INTERACTION OF GROWTH-RATIOS 195 



below that made by the controls left in place. Thus we clearly 

 see that the relative size of any organ is due to two variables 

 — its inherent growth-capacity, and a regulatory partition- 

 coefficient which allots material as between organ and rest- 

 of-body. 1 We could have deduced this from our studies on 

 the chelae of Uca, etc., but this work clinches the matter 

 decisively by means of experimental tests. 



This work of Harrison's has also shed important light on 

 growth-gradients. In addition to grafting whole eyes, he made 

 separate inter-specific grafts (a) of optic vesicle and (b) of 

 lens-producing epithelium (Harrison, 19293). Eyes were thus 

 produced which contained rapid-growing optic cup and slow- 

 growing lens, and others with slow-growing optic cup and fast- 

 growing lens ; and either sort in both kinds of host. In every 

 case, there was a mutual influence of the components upon 

 each other. The growth of the eyes as a whole was of inter- 

 mediate rate, that of the rapid component was slowed down, 

 that of the slow component speeded up. 2 



Fig. 87, constructed from Harrison's data, shows the results 

 quantitatively. The association of a fast-growing optic vesicle 

 with a slow-growing lens retards the growth of the optic vesicle 

 by 22 per cent, in a punctatum body (No. 2 in Fig. 87 : ratio 

 1-26 instead of i-6i) ; and by 17 per cent, in a tigrinum body 

 (No. 4 : ratio 0-83 instead of i-o) ; similarly, it accelerates the 

 growth of the lens by 13 per cent, in a punctatum body (No. 9) 

 and by 6 per cent, in a tigrinum body (No. 11). The reverse 



1 It is interesting to note that Twitty and Schwind (1. c.) in their 

 latest paper, published after this chapter was first written, also adopt 

 the phrase partition-coefficient of growth. They, however, think of the 

 partition-coefficient as a percentage ratio, and therefore, since the 

 ratio eye-diameter : body-length decreases considerably during larval 

 life, find it necessary to postulate a progressive change in the coeffi- 

 cient. If our conception be adopted of a partition-coefficient which 

 is an exponent of body-size, this difficulty disappears, and the coeffi- 

 cient can be considered to remain constant throughout growth. 



2 That such interaction of neighbouring parts as regards their growth- 

 intensity need not always occur is shown by the recent work of Schwind 

 (193 1), who made heteroplastic grafts of parts of the shoulder-girdle 

 between the same two species. He found that the presence of a 

 grafted coracoid region growing at a different rate from the host's 

 organs had no effect on the growth of the host's scapular region. Thus 

 we have here an example of mosaic growth-rates, as opposed to mutual 

 regulation, which is perhaps comparable to the mosaic differentiation 

 of some developing eggs as opposed to the regulatory capacities of 

 others. 



