28 Arthur B. Latiib 



explained on the assumption of an intermittent or non-synchronous 

 activity of the centers, as Reinke^ has shown. - 



Assuming, then, the existence of some polar force exerting its 

 action at a distance, we are confronted with two possible alterna- 

 tives regarding the sign of this action. That is, we may imagine 

 either that the centrosomes attract, or that they repel each other. 

 Wilson^ has urged that the astral centers represent centers o{ trac- 

 tion, caused, perhaps, by a volume change at those places. This is 

 in entire agreement with the configuration assumed by the astral 

 rays and the spindle fibers. They simulate the magnetic field 

 between opposite poles, as pointed out above. But this view is 

 quite at variance with the actual movements of the centrosomes. 

 They move apart, even at a stage when astral rays are well devel- 

 oped and hence seem to repel each other and not attract as they 

 ought if they represent opposite poles. Lillie^ adopts the other 

 alternative, as did Meves.^ He considers the astral centers to 

 repel each other. In this way he explains the movements of the 

 centrosomes satisfactorily enough, but is confronted by the difl[i- 

 culty of accounting for the configuration of the astral rays. Lillie 

 assumes that electric charges located on the centrosomes are the 

 particular forces which produce the repulsion. He would explain 

 the unexpected configuration of the fibers and the astral rays by 

 the rather dubious assumption of a localized positive inter-astral 

 area which superposes its effect on the purely repellent action of 

 the astral centers. Looking at the matter more closely we see 

 that for every unit of negative electricity on the chromatin sub- 

 stance there should be a corresponding unit of positive electric- 



1 Reinke, Fr.: Arch. f. Entwicklungsmech., ix, 1900. 



■^ Rhumbler: Ihid., iii, iv and v, 1896, 1897 and 1899, has suggested a non-polar force to explain the 

 astral rays independently progressing rays of crystallization out of a supersaturated solution. While 

 avoiding the difficulty of crossed fibers this explanation encounters the still more formidable one of 

 accounting for the universal occurrence of curved fibers. 



^ Ibid., xiii, p. 354-395, 1901. See also his book, The Cell in Development and Inheritance, 3d 

 Ed. The Macmillan Company, New York. It is a pleasure to express my thanks for a most profitable 

 discussion of this whole question with Professor Wilson, who, it seems, had already considered the pos- 

 sibility of a hydrodynamic explanation. 



■• Amer. Jour. Physiol., xv, 46-84, 1905. 



^ Ergebn. d. Anat. u. Entwick., vii, viii. Merkel u. Bonnet. 



