:;50 



THE AMERICAN NATURALIST [ Vol. XLVII 



up a large part of the work of science; does the search 

 for other sorts of conditions form any part of its work? 



Men do indeed infer certain things confessedly not dis- 

 coverable by observation or experiment, but these evi- 

 dently deserve, and commonly receive, a classification by 

 themselves, as something else than science; otherwise 

 science itself would require division into experiential and 

 non-experiential, the former including what is commonly 

 practised as science. Our rule is the test for this classi- 

 fication; a question that could not be answered by any 

 conceivable experiment (or series of experiments) does 

 not belong to (experiential) science. 



But what does " conceivable experiment" include and 

 exclude? An experiment is a change in one or more of a 

 given set of conditions ; ideally carried out it involves the 

 presence of two similar systems, known to act in the same 

 way; on one of the systems a certain condition is then 

 altered, and the difference this brings about is observed. 

 In cases where this ideal can not be fulfilled, it forms the 

 standard for mental reference with relation to the ex- 

 periment as actually tried. The possibility of experi- 

 menting comes from the observed fact that conditions 

 which sometimes occur or act together need not always 

 do so. Xow, a proposition to separate such conditions 

 as are in the nature of things inseparable would not be 

 a conceivable experiment. Is such a proposition involved 

 in the question whether psychic processes affect physical 

 ones? 



But often a change in some one of a given or specified 

 set of conditions is conceivable where it is not practi- 

 cable. This may be for technical reasons; we have not 

 obtained control of the conditions. Or the system under 

 consideration may belong to past time. But in both these 

 cases, when we assert that a specified condition is the 

 cause of a certain result, we mean that if this condition 

 could be or had been altered, as is done in experimenta- 

 tion, the result would have been different. 



It is this mental reference to an experimental situation 



