376 Dr. J. O. W. Barratt. [Apr. 4, 



chromosomes are for the most part large spheroidal or ellipsoidal masses, 

 minor differences being recognisable in the smaller gemini. 



In addition to the above type, another less frequent form of mitosis was 

 encountered. This is exhibited in the dyaster shown in fig. 5. Here the 

 chromosomes, which form elongated narrow rods, curved or bent, are also 

 reduced in number, varying from 16 to 18 in each aster. This form may 

 represent a post-synaptic division. If this is so, it is impossible to determine 

 whether it is the first or a subsequent post-synaptic division, since there 

 is no means of identifying its particular generation. 



Spireme forms are fairly common. In these the chromatin granules are 

 not unfrequently abundant, but well-defined double beading of the spireme 

 thread has not so far been observed. It may be noted, however, that in the 

 testis of the rabbit double beading could not be clearly exhibited. Late 

 prophases, in which different permanent forms of the chromosomes com- 

 posing the gemini, similar to those described by Moore and Arnold,* were 

 recognisable, have not been met with, nor have masses of archoplasm or 

 archoplasmic vesicles been up to the present encountered in proliferating 

 epithelium of the rabbit's ear. 



No suggestion can be offered, from the observations made in the course of 

 this investigation, as to the significance of the reduction division which takes 

 place. Nor has it been possible to ascertain the further history of the 

 reduced cells, and, in particular, to determine whether division with the 

 reduced number of chromosomes may be repeated for more than one 

 generation. 



It may be observed that the number of chromosomes counted has not 

 been constant, but has varied from 28 to 36 in somatic mitoses, and from 

 14 to 18 in the reduction mitoses. The difficulty of accurately ascertaining 

 the number present under the conditions obtaining in stained sections 

 viewed under high magnification may in part account for the variation met 

 with, particularly in the former case. It is, however, not at all clear 

 that this explanation is completely satisfactory ; on the contrary, the 

 constancy with which variation is met with suggests, particularly when 

 reduction occurs, and enumeration of chromosomes becomes in consequence 

 much less difficult, that the number is not absolutely fixed, but exhibits 

 a certain degree of variation. 



Considerable variation was exhibited in different experiments in respect of 

 the number of division figures encountered. As a rule, the number was 

 greatest when the epithelial proliferation was most marked. An extreme 

 degree of mitosis was observed at the end of a week in some of the implanted 



* ' Boy. Soc. Proc.,' B, 1906, vol. 77, p. 563 ; also Moore and Enibleton, ibid., p. 555. 



