CAUSE AND EFFECT— PROBABILITY 115 



motion ^ must be that which determines a sequence of 

 sense-impressions, or, in other words, it is the source of a 

 routine of perceptions. But the source of such routine, as 

 we have seen, Hes either in the field of the unthinkable 

 beyond sense-impressions, or else in the nature of the 

 perceptive faculty itself The " cause of change in motion " 

 thus either lies in the unthinkable or is a substantive part 

 of the machinery of perception ; in neither case can it 

 with any intelligible meaning of the words be spoken of 

 as a " dead agent." In the former case the cause of 

 change is unknowable, in the latter it is unknown, and 

 may long remain so, for we are very far at present from 

 understanding how the perceptive faculty can condition a 

 routine of perceptions. Science does not deal with the 

 unknowable, and if force be not unknowable, but unknown, 

 then mechanics as the science of force would as yet have 

 made no progress. The reality is indeed different from this. 

 One of the greatest of German physicists, Kirchhoff, thus 

 commences his classical treatise on mechanics ' : — 



" Mechanics is the science of motion ; we define as its 

 object the complete description in the simplest possible 

 manner of such motions as occur in nature." 



In this definition of Kirchhoff's lies, I venture to think, 

 the only consistent view of mechanism and the true con- 

 ception of scientific law. Mechanics does not differ, as so 

 often has been asserted, from biology or any other branch 

 of science in its essential principles. The laws of motion 

 no more account than the laws of cell-development for the 

 routine of perception ; both solely attempt to describe as 

 completely and simply as possible the repeated sequences 

 of our sense-impressions. Mechanical science no more 

 explains or accounts for the motions of a molecule or of a 

 planet than biological science accounts for the growth of 



^ We shall see reason in the sequel for asserting that "motion " is a con- 

 ception, rather than a perception — a scientific mode of representing change of 

 sense-impressions, rather than a sense-impression itself. In this chapter, 

 however, the term "motion" is used in its popular sense for a well-marked 

 class of sequences of sense-impressions. 



^ Vorlesungen ilber mathematische Physik. Band I. Mechanik, S. i. 

 Berlin, 1876. 



