II CLEAVAGE 47 



divide independently, they are not divided by the spindle. 

 The function of the spindle-fibres is merely to pull apart 

 their halves. But the spindle does take an active part in 

 cell-division. In plants the cell-plate, the future cell-wall, 

 arises by equatorial thickening of the spindle-fibres, and 

 there is a parallel phenomenon in the animal cell, for after 

 the separation of the daughter chromosomes there arises a 

 plate of material in the spindle equator which has a less 

 surface-tension than the remainder of the cytoplasm. It is 

 the reduction of the surface-tension, or rather the greater 

 surface-tension of the remainder, which pulls the two blasto- 

 meres apart. This not only appears to be a theoretical 

 necessity but is experimentally demonstrable. For if a drop 

 of rancid olive oil be floated on water, or better on a mixture 

 of alcohol and water, and a thread soaked in weak potash 

 be laid across a diameter of the drop, the soap now formed 

 along this diameter having a less surface-tension than the 

 remainder, the drop divides. The experiment is due to 

 Robertson : it may be readily verified. 



We have finally to inquire whether the last event of ordinary 

 fertilization, the alteration of the egg symmetry, finds a 

 parallel in artificial parthenogenesis. In only one sense, so 

 far, has it at present been found possible to give an answer 

 to this question, in the frog's egg, where the grey crescent 

 appears as a result of fertilization on that side of the egg 

 opposite to the point of entrance of the spermatozoon. 

 Brachet has shown that in the eggs stimulated by puncture 

 a grey crescent appears, and at the same time as in the 

 controls, but that it bears no definite relation to the point of 

 puncture, being variably on the same side, on the opposite 

 side, or at right angles to the latter. There is on the other 

 hand an invariable coincidence between the plane of symmetry 

 of the egg, as thus defined, and the median plane of the 

 embryo, for the dorsal lip of the blastopore always appears 

 in the region of the grey crescent. Brachet is therefore 

 driven to the somewhat remarkable conclusion that the 

 unfertilized egg is only apparently radially symmetrical, and 



