June io, 1922] 



NATURE 



735 



trrms. What is important is that biologists understand 



(Hie another, and, in spite of Sir Archdall Reid, we 



Midtest that they do. One expert may say, " I prefer 



use the term fluctuations not for slight intergrading 



iiiations, but for certain kinds of modifications"; 



other may say : " What you diagnose as ' fluctuating 

 variations ' are really ' fluctuating modifications,' though 

 I confess I cannot put my finger on the peculiarity in 

 nurture that induces them." Both may be right ; 

 both may be wrong ; but there is not in either of them 

 any confusion of thought. Prof. MacBride finds 

 evidence of the transmission of acquired evergreenness 

 in peach-trees ; we think that the results may have 



I en due to antenatal infection with certain metabolites, 

 tthe seed sojourns for a long time in intimate union 

 ■1 the parent. This is a question of interpretation ; 

 me is no confusion of thought in regard to the 

 piling of ' acquired character.' 



In his year of letters Sir Archdall Reid discovered so 

 many mares' nests that we cannot get rid of the 

 suspicion that he has been laughing up his sleeve all 

 the time. Since the scientific study of heredity began, 

 with Darwin let us say, it has been clear that words 

 like ' inherit ' and ' transmit ' are metaphorical, but 

 has any biologist been misled ? Heredity, after all, 

 is but a convenient term for the genetic relation between 

 successive generations, — a relation in which the persist- 

 ence of a specific organisation is secured through the 

 continuity of the germ-plasm, and yet not so rigidly 

 secured as to shut out the possibility of ne\Y germinal 

 variations. Except colloquially, the antithesis is not 

 between heredity and variation, but between complete 

 hereditary resemblance and divergence. The inherit- 

 ance is all that the living matter of the fertilised egg- 

 cell includes in virtue of its strictly hereditary relation. 

 It would not in the strict sense include either a symbiotic 

 alga or an antagonistic microbe. 



The vehicles of the specific organisation are 

 ' factors,' ' genes,' ' determinants,' many of which 

 are certainly located in the chromosomes of the nuclei, 

 but there are also extra-nuclear vehicles, e.g. organ- 

 forming substances in the cytoplasm. What the genes 

 precisely are remains uncertain ; they are the germinal 

 lepresentatives or initiatives of subsequent differentia- 

 tions ; in a few cases already it seems possible to say 

 that a particular gene, e.g. one affecting eye-colour in 

 so-and-so, lies about the middle of the third chromo- 

 some ! A character in the fully-formed organism is 

 usually the expression of several genes, and the same 

 gene may affect several characters. Every biologist 

 agrees with Sir Archdall Reid that the inheritance is 

 made up of factors for characters, not of the characters 

 themselves, for a character is the product of nature and 

 nurture. But just as Darwin sometimes spoke, for 

 NO. 2745, VOL. 109] 



short, of an adaptation as the result of selection, when 

 he meant, of course, selection acting on successive 

 crops of heritable variations, so it seems unnecessarily 

 purist to insist always on speaking of hereditary 

 factors rather than of hereditary characters. Gerould 

 has shown that conspicuous blue-green caterpillars (of 

 Colias philodice) may arise as mutations in a pure race 

 of inconspicuous grass-green caterpillars, and the 

 offspring of the adults into which two blue-greens 

 develop will breed true. But what is inherited is not 

 the coloration of the blood, for that is due to the 

 xanthophyll and chlorophyll of the food-plant. What 

 is inherited is some subtle gene (a nuclear enzyme 

 perhaps) which acts as a decoloriser or inhibitor of the 

 xanthophyll. Yet would it not be a trifle pedantic 

 to insist that the blue-green character is non-heritable. 



Whether the specific organisation which persists in 

 the germ-cell lineage can be added to in a definite way 

 by nurtural changes or modifications wrought in the 

 body of the parent, remains a question for legitimate 

 discussion (see, for instance, Mr. Cunningham's in- 

 teresting " Hormones and Heredity "), but it is not 

 playing the game to say that " there is absolutely no 

 meaning in the neo-Darwinian statement that acquired 

 characters are not transmissible." Some biologists 

 find convincing evidence that a novel somatic modifica- 

 tion can affect the germinal organisation in a manner 

 so specific that the offspring show some representation 

 of the acquired character. Others remain unconvinced 

 by any of the evidence that has been as yet adduced. 

 What is wanted is not a rumpus about terminology, 

 but more facts, and more critical interpretation. 



Sir Archdall Reid complains that biologists are 

 thirled to a particular classification of characters into 

 ' innate ' and ' acquired,' whereas they ought to take 

 a leaf from the physiologist's book. But this is simply 

 another of the windmills at which the Quixotic knight 

 tilts. When thinking along a particular line biologists 

 must distinguish characters as expressions of an in- 

 tactly persistent germinal organisation, or as expres- 

 sions of germinal rearrangements — shufflings of the 

 factorial cards, or as dints directly due to peculiarities 

 in nurture, and this does not exhaust the classification 

 sub specie hereditatis. But at another time, the biolo- 

 gist is just as open as any physiologist to classifications 

 of characters from other points of view. Are they 

 generic, specific, or varietal ; are they adaptive or non- 

 adaptive ; are they the outcome of natural selection, 

 or of sexual selection, or of neither ; are they exhibited 

 at birth or do they appear in the course of later develop- 

 ment (like the curlew's bill) ; are they activated by 

 hormones or by more general constitutional changes ; 

 are they progressive or involutionary, do they illustrate 

 differentiation or de-differentiation ? There is no 



