THE UNITY OF SCIENCE 19 



continuous laws about the one general sort of phenomenon 

 namely, the relative motion of matter. To put the distinction 

 between the two modes of discontinuity concisely, we may say 

 that two laws may be discontinuous in the sense that one does 

 not follow from, though it conforms to, the other; or in the 

 sense that (the two relating to a single general kind of phe- 

 nomenon) the one neither follows from nor conforms to the 

 other. This distinction has been too much disregarded in dis- 

 cussions about the unity of science. Thus Mr. Herbert 

 Spencer, who perhaps had more than any other nineteenth 

 century writer to say about the unification of knowledge, was 

 wont to imagine that he was exhibiting the unity of all classes 

 of phenomena by showing merely that they conformed to a 

 single law. That law, as he conceived the matter, was pre- 

 cisely the principle of the conservation of energy; this was 

 for him "the highest generalization which is true not of one 

 class of phenomena but of all classes of phenomena, and 

 which is thus the key to all classes of phenomena." But if 

 you will examine the way in which he employs this "key," I 

 think you will find that what he really proves concerning most 

 of the lower generalizations is, not that they are deducible 

 from this "universal truth," but only that they do not violate 

 it. 



It is by this time, I hope, fairly clear what one means or 

 ought to mean by the unification of science, on the one hand, 

 and by the discontinuity of scientific laws, on the other 

 notions by no means so simple and obvious as they are some- 

 times supposed to be. After so much explanation, we should 

 be prepared to appreciate the interest and significance of the 

 scientific movement of our time in its relation to the age-long 

 struggle after a complete unification of the sciences. That in- 

 terest and significance will be still more apparent if we very 

 summarily review the past history of this struggle. In such a 

 review, it will be necessary to use as illustrations some details 

 of the special sciences which will, perhaps, not be altogether 



