February 29, 191 2] 



NATURE 



585 



ible), find a good measure of their degree of similitude 

 with a given orientation. 



The problem is one over which the late Sir Francis 

 Galton was at times much exercised when discussing the 

 resemblance of portraits of the silhouette type. It was 

 further considered very fully when the proposal to pre- 

 pare average or type cranial contours was originally dis- 

 cussed in the Biometric Laboratory some five or six years 

 ago. Prof. D'Arcy Thompson's scheme is suggestive, but 

 it is very far from unique. I feel doubtful whether any 

 scheme for all these contours could possibly be other than 

 conventional, but I suggest that, even for a good con- 

 ventional scheme to be reached, we must have further 

 knowledge of the mathematics of the subject, i.e. we 

 want to study measures of the similarity or dissimilarity 

 ■of what we may perhaps call " resemblant contours." 



Karl Pearson. 



Biometric Laboratory, London, February ii. 



The Mnemic Theory of Heredity. 



If it were explained clearly in what respects an 

 " acquired " character is more acquired and less innate, 

 germinal, and inherited than an " inborn " trait, a real 

 service would be rendered to science, and, possibly, a 

 controversy which at present seems interminable might be 

 «nded. A unicellular organism distributes itself between 

 its daughter-cells. Here, obviously, there is actual inherit- 

 ance ; and, if the acquirements of the parents persist in 

 the offspring, there is inheritance of acquirements. But a 

 multicellular organism does not distribute itself. It is a 

 cell-community, and, so far as is known, offspring are 

 derived not from it as a whole, but from particular 

 members of it — the germ-cells. There is thus no inherit- 

 ance from the " parent " in the sense that there is inherit- 

 ance among unicellular types. For example, the child 

 ■does not inherit the parent's nose, leaving the parent 

 derelict. The latter keeps the whole of his nose for him- 

 self. 



The germ-cell is a bundle of potentialities for develop- 

 ment. It develops into an animal or plant of the species 

 whence it is derived under the influence of various stimuli 

 — food, temperature, light, moisture, internal secretions, 

 use, injury, and the like. Thus in man one kind of 

 stimulus causes a hand to develop, another a scar, a third 

 a use-callosity. Nothing develops in the individual, nothing 

 can develop, unless both the potentiality and the appro- 

 priate stimulus are present. All kinds of potentialities are 

 equally products of evolution, and are equally rooted in 

 the germ-plasm. Thus the potentiality to develop a scar 

 is as much a part of the germ-plasm as the potentiality 

 to develop a head. Some characters develop more certainly 

 than others, but this is only because the stimulus (not the 

 potentiality) under which they grow is more certainly 

 present. Thus a head develops more certainly than a 

 particular scar, but the scar would develop as certainly as 

 the head were its stimulus (a particular injury) as con- 

 stantly present. In man the scar left by the destruction of 

 the umbilical cord is as constant as the head. 



It is customary to term traits which develop under the 

 stimulus of use and injury acquired, while all others are 

 called inborn. But if all potentialities are equally present 

 in the germ-cell, if all characters are alike products of a 

 reaction between internal potentiality and external 

 stimulus, what is the peculiarity that makes one kind of 

 character more inborn and inheritable than another? As 

 far as I am able to judge, the Lamarckian controversy 

 lias been conducted on the basis of a misuse of terms, or 

 on the (at present unwarrantable) assumption that the 

 multicellular organism is derived from its parent in the 

 sam<' SI ti-. ,1-; a unicellular is derived, or under the belief 

 (riUo Linuaii .citable) that the only characters that arise 

 in response to stimulus from the environment are those 

 which grow through the influence of use and injury. I 

 am able to understand, for instance, how a negro who has 

 a scar differs both innately and by acquirement from a 

 white man who has no such scar. His potentialities are 

 different, and therefore he differs innately ; the stimuli to 

 which he was exposed differed, and therefore he differs 

 by acquirement. But it is one thing to apply these terms 

 to likenesses and differences between individuals and 

 another to apply them to charnctiM-s as such. I take it 



NO. 2209, VOL. 88] 



that the words " inborn," " acquired," and " inheritable " 

 have been illegitimately transferred from a connection in 

 which they have meaning to a connection where they are 

 unintelligible : for can anyone state precisely in what sense 

 the skin colour of a negro is more innate or germinal than 

 his scar? 



When it is maintained that " acquirements are trans- 

 missible," it is held, in effect, that characters {e.g. scars 

 and use-callosities) which the parent was able to acquire 

 in a certain way (as reactions to injury and use), because 

 a long course of evolution had rendered such acquisitions 

 possible to members of his species, tend, at the time of 

 observation, to be reproduced by the offspring in a different 

 category of characters and in ways (as reactions to other 

 stimuli) in which no ancestor had acquired them before, 

 and with which, therefore, evolution had nothing to do. 

 The evidence on which we are asked to accept this improb- 

 able supposition is usually equivocal, and, in recent times, 

 invariably such as cannot easily be verified. 



But turn to common experience. Facts are not the less 

 valuable or certainly true because they are familiar. Take 

 characters which develop under the stimulus of use, or, 

 what in the case of mind is the same thing, experience. 

 The development of some physical and mental traits, for 

 example, the hair, the teeth, external ears, reflexes, and 

 instincts, is not influenced by this stimulus. Other 

 characters, for instance, in man, the limbs, heart, kidneys, 

 brain, and all that is learnt, all that is intellectual, owe 

 their growth after birth mainly to it. Such characters 

 tend to atrophy when disused or little used, and to hyper- 

 trophy when much used. Low in the scale of life, animals 

 develop less under the influence of use and more under 

 other stimuli. But all the higher animals, in proportion 

 as they are highly placed, impelled by an instinct, sport 

 during youth, and thus stimulate mind and body to the 

 acquisition of traits without which maturity is incomplete. 

 Parental care after the beginning of conscious life is an 

 adaptation the function of which is to afford time and 

 opportunity for the acquisition of use-acquirements. It is 

 not found low in the scale of life among animals that, at 

 each stage, come ready armed by " inborn traits " to the 

 struggle for existence, and is most elaborate and prolonged 

 among the highest types. We call an animal intelligent 

 in proportion as it is capable of profiting from experience. 

 A human idiot is nothing other than an individual who, 

 reverting to a remote ancestral type, has lost the power 

 of growing mentally under the influence of experience. 



Manifestly the so-called acquirements are more 

 advantageous as responses to injury and use than they 

 would be if they grew in response to the more unvarying 

 stimuli. As they are, they render the animal adaptable, 

 capable of fitting himself to a diversity of environment. 

 Compare the adaptability of a man with that of a beetle. 

 Manifestly also " inborn traits " have undergone great 

 retrogression and use-acquirements great progression in the 

 higher animals, which, presumably, are derived from lower 

 types. It follows that, while a supposition that " inborn 

 traits " tend to be transmuted into " acquirements " might 

 be maintained with some appearance of plausibility, the 

 contrary Lamarckian doctrine that " acquirements " tend 

 to be transmuted into " innate traits " is untenable. The 

 mnemic hypothesis does not demonstrate the transmission 

 of acquirements. It merely makes confusion worse con- 

 founded by misusing another word. According to it, the 

 germ-cell remembers that which it never knew, and forgets 

 that which it knew. 



Southsea, February 17. G. .Arciidali. Reid. 



The reply to Prof. Dendy's comments upon my letter 

 (N.^TURE, February 8, p. 482) is briefly as follows. The 

 germ-cells are unicellular living organisms with a life- 

 cycle of their own, part of which they pass in a 

 metazoan individual. When they enter it, they are all in 

 potentialities so many twins identical with this. For the 

 time being its environment is theirs. The non-existent 

 protoplasmic bridges need not be postulated. If the germ- 

 cells could not " remember events in the past history of 

 the race," I fail to perceive how any developmental un- 

 folding would be possible. The relation of the doctrine of 

 acquired characters to the theory depends solely upon the 

 embryological facts of the cycle of animal life. 



