4o6 



NATURE 



[November 24, 192 1 



to the organism. Just as the various steps in the 

 metabohc process are dependent on those which 

 preceded them, so when an organism becomes 

 differentiated into parts, when the main process 

 becomes sub-divided into subsidiary ones, these 

 react on each other. What is internal to the 

 whole becomes external to the part. An external 

 stimulus may set up an internal metabolic change, 

 giving rise to a response the extent and nature 

 of which depend on the structure of the 

 mechanism and its state when stimulated, that is 

 to say, on the effect of previous responses. Such 

 a response may act as an internal stimulus giving 

 rise to a further response, which may modify the 

 first, and so on. Parts thus become marvellously 

 fitted to set going, inhibit, or regulate each other's 

 action ; and thus arises that power of individual 

 adaptation, or self-regulation, so characteristic of 

 living organisms. The processes of temperature 

 regulation, of respiration, of excretion are ex- 

 amples of such delicate self -regulating mechanisms 

 in ourselves. But one of the great advantages 

 thereby gained by organisms is that they can 

 regulate their own growth and ensure their own 

 " right " development. Whereas the simplest 

 plants and animals are to a great extent, so to 

 speak, at the mercy of their external environment, 

 except in so far as they can move from unfavour- 

 able to more favourable surroundings ; whereas 

 their characters appear in response to external 

 stimuli which may or may not be present, and 

 over which they have little or no control--the 

 higher organisms (more especially the higher 

 animals), as it were, gradually substitute internal 

 for external stimuli. Food material is provided 

 in the ovum, and the size, structure, and time of 

 appearance of various characters are regulated to 

 a great extent by use and by the secretions of 

 various endocrinal glands, the action of which has 

 been so successfully studied, among others, by 

 Sir E. Sharpey Schafer in this university. Thus, 

 as is well shown in man, the higher animals acquire 

 considerable independence, and are little affected 

 in their development by minor changes of environ- 

 ment. Inheritance is thus made secure by en- 

 suring that the necessary conditions are always 

 present. 



We may seem to have wandered far from our 

 original question ; but the answer now appears to 

 be that only those characters can be regularly 

 inherited which depend for their appearance on 

 conditions always fulfilled in the normal environ- 

 ment (external or internal); and those characters 

 will not be regularly inherited which depend on 

 stimuli that may or may not be present. 



Now it will be said, and not without some truth, 

 that all this is mere commonplace admitted by all ; 

 but, if so, it is, I think, often ignored or mis- 

 understood in discussions on heredity, more espe- 

 cially in semi-popular writings, and sometimes 

 even in scientific works. However, I quite will- 

 ingly admit that the real problems Darwin left to 

 be solved by the evolutionist are the nature of 

 the germinal factors themselves, and more espe- 

 cially the origin of the differences between them, 

 NO. 2717, VOL. 108] 



the origin of those changes which give rise to 

 mutations. 



That these factors " must at least be self-pro- 

 pagating substances, subsidiary vortices in the 

 main stream of metabolising living protoplasm, is 

 certain, since they grow and multiply repeatedly, 

 to be distributed to new generations of germ-cells. 

 That they may be relatively constant and remain 

 unaltered for generations seems also certain, since 

 organisms or their parts can continue almost un- 

 changed for untold ages. That they can act in- 

 dependently, can be separately distributed into 

 different germ-cells, and can be re-combined seems 

 likewise to have been proved by the brilliant work 

 of Mendel and his followers. So independent and 

 constant do they appear to be that modern 

 students of heredity tend to treat them as so 

 many beads in a row, as separate particles them- 

 selves endowed with all the properties of inde- 

 pendent living organisms, the very properties we 

 wish to explain. While not prepared to accept 

 these views without qualification, it seems to me 

 that it can scarcely be doubted that some such 

 units must exist whether in the form of discrete 

 particles or merely of separable substances. But 

 not until these factors have been brought into 

 relation with the general metabolism of the 

 organism, as links in the chain of processes, will 

 the problem of inheritance approach solution. If 

 the theory is to be completed it must attempt to 

 explain how they come to differ, how their orderly 

 behaviour is regulated, in what functional relation 

 they stand to each other, what is the metabolic 

 bond between them. That harmonious processes 

 may be carried out by discrete elements in co- 

 operation is shown in cases of symbiotic combina- 

 tions such as the lichens, or the green algae in 

 such animals as Hydra and Convoluta. Here an 

 originally independent organism takes its place 

 and does its work regularly in another organism, 

 and may even be propagated and transmitted from 

 one generation to the next in the germ-cell ! Most 

 instructive, also, are the recently studied cases of 

 bacteria and yeasts living regularly in certain 

 special tissues of various species of insects, where 

 thev exert a definite influence on the metabolism 

 (see the works of Pierantoni, Buchner, Glaser).| 

 These no doubt are mere analogies, but theyj 

 serve. 



In all probability, then, factors of inheritance 

 exist, and the fundamental problem of biology is 

 how are the factors of an organism changed, ol 

 how does it acquire new factors? In spite of itf 

 vast importance, it must be confessed that littU 

 advance has been made towards the solution o\ 

 this problem since the time of Darwin, who con- 

 sidered that variation must ultimately be due tr 

 the action of the environment. This conclusioi 

 is inevitable, since any closed system will read 



7 Herbert Spencer's "physiological units," Darwin's "pangens,^ 

 Weismann's "determinants," are all terms denoting factors, but with some-1 

 what different meanings. More recently Prof. W. Johannsen ( Elemente 

 der exakten Erblichkeltslehre." 1Q09) has proposed the term gene for a 

 factor, "genotype" for the whole assemblage of factors transniitted by a 

 specie-;, ard "phenotype" for the characters developed from them. 1 his 

 clear system of nomenclature, although much used in America, has not been 

 generally adopted in this country. 



