584 ON INCREASE IN COMPLEXITY [pt. iii 



amphibian egg, for instance, leading from the protoplasm-rich 

 rapidly dividing animal pole to the yolk-rich slowly dividing vegetal 

 pole, has an enormous effect on the type of development which takes 

 place. It is plain, too, that much more might be said about the 

 relation of gradients to the various classes of eggs, alecithic, telo- 

 lecithic, and centrolecithic. In 1905 Morgan advanced the hypothesis 

 that the gradation of materials in the egg was a factor in establishing 

 physiological polarity, and Boveri came to very similar conclusions 

 about the Ascaris egg. 



Child made a step forward from these simple facts when he 

 propounded the conception of primary protoplasmic gradients of 

 which the morphological gradients were the obvious result, for he 

 brought the subject to some extent nearer the point where physico- 

 chemical analysis could begin. Moreover, he was above criticism 

 when he did not specify what sort of physico-chemical activity it was 

 that was responsible for the gradient. But it is difficult to follow 

 him when he concludes that the gradient is one of "oxidative 

 metabolism", "oxidising power" or metabolic rate. Apart altogether 

 from the fact that the experimental evidence will not carry this 

 conclusion, there are theoretical difficulties involved in it. (It 

 seems to have its roots, indeed, in that now abandoned idea, 

 ^ which was once common among the followers of Jacques Loeb, 

 that oxidation-processes and growth were very closely allied, even 

 that the master reaction of the growth-phenomenon was an oxida- 

 tion. If no physiologist now adopts this notion it is because so 

 many researches have shown it to be false. The work of Crozier 

 and his school on temperature characteristics might be mentioned, 

 in which the [x value for growth in general and embryonic 

 growth in particular practically never turns out to be 16,000. 

 Murray's demonstration of the diametrically opposite course 

 taken by growth-rate and metabolic rate, again, is an instance 

 of the same thing. All the tendencies of recent years have been 

 against any close identification of oxidative processes with growth. 

 A little reflection is enough, moreover, to convince one that such an 

 association is far from being a priori necessary. In the case of the 

 embryo of the chick in the egg, for example, its increase in size 

 could be represented by the curve a-b, in Fig. 95. At the beginning 

 and at the end of the period x-j> the size of the embryo is of course 

 given by the height of the curve above the abscissa, but to conclude 



