444 WESLEY R. COE, 



transferred from the old aster to one of the new, we may assume^ 

 perhaps, that after the separation of the daughter centrosomes, new 

 fibres grew outward in all directions to forna an aster as usual. Those 

 fibres which came in contact with the inner ends of fibres of the old 

 system which had the proper size and direction joined them end to 

 end, while those not coming into contact with such already-formed 

 fibres grew out independently. At the same time those fibres not 

 absorbed into the new systems degenerate. This process seems to be 

 beautifully illustrated in the division of the sperm-aster, as will be 

 described below. 



The process of formation of the second polar spindle in the 

 Mollusk JDiaulula has been worked out by MacFarland (18) with 

 the greatest detail. In Diaulula the process diff'ers from that de- 

 scribed above in that there is a most distinct and beautiful central 

 spindle between the divided centrosomes ^). This central spindle 

 develops directly into the polar spindle. MacFarland finds, moreover, 

 that "kein einziges Fädchen der alten Astrosphäre geht als solches in 

 eine Fibrille der beiden neuen über." 



The forming spindle early begins to rotate (Figs. 21, 22), and 

 the chromosomes are drawn into an equatorial plate between the 

 asters. The second polar spindle resembles the first in all essential 

 respects except that it is smaller (Fig. 10), and the structure of the 

 chromosomes is different. The centrosomes at the poles of the spindle 

 often divide as in the first polar spindle 2), though they do not invariably 

 do so. The centrosomes of the outer pole disappear before the polar body 

 is formed, and those of the inner pole soon after. When the second polar 

 body is cut off, a few of the spindle fibres thicken up and stain much 

 more intensely in their middle portions, so that a very slight Z wi schen- 

 körper is formed as has been described for many other eggs. 



1) The term "ceûtrosom" is used by MacFakland for the structure 

 which, in view of its later history, I have in this article called the 

 "centrosphere", and MacFaeland's "Centralkorn" corresponds with the 

 "centrosome" in Cerehratulus. 



2) Occasionally an egg is met with which exhibits a most ab- 

 normal activity of the centrosomes of the second polar spindle, both of 

 the outer and of the inner pole. The centrosomes not only divide 

 several times but they separate and form spindles between them. In 

 this way multipolar spindles are formed which have , sometimes , five 

 or six poles. This fact would indicate that the centrosomes remaining 

 in the egg are not exactly "worn out", and that they can and do divide 

 under certain abnormal conditions. 



