526 



SCIENCE 



[N. S. Vol. XLIII. No. (Ulll 



ergo propter hoc. The fact is we do not 

 know what causes the division of a eentro- 

 some, or chromosome or cell or a many- 

 celled organism. 



Spencer held that since the volume of 

 any organic body increases as the cube of 

 its diameter, whereas its surface, through 

 which it must receive nutriment, increases 

 only as the square, it must divide after 

 reaching maximum size in order to restore 

 a proper ratio of surface to volume; but 

 although this may be true in general, the 

 sizes of cells, or of other organic bodies, 

 vary enormously and it does not seem pos- 

 sible to explain all these differences in size 

 in accordance with Spencer's hypothesis 

 alone; furthermore, there is no indication 

 of the mechanism by which this general 

 need to divide actually causes division. 

 Boveri assumed that chromosomes and 

 nuclei grow until they are equal in size to 

 the parent structures from which they 

 came and that they then divide ; but this is 

 far from being true in some cases. In the 

 cleavage of the egg the cells, nuclei, 

 chromosomes and centrosomes progressively 

 grow smaller, and this not at any uniform 

 rate for all cells, some growing smaller 

 much more rapidly than others. R. Hert- 

 wig finds the cause of cell division in the 

 preservation of a proper ratio between the 

 nuclear volume and the cell volume, but as 

 I have shown there is no contant nucleus- 

 plasma ratio since this ratio differs greatly 

 even in different cells of the same embryo. 

 Strassburger held that the cause of cell 

 division was to be found in the limit of the 

 "working sphere of the nucleus," and 

 that when in the growth of the cell this 

 limit was reached, the cell divided; but 

 again it may be objected that there is no 

 fixed limit to the "working sphere of the 

 nucleus" even in the same animal; in some 

 cells of Crepidula the volume of the nu- 

 cleus at the time of division is three times 



that of the cytoplasm, in others the cyto- 

 plasm is fifteen times that of the nucleus. 

 Apparently no single one of these factors 

 is the determining cause of cell division, 

 and it seems probable that the latter is 

 brought on by the coincidence of several 

 more or less independent factors. 



In a series of contributions and in two 

 recent books Child has emphasized the im- 

 portance of polar "gradients of metabo- 

 lism" as the basis of organic individuality. 

 He finds, for example, that metabolism is 

 most active at the anterior or head ends of 

 certain protozoa, hydroids, flatworms, em- 

 bryos, etc., and that it becomes less active 

 toward the opposite ends. Regions of 

 higher activity "dominate" those of lower 

 activity, and whenever the metabolic activ- 

 ity of the head region ceases to dominate 

 the entire body, secondary regions of 

 higher metabolic activity appear and may 

 lead to division, one individual thus be- 

 coming two; the basis of individuality is 

 thus reduced to polar gradients in metab- 

 olism. But in existing organisms physio- 

 logical gradients are associated with corre- 

 sponding gradients in material structure, 

 since structure and function are insep- 

 arable in living things. Disembodied 

 functions are as unknown in biology as 

 are disembodied spirits. Doubtless gradi- 

 ents of metabolism as well as of growth, 

 division, differentiation and sensitivity 

 exist in organisms; but there is good rea- 

 son to maintain that such gradients in 

 physiological processes are associated with 

 corresponding gradients in material sub- 

 stances, and this is merely to hold that 

 axial differentiations, both physiological 

 and morphological, exist in organisms. 

 That such differentiations frequently ac- 

 company the division of cells or of multi- 

 cellular organisms is well known, but that 

 they cause these divisions is unproved. 

 The simplest individuals, such as chromo- 



