72 POPULAR SCIENCE MONTHLY. 



ing scheme fitted to the new state of knowledge developed. The 

 method is specially applicable to elaborate inquiries, particularly those 

 in which the premises are imperfect and a long logical chain is hung 

 upon them. The discussions of our great fundamental conceptions fur- 

 nish the best examples, chiefly examples of the lack of a systematic re- 

 generative method. Among these, two general classes may be recog- 

 nized, (1) those of a rather rigorous type, as, for a distinguished 

 example, the researches of George Darwin on tidal reaction and the 

 history of the earth and moon, and (2) those of a looser and some- 

 times rather metaphysical type, which I shall try to illustrate by the 

 doctrine of determinism. In all cases, assumptions are made the basis 

 of the procedure. Absolute premises are not available. Taking its 

 start from these assumptions, the process pursues a long course and at 

 the end conclusions of great import are often drawn. Usually the 

 process rests there, and in this lies a serious shortcoming. It should 

 give rise to a new process of a higher order. Not seldom, a critical 

 study of the results will reveal features that were not recognized nor 

 suspected in the original assumptions, though really there. Out of 

 these revelations should grow new assumptions and a new process. 

 The second conclusion may in like manner betray unsuspected qualities 

 and these should beget still other assumptions, and so the procedure 

 should continue until the field is exhausted. 



To choose a specific illustration is not a little delicate, for to be 

 most familiar it must be of the negative type, but I fear such an 

 illustration is the only way to clearly convey the meaning here in- 

 tended. I therefore venture to choose one so eminent and so admirable, 

 even with its limitations, that any suggestion of shortcoming will in no 

 wise dim the luster of a great achievement. In the classic investiga- 

 tions of George Darwin on tides and their astronomic consequences, a 

 viscous earth is assumed as the starting-point, with properties such that 

 the tidal protuberance is carried forward by the rotation of the earth 

 to the point which gives the maximum effect on the motions of the 

 earth and moon. These assumptions run potentially through the 

 whole train of brilliant mathematical deduction. At the end of the 

 inquiry, or if not of this particular inquiry, at least of collateral in- 

 quiries, the conclusion is reached that the earth is a rigid body com- 

 parable to steel. Between such a rigid body and such a viscous earth 

 as was assumed at the outset of the inquiry, there is a seeming incon- 

 gruity. This, under the regenerative method, suggests a new investiga- 

 tion on the assumption that the earth is a very rigid body, with the 

 further assumption that it has high elasticity of form, such that its 

 protuberance may perhaps not be carried forward to the degree pre- 

 viously postulated. These new assumptions are the more imperative, 

 because they are supported by inquiries based on quite independent 

 lines. In framing the new hypothesis an advance in detail and in 

 organization is to be sought on the evolutionary principle already in- 



