pinney: chromosomes of anasa tristis. 365 



with the others. This may be regarded as due to the fact that 

 the same conditions are present in all of the chromosomes pre- 

 vious to division ; that is, the degree of union between the chro- 

 matids is the same in each diad, while it will be remembered 

 that this relation between the diads which composed the tetrads 

 of the first spermatocyte varied. 



When the chromosomes separate sufficiently to expose a clear 

 zone of fibers in the region occupied by the equatorial plate 

 we find the accessory chromatid still in its place, and it keeps 

 this position until the remaining chromatids have almost 

 reached the opposite poles. Figures 13 to 18, inclusive, plate 

 LXIX, show the successive steps in this division. Well coa- 

 lesced daughter groups of this division are observed at oppo- 

 site ends of the spindle, while the accessory lingers half way be- 

 tween. The accessory finally approaches one of the groups and 

 almost immediately the dividing cell wall is formed. This 

 leaves the accessory near the newly formed cell wall, entirely 

 apart from the group to which it evidently belongs. Figure 19, 

 plate LXIX;, is a group of such bodies drawn from cells in this 

 stage. The size of this lagging body was thus found to be 

 constant. The spindle fibers between the two groups persist 

 until after the nuclear membranes of the daughter cells are 

 formed. Before this occurs, however, the changes shown in fig- 

 ures 20 to 25, inclusive, plate LXIX, take place. This consists 

 in the withdrawal of the accessory from its position near the 

 dividing wall into the now shapeless chromatin mass of the new 

 nucleus. In studying these processes it was very easy to obtain 

 sections which showed the two daughter cells in the same plane. 

 The connecting band of spindle fibers leaves no doubt as to 

 their relation even after the new cell wall has been formed. 

 The accessory occurs in only one of these cells. A further 

 significant fact is that in every such case where both daughter 

 cells can be identified beyond doubt one always shows the ac- 

 cessory at some stage of its characteristic activity, while in the 

 other no corresponding body is to be found. 



As soon as the membrane forms around the spermatid nu- 

 cleus which contains the accessory this element becomes de- 

 tached from the mass of which it forms a part and moves again 

 to the nuclear wall. Its individuality again becomes prominent. 

 Figures 26 to 31, plate LXIX, show this condition. Here also 

 the same precautions in determining its identity were observed. 



