for study. In polar view these show either fourteen or 

 thirteen chromosomes (Fig. 4, h, /'), the two classes existing in 

 approximately equal numbers, and side by side in the same cyst. 

 At first sight neither of the small chromosomes of the first division 

 can be distinguished in polar view of the second. This is owing 

 to two causes: First, the small heterotropic chromosome, having 

 failed to divide while all the others are but half as large as before, 



