1894- CHARACTERS IN BIOLOGY. 269 



animal is generally thought of and observed in some special direction, 

 with the result that some appearances soon eclipse the others and 

 come to be regarded as the characters par excellence of the form. 

 Anything particularly striking to the eye or fancy at once assumes 

 undue importance and becomes marked off as a unit. And still more 

 often, those characters which are of the highest definitive significance, 

 and consequently, as a rule, of the lowest physiological importance, 

 those which are most " characteristic," take a more prominent 

 position in our minds than others, quite rightly for some purposes, 

 but wrongly for the study of inheritance. For thus one aspect of the 

 organism comes to stand for all. 



But, as a rule, characters are quite freely regarded as separate 

 from one another and as independent of the rest of the organism, and 

 are formed arbitrarily and used singly in the observation of the facts 

 of inheritance. The animal is made up of, or at least capable of 

 definite analysis mto, separate incoordinate individual appearances, 

 the presence or absence of which, in various proportion, determines 

 its specific and individual relations. And in marking off characters 

 from one another, one principle is always present, though loosely 

 adhered to; that is — that those appearances which usually vary 

 without any obvious accompanying change in the other relations of 

 the organism, are included in one term. Thus a patch of colour is 

 one character, being present or absent from animals otherwise not 

 different. Any relation of paired parts is always called one character, 

 not two, because such always bear a fixed relation to one another. 

 An instinct or habit is a character, being for us irreducible to its com- 

 plex causes. But any phenomenon for which we can find a cause, or 

 even a distant causal relation in another, is grouped with the latter as 

 one character. Wherever our knowledge of physiological relations 

 between parts breaks down, there we too readily assume a breach in 

 organic relations, calling all that lies between such breaches a unit 

 that varies independently. So, to modern biology, a seeming inde- 

 pendence in quantity or quality of one or many structures or functions 

 or both, from the rest of the organism, together with a real and 

 obvious causal relation between the former, are enough to give them 

 the status of a unit for heredity, having a set of little determining 

 demon particles all to themselves. And yet we are asked how it is 

 that characters can vary independently of one another ; and ingenious 

 theories are built to account for so strange a fact and to fit an 

 apparent union on to this disunion. Why, but for the very simple 

 reason that just those phenomena which, to superficial observation, 

 range themselves in independent groups are chosen for the dignity of 

 being styled separate characters. Why A = B, is a question that 

 needs little discussion if higher up the page you have defined A solely 

 on the ground of that equation. If this be not the principle of 

 definition of a character, why consider the length of right and left 

 legs not two relations but one ? If they ever differed seriously in 



