Studies OH Chromosomes 175 



cases it seems clear that each pole receives 13 chromosomes, as 

 follows : 



a 10 + / + 2.t = 13 6 10 4- / + 2i = 13 



Fig. 10, 0, on the other hand, shows a perfectly clear case in 

 which the hexad element has separated into a 2-group and 4-group: 

 Fig. 10, p, shows what is probably a later stage of the same type. 

 In both these cases one pole appears to receive 12 and one 14 as 

 follows : 



i2 10 + 7 + 35 + 14 b 10 + i + s = 12 



one pole receiving but one supernumerary, and the other three. 

 The cases in which all of the components may be clearly recog- 

 nized in the anaphases are comparatively rare, and in the greater 

 number of them the distribution of the supernumeraries appears to 

 be symmetrical. Of their unsymmetrical distribution in some 

 cases there can be no doubt (and the same is true of the T4-chromo, 

 some form, as described beyond). The few undoubted cases of 

 this all show one to one pole and three to the other (as in Fig. 10, 0- 

 p), and I have never found a case in which all four pass to the same 

 pole. 



It seems, therefore, probable that in the 26-chromosome type 

 there are at least six classes of spermatozoa, as follows: 



(l) 10 + 7 + 2i = 13 (2) 10 + / + 2i = 13 



(3) 10 + 7 + s = 12 (4) 10 + ;' + 3s = 14 



(5) 10 + 7 + 3^ = 14 (6) 10 + ; + i = 12 



It is possible that the following four additional classes may be 

 produced : 



(7) io + 7 + 4i=i5 (8) 10+/ =11 



(9) 10+7 =11 (10) 10 + / + 4i = 15 



Perfectly clear spermatogonial figures of this type were rarely 

 found, though many of them show approximately 26. The nor- 

 mal group of fern., No. 42, is shown in Fig. 2, h. Two groups from 

 fem. No. 40 (with two small and two large supernumeraries) are 

 shown in Fig. 9, n, 0, each having 26 chromosomes including four 

 small ones (cf. Fig. 2, k). Two ovarian groups from gran.. No. 61, 



