274 H. LYNDHUKST DUKE. 



inatic substance, showing witliin it a granule of deeply 

 stained substance. The figure was very suggestive of the 

 state of affairs seen in fig. 18 a and I), with, however, a single 

 polar granule. The spai'sity of material unfortunately renders 

 a complete account of the first division-phenomena out of the 

 question. From a careful study of the slides at my disposal 

 I suggest the following as the more striking points, the 

 significance of which I shall revert to later on (see p. 278). 

 Firstly, the depth to which the spindle proper stains with 

 both Delafield's and Heidenhain's haematoxylin : secondly, 

 the proximity of the karyosome to the origin of the achro- 

 matic mass, and, later on, the very definite spindle-fibres 

 running dowji in among the karyosome remnants and the site 

 of the old nucleus : thirdly, the absence of regular chromo- 

 somes such as can at any stage be outlined or counted with 

 anything approaching certainty: fourtlily, the vesicles at 

 the poles of the latei- sp'indles, which form the centres of 

 definite astral figures. The nature of these vesicles it is 

 difficult to decide. Aie they centrosomes or incipient 

 daughter-nuclei ? As will be seen later, the daughter-nuclei 

 are strikingly vesicular ; and the fact that, if these vesicles 

 are considei-ed as centrosomes pure and simple, there aie no 

 other defined chromatic elements in the spindle figure, seems 

 to indicate their being early stages of the daughter-nuclei. 

 This being the case, the centrosome must be sought either in 

 one of the granules on the circumference of the vesicle, or 

 distal to the latter. On this point, though tempted to an 

 explanation, I dare not base a theory upon a drawing so 

 diaoframmatic as fio^. 16. 



Proceeding to the further division of the daughter-nuclei, 

 all uncertainty about the centrosome vanishes. In the 

 earliest stages, where eight or nine nuclei are present in each 

 cyst (fig. 17 a, h, and c), the astral radiations are very marked, 

 and the centrosome consists of a deeply stained mass at the 

 periphery of the nuclear vesicle, from which emanate the 

 strise. These, where they spring from the centrosome, are 

 extremely obvious. In fig. 19 c and d, stained with ha3ma- 



