348 ROBERT H. BOWEN. 



to the same conclusion in respect to the division of the Golgi ap- 

 paratus, except possibly in mammals where the descriptions are 

 contradictory. With the mitochondria however the case is dif- 

 ferent, and there seems on the surface to be no general rule. 

 May this not be, after all, only an apparent difference? It seems 

 difficult to conceive of a purely chance distribution of mitochon- 

 dria which would produce the relatively constant results in Opis- 

 thacanthus for example, and it is still more difficult to conceive 

 of a chance arrangement which would allow the dividing cell to 

 sever the chondriosome-ring of Centrums into exactly equal 

 parts. Why should chance distribute the granules in the guinea 

 pig germ cell so evenly? Why, indeed, should the granules be 

 distributed at all, since in earlier stages they tend to be aggre- 

 gated at one pole of the cell ? In this connection the observation 

 of Meves ('14) on the dividing egg of Ascarls is of interest, for 

 he found that the mitochondria in the form of very small gran- 

 ules tended to collect around the poles of the cleavage spindles, 

 i.e., around the centrioles. Finally there is the case of the sper- 

 matocyte divisions of Ascaris in which the axial dimensions of 

 the mitochondria reveal the presence of an underlying polariza- 

 tion which, if the chondriosomes were spherical, would never be 

 suspected. 



If we combine these many facts, it is difficult to escape the 

 conviction that the division o<f the mitochondria and the Golgi 

 apparatus is always in more or less definite relation to the cen- 

 trioles and that the regularity which has been shown to exist in 

 cytoplasmic distribution is thus to be traced ultimately to the 

 same influences which control the formation of the mitotic figure 

 and the distribution of the chromosomes. I have come thus to a 

 concept of the dividing cell as a whole which differs somewhat 

 from that now commonly accepted. It seems to me probable that 

 the centriole is to be considered not merely the morphological 

 expression of the influences which construct the spindle and 

 marshal the chromosomes for division, but that it is in a much 

 wider sense the dynamic center of the whole cell, in which are 

 centered at the time of division influences that extend to prac- 

 tically all the elements of the cell and direct their division in an 

 orderly manner. This view of the functional significance of the 



