36S 



NATURE 



[August 17, 1893 



regions. Suppose, e.g. that in some region the rainfall in a long 

 series of years varied, not as in the cases above considered, 

 but in a certain regular correspondence with the sunspot curve ; 

 and in another region (perhaps farther south) in opposite cor- 

 respondence ; also that these variations were traced to the shifting 

 of a depression path. The opposite correspondence would obviously 

 not be a good reason for denying sunspot influence, but rather 

 corroborative evidence of such influence. Again, it will be 

 admitted, I think, as conceivable that we might find certain 

 great anticyclonic systems to vary in position or extent with the 

 sunspot variations. Suppose, then, an anticyclone which lay 

 over a region [a) at the time of minimum sunspots, were moved 

 in a given direction, say northwards, so that it came to cover a 

 region {b) at the maximum of sunspots and that it returned to a 

 by the next minimum. In that case a place, e.g., in the south 

 part of region a, would have high barometer at minimum sun- 

 spots, while a place in the north part of region /; would have 

 low barometer. And at the maximum of sunspots, on the other 

 hand, the two places would again have opposite conditions of 

 pressure (to each other and to the first). These are some out of 

 many aspects of the matter which seem to me to render doubt- 

 ful the affirmation that if the solar cycle influences weather, it 

 cannot produce an opposite succession of effects in different 

 (even neighbouring) regions. 



To revert, for a momant, to the shifting of depression-paths, 

 might it not, in some cases, account for certain changes ob- 

 served in the relative proportion of different wind directions ? 

 Suppose eg. that, by the shifting of a path a little southwards, 

 a pla;e which has been for some yea's in its southern border 

 comes to lie in the northern border, might it not thus come to 

 have more easterly wind ami less westerly ? A. B. M. 



The Non-Inheritance of Acquired Characters. 



Dk. Wallace, in a letter which appeared in Natijre on 

 July 20, asks for the opinion of naturalists as to the in- 

 terpretation of certain facts bearing upon the qiestion of the 

 " N> n-inheritance of Acquired Characters," and as I have given 

 much thought to the subject I venture to offer my opinion. 



In two papers pulilished in Natural SHence, V)l. i. (1892), I 

 set forth at some length a theory of heredity which has hitherto, 

 so far as I am aware, met with no pu'ilic criticism, and which I 

 believe sets the qiestion at rest, not by establishing the views 

 of either of the rival schools associated with the names of 

 Weismann and Lamarck respectively, but by showing that 

 another interpretation is possible, and one which while fundi- 

 mentally opposed to both of these makes it possible that there 

 may be some truth hidden in the almost meaningless statements 

 of the Weismannians and of the Lamarckians alike. 



Till " heredity " is defined, and till we know exactly what 

 we mean by " inheritance of characters" (hi they " acquired " 

 or " blastogenic "), it is useless to argue as to whether charac- 

 ters are "inherited" or not. 



Is the word "heredity" an abstract noun, the name of a 

 quality, a sort of magnified "family-likeness," or is it not? 

 Those who write of heredity are too prone to speak of " here- 

 dity" as if it were a force or combination o' forces producing an 

 effect ; as an "inherent tendency,"to resemble parents or other 

 ancestors which it is perhaps not unfair to compare to the " in- 

 herent tendency" of a watch to tell the time or of a weather- 

 cock to p lint to the south-west. There are those who even speak 

 of it as being " latent " for a lime and then, owing to some 

 u iknown cause, "springing into activity " anew and giving rise 

 to what we call "atavism." Eve\ "atavism" is not infre- 

 quently spoken of, as if it were of the nature of a force or 

 c )mbination of forces, comparable to a "latent tendency," 

 which after lying "dormant" or "latent" for a time in a 

 weathercock, suddenly springs into ncA' activity and causes it 

 to point as of old to the south-west. 



It appears to me that if we once grasp the idea that " here- 

 dity" is the name of a quality, a particular kind of "like- 

 ness" or "similarity," and nothing else, we shall be saved 

 from much useless discussion of propositions whijh are intrin- 

 sicilly almost, if not quite, meanin.;less. 



Artemia salina is the collective name given to a large num- 

 ber of indivi luals which have certain characters in common. 

 It would hardly seem to be necessary to suggest the probability 

 that this possession of many characters in cominon is due to 

 the action of Natural Selection ; that each new individual 

 possesses the characters in question solely by virtue of the fact 



NO. 1242, VOL 48] 



that Natural Selection has led to the production of individuals 

 possessing the power to produce, under given constant con- 

 ditions, eggs, which by virtue of their cons'itution will develop 

 under given conditions into adults possessing the chxracters 

 which natural selection has under those conditions rendered 

 nearly constant. 



It has been found that this same constitution does not 

 necessarily lead to the same series of developmental changes 

 under other conditions, and that in strong brine the eggs develop 

 into animals which, though capable of living and multiplying 

 under those conditions, differ in form from the ancestral A. 

 salina. This new form has no more right to rank as a species 

 than has a "worker" bee whose adult form differs from that 

 of its parent merely on account of certain conditions to which 

 it is exposed during development. 



It appears to me to be absurd to ask whether the "acquired 

 characters" of the so-called Artemia Milhausenii are inherit- 

 able or not. Experiment has shown that the constitution of 

 the species A. salina has so little changed that it still has the 

 power to produce eggs which under one set of conditions develop 

 into A. salina and under another set of conditions into A, 

 Milhaascnii. The average constitution of the species has not 

 varied : it still produces ova which will develop into either A. 

 salina or A. Milhausenii, according to the conditions to which 

 it is exposed. If we look upon the species as a whole, it is 

 not too much to say that it exhibits no acquired characters. If 

 bred in strong brine the individuals of many generations are 

 alike, having been moulded by like influences, intrinsic as well 

 as extrinsic. If the extrinsic influences change, new individuals 

 differ from the old ones, simply because the constitution of the 

 individuals as well as that of the species is such that under the 

 new conditions the developmental changes occurring differ from 

 those which would have occurred under the old conditions. 



Whether this is true of all species and under all conditions 

 consistent with life and multiplication, or is not true of some, is 

 a matter fir experiment, and can never be decided by argument. 

 The experiment has been made by nature, and also by man in 

 the case of Amhlystoma, and with a result in exact conformity 

 with the result in the case of Artemia. 



The experiment has also been made with white viice in 

 Freiburg, and it has been conclusively shown that under con- 

 stant conditions the characters of successive generations are 

 constant. One element of the environment in one series of cases 

 was Prof. Weismann armed with tools for amputation of the 

 the tails of the young mice, plus a determination to amputate 

 those tails. So long as this remained a constant factor in the 

 environment, so long and no longer did the taillessness of the 

 adult mice remain a constant character of the species. 



The Texan species of Saturnia, so long as the exclusive sup- 

 ply of Juglans regia is a constant factor in the environment, 

 may or may not have a constant group of characters. That is 

 a matter for experiment ; but innumerable experiments, called 

 collectively "domestication," have shown that whatever effect 

 changes of certain details of the environment — such as food — 

 may have, the suspension of natural selection will in the long 

 run lead to inconstancy of all those characters which are re- 

 lieved from its restraining influence. 



If anything has ever been rendered certain in biology by pro- 

 longed experiment and observation, it is the fact that specific 

 characters are maintained constant by selection and by that 

 alone. Long continued selection — natural or artificial— may 

 produce a seeming constancy of characters [7vhich we call 

 "heredity" !), but in the long run this constancy will vanish 

 when the particular selection which has induced it has been 

 suspended for a sufficiently long period. 



The discusssion upon the "Inheritance of Acquired Charac- 

 ters," though it has led to many valuable results, has been 

 throughout little more than a quibble, for in the whole dis- 

 cussion, so far as I am aware, the meaning of the word 

 " inheritance" has never been defined. Most of the disputants 

 appear to use the term as the name for the action of a force or 

 combination of forces, which s ime have called "heredity"-^ 

 a force or influence — either simple or complex — of which it is 

 perfectly safe to deny the existence. There is no such thing 

 as heredity — heredity is only a quality, a likeness or similarity, 

 and nothing more. That likeness of characters is simply and 

 solely due to the likeness of the influences which have produced 

 the like characters, and pre eminent amongst those influences 

 is natural selection, though every factor of the environment has 

 also had its part to play. So long as those like influences— 



