509 



it is not peculiar to the guinea is shown from the similar phenomenon 

 which I have recorded in my study of the pigeon. Furthermore, I have 

 looked sufficiently into my material prepared for a similar investigation 

 of the common chicken (Gallus gallus) to feel practically certain 

 that the same phenomenon is found there. 



The second conjugation does not imply, of course, that there is 

 necessarily a second reduction division. The scanty indications that 

 exist, indeed, point rather to a longitudinal division of the chromo- 

 somes (Fig. 30). This much seems certain, that the eight presumably 

 univalent chromosomes which pass to the one end of the spindle in 

 the primary spermatocyte pair again in such a way as to form four 

 bivalent chromosomes, and that eight of the nine which pass to the 

 opposite pole pair similarly, the odd remaining unpaired. The idea 

 suggests itself that in this second reduction in numbers we may have 

 an exemplification of a tendency toward diminution in the number of 

 chromosomes in the more highly organized animals due possibly to 

 more closely knit correlations in the germinal substances. Thus, an 

 ultimate decrease in numbers of chromosomes may come about through 

 fusions instead of the actual disappearance of individual chromosomes. 

 However, our knowledge of the relations between the germinal sub- 

 stances and adult characteristics is so much a matter of speculation 

 in itself that until further evidence is forthcoming any attempt toward 

 a solution of the problem under consideration can be little more than 

 a guess. 



Before leaving the discussion of the secondary spermatocytes it 

 should be mentioned that a few divisions were observed in which six 

 chromosomes were participating instead of the typical four or five; 

 presumably pairing of all the chromosomes for some reason had not 

 occurred. 



As a result of the divisions of the spermatocytes it will be seen 

 that there are two classes of spermatids, namely those which receive 

 four and those which receive five chromosomes respectively, so that 

 ultimately there must be two kinds of spermatozoa. The difference 

 between the two kinds of spermatids is usually well marked because 

 of the relative sizes of the masses of chromatin which each kind has 

 received. The odd chromosome is so large that it creates considerable 

 additional bulk in the spermatid nucleus. Furthermore, it not in- 

 frequently exists for some time as a distinct mass apart from the 

 other chromosomes which, having passed to the pole considerably in 

 advance of it, have ordinarily fused into a quadripartite or irregular 

 mass (Fig. 31). It ultimately fuses more or less definitely with the 



