I 1 8 FRANK K. LILLIE. 



shown by some of the phenomena of centrifuged eggs ; when, 

 for instance, the direction of centrifugal force is at right angles to 

 the axis of the spindle, granules must stream by it. Now if the 

 density of the spindle area be no more than that of the surround- 

 ing ground substance and if there be no repellent force from the 

 spindle, such granules would pass through the spindle readily. 

 But on the contrary they are blocked by the spindle, and heap 

 up against it. Fig. 6 illustrates this condition. The accumula- 

 tion of granules on the distal side of the spindle is fatty in char- 

 acter, at least the granules are blackened by Flemming's fluid 

 which was used for killing, and they are the same in character as 

 those accumulated at the proximal pole. A pathway of such 

 granules runs from the pole of the spindle towards the proximal 

 accumulation, rendering the interpretation certain. We can 

 readily believe, therefore, that the relatively dense spindle moves 

 as a unit through the protoplasm. If, then, it be thrown into the 

 basophile cap, it will naturally tend to maintain its character ; 

 many such cases are in fact found. In other cases, the part of 

 the spindle within the basophile cap is partly changed, partly un- 

 changed ; but in very many cases the part within the cap is com- 

 pletely changed, so that the staining reaction of the spindle agrees 

 with that of the cap. 



Whether we consider the movements of the karyokinetic figure 

 produced within the egg by centrifuging, or whether we consider 

 the results of driving granules into the area of the karyokinetic 

 figure, the general conclusion that the poles of the spindles are 

 centers of force appears to me to be inevitable ; no system of 

 antagonistic fibrillse could behave in such a way. The difficulties 

 that the centrosome hypothesis has to meet are well known, and 

 their discussion does not enter within the scope of this contribu- 

 tion, which is in a sense a by-product of another problem. I 

 shall therefore be satisfied merely to present the evidence and to 

 indicate the direction in which it appears to point. 



In the development of the karyokinetic figure, the rays or 

 fibers appear to grow out of the centers up to the time of the 

 mesophase at least, and this must be due to increasing power of 

 the centers. In a magnetic model on the other hand the orienta- 

 tion of particles along the lines of force takes place simultane- 



