22 BIOLOGICAL LECTURES. 



of the cell ; at the same time the central clear area disappears. 

 This occurs in the formation of every nuclear spindle ; in the 

 later stages of the nuclear division, when the central portion 

 of the spindle begins to disappear, the granules which were 

 distributed along the radiating fibres are gathered together 

 in such a way that one or more rows of them are pressed 

 closely together to form the definite boundary of the aster, 

 while within this boundary the granules are less compactly 

 arranged. This granular .zone, with its radiations, corres- 

 ponds, I believe, to the cortical zone of Van Beneden and 

 the archoplasm of Boveri. 



The central clear area probably corresponds to the medullary 

 zone of Van Beneden and the astrocoel of Fol, and in speaking 

 of it hereafter I shall employ the latter name. The astrocoel 

 always contains a large number of irregularly scattered granules, 

 which are considerably larger than those of the outer zone but 

 are not peculiar in any other respect. They show the same 

 micro-chemical reactions as the microsomes, and at an early 

 stage in the history of each aster they are closely connected 

 with the granules of the outer zone. After the nuclear spindle 

 has been formed these granules disaj^pear, and in their 

 place is found, at each pole of the spindle, a darkly staining 

 body much larger than any one of the central granules. 

 This is doubtless the central corpuscle of Van Beneden or 

 centrosome of Boveri, and I cannot doubt that the numerous 

 central granules represent a fragmented or scattered coitro- 

 somc. 



The structure of the outer granular zone of the aster, its 

 micro-chemical reactions, and particularly its method of growth, 

 seem to me to indicate plainly that this portion of the aster 

 is merely a part of the general cytoplasm temporarily modified 

 or differentiated for a particular function. Regarding the 

 astrocoel and centrosomes there is more doubt, but I believe 

 that the evidence is clearly in fa\'or of the view recently 

 expressed by Watasc,^ that these structures are also a part 

 of the cytoplasm. Whatever the ultimate origin of the 

 centrosomes may be, there is no doubt that in Crepidula they 



1 Watase. Homology of the f'entrosome. Jour. Morpli.. Vol. 8, No. 2. 



