46 I'UYSIOLOGICAL GENETICS 



different processes with a different temperature coefficient were 

 involved which therefore were influenced differentially (speeded 

 or slowed) by the experimental temperature. The same inter- 

 pretation of the facts was later used by Zeleny (1923, 1933) and 

 by Harnly (1935). I think, however, that this interpretation 

 fits only the results of experiments within the range of physiologi- 

 cal temperatures. Outside this range, temperature shocks will 

 act always by differentially slowing down some processes. 



All the recent investigations bear out the essential correctness 

 of such a conclusion. Applied to the mutant gene, which acts 

 with identical effect as those shocks, this means that the mutant 

 gene acts by changing the rate of definite developmental reactions 

 or processes. Based on this conclusion, the further inference is 

 drawn that each gene acts through the control of rates of reaction 

 (Goldschmidt, 19176, d, 19206, 1927c). 



It seems that this conclusion, which the author has always been 

 inclined to consider the basic insight into the action of the gene 

 and which originally was derived from experimental facts reviewed 

 on page 52, is being more and more verified in recent work. 

 Many of the authors who have performed experiments bearing 

 on this question have now accepted this viewpoint: Plunkett 

 and Child, after a study of bristles in Drosophila; Zeleny, in his 

 experimentation on Bar eye; Harnly, who studied vestigial wang; 

 Dunn, after a study of minutes in Drosophila; Huxley and Ford, 

 studying eyes of Gammarus; and many others, analyzing different 

 types of cases (Wright, Honing, Tammes, Sinnott, etc.). We 

 shall refer to these facts at their proper place in our discussion. 



E. Some Pertinent Facts from Experimental Embryology 



At this point, some facts ought to be mentioned which in a 

 general way supplement the conclusions thus far derived. We 

 have already described Danforth's experiments, producing the 

 rumpless type of fowl as a phenocopy by experimenting upon the 

 embryo. In this case, the type produced corresponded to a 

 well-known type produced by a mutant gene. Many similar 

 cases are known, in which the treatment produces definite 

 abnormalities, definite types of monsters, though not exactly of 

 the same type as known mutants. 



Many experiments haVe been performed upon lower vertebrates 

 which resulted in the production of typical monsters. In fact, 



