I'll 1C NORMAL SKLF, ETC. 497 method that it should actually be preceded by ' definitions ' so-called. In other words, the methodological assumption^, from which and by the help of which it proceeds to grapple with the unknown that is to say, its ' leading questions,' which as presumably intelligent questions must be already answered in part are most naturally and succinctly ex- pressed in the shape of a certain number of tentative formula! giving precise conceptual expression to wider or narrower pieces of alleged fact. Further, the ' answers ' of the science which are actually answers in point of form, namely its much-prized generalisations or discoveries, must themselves in turn be ' defined,' that is, reduced to a relatively deter- minate and ' workable ' consistency, if they are to serve as the stepping-stones to fresh discovery. Here, then, is a twofold need for the aid of the Method of Definition. It may be objected that neither the question-answers nor the answer-questions thus temporarily posited are definitions in the strict sense ; that on the contrary they are mere descriptions ' generalisations,' if you will, but never ' universalisations '. That is quite true. At the same time evolutionary science has no cause to abandon the search after elSo? for any such reason. Thus if the votary of Formal Science claim that he alone is able to put this ideal to a really successful use, the retort is handy that he attains to his boasted consistency merely by a magnificent effort of abstraction by ignoring once for all the fact that his thought is the outcome of a process of thinking. Moreover he ought to be the last person to deny that the formulatory canon employed in Abstract and Concrete Science alike is not merely similar but identical. Let the correlativity of the two great methods of research be therefore admitted, and made the ground of a resolve to combine their use so far as such combination may be possible. Evolutionary Science must be allowed its formulae or definitions (in the qualified sense), and its Methodology must recognise a (would-be) Formal Part the function of which is to give stability and concentration to the discursive speculations of its (never merely) Material Part. Now Ethics is, or ought to be, a concrete science in the fullest sense. It is certainly no mere branch of Introspective Psychology, however much the academic mode of treatment till lately in vogue may have done to popularise the notion ; nor indeed if it were to be identified as a science with its introspective side could it be termed ' abstract ' in the sense that Formal Logic is abstract, that is, rendered independent of process once for all. On the contrary, it is on the face of 32