130 THE AIM AND ACHIEVEMENTS OF SCIENTIFIC METHOD. 



of nervous tissue, the notion belongs just as clearly to the third 

 class to be considered below. 



62. 



In the second type of hypothesis^ the elements which are 

 added to make the secondary constructions are not spatio- 

 temporal existences but relations between such existences. 

 Such an hypothesis was Newton's belief that the attraction of the 

 earth for the moon could be calculated from its attraction for a 

 stone on the earth's surface in accordance with the law of 

 inverse squares, or Joule's conviction, maintained for years in 

 spite of contradictory experimental results, that a definite 

 equivalence existed between heat and work. Hypotheses of this 

 kind share with the former type the characteristic of being, 

 at least ideally, verifiable. 



63. 



In the last class we find the typical hypothesis of science as 

 opposed to the hypothesis of history and common sense, the 

 hypothesis which Ostwald has attempted to banish from 

 scientific method. In general, its marks are (1) a lack of the 

 homogeneity between the data and the added' or interpolated 

 elements which characterised the first type ; (2) the unverifiable 

 character of the added elements; and (3) that the secondary 

 construction does not merely complete the data but actually 

 replaces them. 



The first two of these marks, at least, are present in the 

 case of the hypothesis of heat, by means of which temperature 

 changes are explained. The entity which is thought of as 

 " flowing " from the hot body to the cooler body is not thought 

 of as of the same order of existence as the sensations of 'hotness 

 and coldness which are the actual data here ; and quite 

 obviously it is completely unverifiable no one has ever 

 pretended to exhibit heat apart from the phenomena of hotness 

 and coldness which it is invoked to render intelligible. 



