8 HUMAN SPERMATOGENESIS: A STUDY OF INHERITANCE. 



and where consequently its normal division was delayed until the second matura- 

 tion mitosis. Each resulting spermatid would contain one more monad than is 

 normal. 



Condition h 4, non-homologous dyads at opposite poles of the spindle (fig. 

 52), found in 1 case. In this case each spermatid would contain the usual number 

 of monadic chromosomes, but each would lack one of a kind normally found in it. 



Fig. 53 shows an anaphase, and figs. 54, 55 show telophases of the second matu- 

 ration mitosis, the most distinct chromosomes only being drawn. In fig. 55 is seen 

 quite a regular condition for man, but one not observed by me in other objects, 

 the nuclei of the spermatids developing at unequal rates. 



Brief mention may be made of the other structures during the maturation 

 mitoses. The mantle fibers are distinct, the polar fibers much more delicate 

 and often scarcely visible. There is a pair of minute centrioles at each pole during 

 both mitoses. The behavior of the centrioles in the diakinesis could not be made 

 out, nor any trace of the spindle before its complete formation, except that in 

 a few cases (figs. 7, 8) a small body, evidently composed of two angular rods, was 

 found outside of the nucleus; this may be an early stage of the centrioles. 

 Through the maturation period acidophile bodies of variable occurrence and size 

 occur in the cytoplasm and within the spindles, and are shown upon the drawings 

 by light shading ; whether these are chromophilic corpuscles or portions of plasmo- 

 somes, or both, was not determined. Not a trace of an idiosome was found. 



B. DISCUSSION. 



Nothing of importance has been written on the growth period of the sperma- 

 tocytes, except that Gutherz (1911) could distinguish allosomes at this time (cf. 

 his footnote, p. 255). The number of chromosomes in human primary spermato- 

 cytes was computed to be 8 by Bardeleben (1892, 1897, 1898), to be apparently 

 18 by Wilcox (1900), to be 12 by Duesberg (1906) and Guyer (1910). 

 confirm Guyer's conclusion that there are 12, of which 10 are bivalent gemini, 

 each dividing in both maturation mitoses, and 2 are univalent allosomes (acces- 

 sory chromosomes) which divide only once in the two maturation mitoses. 

 Guyer's view is therefore probably correct that the number in the spermatogonia 

 must be 22, and not 24 as reasoned by Duesberg. 2 



But Guyer concluded that the two allosomes always pass undivided to one 

 spindle pole in the primary spermatocytes, reaching then only half of the secon- 

 dary spermatocytes, and in these dividing presumably equationally. He conse- 

 quently argued two classes of spermatozoa are produced in equal numbers, one 

 class containing division products of both allosomes, the other class lacking these. 

 That is to say, he overlooked the variability in behavior of the allosomes specially 

 studied by me. 



Before discussing this variability we will decide whether it should be regarded 

 as a normal or a pathological process. Is it only a pathological peculiarity of 



« The papers of Branca were not accessible to me. 



i 



