CHROMOSOME THEORY OF HEREDITY 243 



and it accomplishes this by being itself a unit of a higher order, 

 whose parts are linked together in a definite and specific way, 

 each of the parts, and therefore the unit-complex as a whole, 

 being self-perpetuating. 



If this is so, then we are presented with a piece of chemical 

 machinery which would appear to be admirably adapted for 

 the task of acting as self-regulator of heredity ; for passing on 

 unchanged from generation to generation that raw material, 

 that constitution which in interaction with the environment 

 gives the adult organism ; and therefore for resisting the very 

 type of eifect which the Lamarckians consider to be operative 

 in evolution. 



A chemical mechanism of this sort will undoubtedly prove 

 in the long run to be susceptible to some external agency or 

 other. No doubt, too, we shall some day find out how to break 

 up the atom and liberate its contained stores of energy ; but 

 so far we have not discovered the way. And we have not yet 

 discovered the way to alter the genes at will. If the recent 

 experiments of Guyer are confirmed (Guyer and Smith, /. Exp. 

 Zool., 31, 1920, p. 171) it appears that some factors can be 

 directly attacked by the methods of immunology. Otherwise, 

 results have been negative. Selection is of no avail. (The 

 only positive results are those of Jennings and of Hegner, in 

 Protozoa (see Proc. Nat. Ac. Sci., 4, 191 8), and these are con- 

 tradicted by current researches of German workers). Gradual 

 acclimatisation to changes in the chemical constitution of the 

 medium have an effect, but not a cumulative one, and not per- 

 sisting after sexual fusion (Jollos, Zeits. Ind. Abst. Vererb., 12, 

 1914, p. 14). Attempts to raise the upper level of heat tolerance 

 in insects — a type of experiment in which, if anywhere, one 

 would expect to find evidence of functional adaptation — have 

 proved unavailing ; indeed, the rearing of one brood at a higher 

 temperature makes it more difficult, instead of easier, to bring 

 up the next generation at the same temperature (Northrop, 

 /. Gen. Physiol. 2, 1920, p. 313). 



Indeed, the more one considers the nature of any such unit- 

 complex of genes as I have endeavoured to sketch, and the 

 more one discovers as to the apparently accidental and cer- 

 tainly very localised process of mutation in single genes, the 

 more difficult is it to imagine any method (save perhaps one, 

 like Guyer's, connected with protein specificity), which will 

 enable us to alter the genes directly. This is not to say that 

 some such method does not in reality exist ; but the difficulty 

 of imagining how a functional adaptation in the soma could 

 be reproduced, in whole or in part, in the next generation, 

 becomes greater and greater the more we analyse the implica- 

 tions of the gene-constitution hypothesis. The functional 



