August 2, 1900] 



NA TURE 



lit. Thus there are some characters, e.g. tendency to certain 

 liseased conditions, which are more frequently transmitted than 

 others, and we ought to have, in each case, precise statistics as 

 to the probabilities of transmission. 



Again, there are some subtle qualities whose heritability must 

 not be assumed without evidence. Thus it is of very great im- 

 portance to students of organic evolution that Prof. Karl Pearson 

 has recently supplied, for certain cases, definite proof of the in- 

 heritance of fecundity, fertility and longevity. 



The familiar saying, " like begets like," should rather read, 

 " like tends to beget like," since variation is quite as import- 

 ant a fact as complete hereditary resemblance. If it seems 

 that in many cases the offspring is practically a facsimile repro- 

 duction of the parent, this may be due to absence of variation, 

 or, what comes almost to the same thing, to great complete- 

 ness of inheritance ; but it is more likely to be due to our 

 ignorance, to our inability to detect the idiosyncrasies. 



But it will be granted that the completeness with which the 

 characters of race, genus, species and stock are reproduced 

 generation after generation is one of the large facts of inherit- 

 ance. But th's does not sum up our experience, and we must 

 face the task of considering the different degrees of hereditary 

 resemblance. For these a confused classification and a trouble- 

 some terminology has been suggested, but it will be enough to 

 restrict attention to three familiar cases — blended, exclusive 

 and particulate inheritance. 



A preliminary consideration must be attended to. It is a 

 matter of observation that there are great differences in the de- 

 gree in which offspring resemble their parents ; but it is surely a 

 matter of conjecture that lack of resemblance is necessarily due 

 to incompleteness in the inheritance. Indeed, the fact that the 

 resemblance so often reappears in the third generation makes it 

 probable that the incompleteness is not in the inheritance, but 

 simply in its expression. The characters which seem to be 

 absent, to " skip a generation," as we say, are probably part of 

 the inheritance, as usual. But they remain latent, neutralised, 

 silenced (we can only use metaphors) by other characters, or else 

 unexpressed because of the absence of the appropriate stimulus, 



(a) In blended inheritance, the characters of the two parents, 

 e.g. in regard to a particular structure, such as the colour of the 

 hair, may be intimately combined in the offspring. This is par- 

 ticularly well seen in some hybrids, where the offspring often 

 seems like the mean of the two parents ; it is probably the most 

 frequent mode of inheritance. 



(b) In exclusive inheritance, the expression of maternal or of 

 paternal characters in relation to a given structure, such as eye- 

 colour, is suppressed. Sometimes the unilateral resemblance is 

 very pronounced, and we say that the boy is *' the very image 

 of his father," or the daughter "her mother over again"; 

 though even more frequently the resemblance seems " crossed," 

 the son taking after the mother, and the daughter after the 

 father. 



(f) It is convenient to have a third category for cases where 

 there is neither blending nor exclusiveness, but where in the ex- 

 pression of a given character, part is wholly paternal and part 

 wholly maternal. This is calkd particulate inheritance. Thus, 

 an English sheep-dog may have a paternal eye on one side, and 

 a maternal eye on the other. Suppose the parents of a foal to 

 be markedly light and dark in colour ; if the foal is light brown 

 the inheritance in that respect is blended, if light or dark it is 

 exclusive, if piebald it is particulate. In the last case there is 

 in the same character an exclusive inheritance from both parents. 



The facts above referred to may be considered in another 

 aspect, in terms of what is called the quality of prepotency. In 

 the development of a character the paternal or the maternal 

 qualities may predominate, as in unequal blending where there 

 is relative prepotency, or in exclusive inheritance where the 

 prepotency in respect to a given character is absolute. It seems 

 doubtful whether we gain much by using the word, since all 

 these general terms are apt to form the dust particles of intel- 

 lectual fog ; but we have to do with the fact that in respect to 

 certain characters the paternal inheritance seems more potent 

 than the maternal, or vice versa. 



It seems that one of the ways in which the quality of pre- 

 potency may be developed is by inbreeding, as Prof. Ewart and 

 others have maintained. 



Therefore, as inbreeding may be frequent in nature, especially 

 in gregarious and isolated groups, and as it tends to develop 

 prepotency, we are able to understand better how new vari- 

 ations may have been fixed in the course of evolution. And we 



NO. 1605, VOL. 62] 



can appreciate the position maintained by Reibmayr, that the 

 evolution of a human race implies alternating periods of 

 dominant inbreeding, and dominant cross-breeding. The in- 

 breeding gives fixity to character, the cross-breeding averts 

 degeneracy and stimulates new variations which form the raw 

 material of progress. 



Until we have more precise statistical data in regard to 

 blended, exclusive and particulate inheritance, we cannot hope 

 to simplify the matter with any security. But perhaps a unified 

 view will be found in the theoretical conception of a germinal 

 struggle in the arcana of the fertilised ovum, a struggle in which 

 the maternal and paternal contributions may blend and har- 

 monise, or may neutralise one another, or in which one may 

 conquer the other, or in which both may persist without com- 

 bining. We have extended the wide conception of the struggle 

 for existence in many directions ; it may be between organisms 

 akin or not akin, between plants and animals, between 

 organisms and their inanimate environment, between the sexes, 

 between the different parts of the body, between the ova, be- 

 tween the spermatozoa, between the ova and the spermatozoa, 

 and Weismann has suggested that it may also be between the 

 constituents of the germ-plasm. 



IV. Regres-SION. 



We have already referred to the fact that there is a sensible 

 stability of type from generation to generation. " The large," 

 Mr. (ialton says, " do not always beget the large, nor the small J 

 the small ; but yet the observed proportion between the large 

 and the small, in each degree of size and in every quality, 

 hardly varies from one generation to another." In other words, 

 there is a tendency to keep up a specific average. This may be 

 partly due to the action of natural elimination, weeding out 

 abnormalities, often before they are born. But it is to be 

 primarily accounted for by what Mr. Galton calls the fact of 

 " filial regression." 



As Mr. Galton puts it, society moves as a vast fraternity. 

 The .sustaining of the specific average is certainly not due ta 

 each individual leaving his like behind him, for we all know 

 that this is not the ca.se. It is due to a regression which tends 

 to bring the offspring of extraordinary parents nearer the aver- 

 age of the stock. In other words, children tend to differ less 

 from mediocrity than their parents. 



This big average fact is to be accounted in terms of that 

 genetic continuity which makes an inheritance not dual, but 

 multiple. " A man," says Mr. Pearson, " is not only the pro- 

 duct of his father, but of all his past ancestry, and unless very 

 careful selection has taken place, the mean of that ancestry is 

 probably not far from that of the general population. In the 

 tenth generation a man has [theoretically] 1024 tenth great- 

 grandparents. He is eventually the product of a population of 

 this size, and their mean can hardly differ from that of the 

 general population." 



At this point one should discuss reversion or atavism, but it is. 

 exceedingly difficult to get a firm basis of fact. The terna 

 reversion includes cases where through inheritance there re- 

 appears in an individual some character which was not expressed 

 in his parents, but which did occur in an ancestor. The char- 

 acter whose reappearance is called a reversion may be found 

 within the verifiable family, within the breed, within the species, . 

 or even in a presumed ancestral species. ' 



The best illustrations of reversion are furnished by hybrids. 

 Thus in one of Prof. Cossar Ewart's experiments a pure white 

 fantail cock pigeon, of old-established breed, which in colour 

 had proved itself prepotent over a blue pouter, was mated with 

 a cross previously made between an owl and an archangel,, 

 which was far more of an owl than an archangel. The 

 result was a couple of fantail-owl-archangel crosses, one re- 

 sembling the Shetland rock-pigeon, and the other the blue rock 

 of India. 



But great carefulness is necessary in arguing from the results 

 of hybridisation to those of ordinary mating, and even if some 

 of the phenomena of exclusive inheritance seem to show reversion 



iProf. Karl Pearson defines a reversion as " the full reappearance in an 

 individual of a character which is recorded to have occurred in a definite 

 ancestor of the same race," and atavism as " a return of an individual to 

 a character not typical of the race at all, but found in allied races supposed 

 to be related to the evolutionary ancestry of th« given race." " In reversion 

 we are considering a variation, normal or abnormal, from the standpoint of 

 htredity in the individual ; in atavism we are considering an abnormal 

 variation from the standpoint of the ances-ry of the race." But the two 

 words seem to be used by some authors in the converse way, or as equiva* 

 lent, and it is surely difficult to define the field of abnormal variation. 



