152 SCIENCE PROGRESS. 



somes quite independently of the nucleus. Whether we 

 suppose that forces are at work in the protoplasm which 

 split the original centrosome and drive the two halves 

 asunder, with the resulting formation of the so-called 

 central spindle, or whether the vis viva originates in the 

 centrosome itself, the appearances would be precisely the 

 same, and, as a matter of fact, both views have been 

 urged. 



But if we allow that the centrosome owes its existence 

 to the energy liberated in the protoplasm, this does not ex- 

 clude the hypothesis that a periodical repetition of certain 

 conditions might as often lead to the formation of a centro- 

 some. Indeed, from what we already know of protoplasm, 

 it is not difficult to imagine, should the presence of such a 

 structure conduce to the more adequate accomplishment of 

 karyokinesis, that it might probably be differentiated in this 

 way. All the facts we possess on the subject of nuclear 

 division go to show the existence of an almost infinite 

 amount of variety in the details of the process, and any 

 possible physical condition which facilitated its completion, 

 or which enabled its course to be run through more smoothly, 

 would probably come to be more or less closely realised. 

 And the very belief that centrosomes ai'-e in some way usetul, 

 is founded on the frequency with which they occur, and on 

 their relation with the spindle. 



But however this may be, we can hardly escape the con- 

 viction, even on the independent-organ hypothesis, that it is 

 the protoplasm which, after all, carries its "division organs" 

 into those positions which determine the direction of divi- 

 sion, a direction that is often, perhaps always, so exactly in 

 accordance with what mathematical theory demands that its 

 plane can be predicted from the contour of the cell. And 

 in many of the apparent exceptions to this rule, as for ex- 

 ample in the eggs rich in yolk, to be found in many animals, 

 and perhaps also, though less clearly, in the oospheres of 

 Pimis, the difficulty vanishes as soon as one realises that 

 the protoplasm is not homogeneous throughout the whole 

 cell, but is obviously of a different texture in one part, as 

 compared with the rest. 



