556 ON INCREASE IN COMPLEXITY [pt. iii 



release of morphogenetic activities already potentially in the de- 

 veloping embryo for a genuine morphogenetic action". Now it is 

 easy to adduce in the light of biochemical researches, chains of causes 

 which can have the effects seen in the developing embryo. One could 

 start with the experiments of Huxley on the morphogenetic action 

 of thyroxin. We know from Ahlgren's work that thyroxin has definite 

 effects on oxidation-reduction mechanisms in vitro, just as other 

 hormones have, and anything that may locally affect oxidation- 

 reduction mechanisms has every chance of affecting the local fatty 

 acid concentration, as is indicated by the work of Hopkins, so that 

 in due course the lipocytic constant or some other such cellular 

 value (cf. Mayer & Schaeffer) will alter and change correspondingly 

 the surfaces of the intracellular phases in that region, with the final 

 outcome that, as in the models of Warburg, one chemical substance 

 may be formed instead of another. These two alternatives may be 

 thought of, for instance, as scleroprotein on the one hand or phos- 

 phoprotein on the other, and thence it requires little imagination 

 to picture the most profound morphological changes taking place. 

 These causal chains are being unravelled every day. 



But Rignano was probably prepared to admit the cogency of these 

 mechanisms within their own sphere; what he wanted to know was, 

 why should one embryo pass through all these changes and come 

 out at the end a dogfish, and another pass through them and come 

 out a skate. The standard answer of exact biology has, of course, 

 been that the genetic constitution of the former animal governs 

 the chemical morphogenetic processes, catalysing this and inhibiting 

 that, so as to produce the results we find. Rignano, Haldane, and 

 their associates have often replied that such a preformationism implies 

 a complexity too great to be imagined when it is faced with the 

 facts of biology as a whole, but this depends on one's imagination. 

 The alternatives, it might be argued, are much worse. 



The adversaries of genetics too often seem to suppose that every 

 one of the infinite number of characters which they see in a given 

 animal has to be represented in some way, within the nucleus. They 

 forget that large blocks, as it were, of the specific- characters may be 

 the result of single processes set in action by a gene, and do not 

 consider the possibility that a good deal of morphogenesis may be 

 associated with a "delegation of function", the gene activating 

 secondary key-factors, just as statesmen delegate many of their 



