374 Dr. Dorothy "Wrinch and Dr. H. Jeffreys on Certain 



of testimony have been discussed. Science therefore rests 

 on individual judgments, and so far is similar to mysticism 

 or art ; the judgments that electric potential satisfies 

 Laplace's equation and that ' The Magic Flute ' is a 

 superior work to w The Bohemian Girl ' are distinguished 

 by our own feelings about them and not by any external 

 criterion. Even the a priori postulates, in their actual 

 application, are individual judgments. 



On certain important Principles in the Theory of 

 Scientific Knowledge. 



Before we proceed to indicate the nature of the postulates 

 involved in scientific knowledge, it will be useful to point out 

 that the results of empirical science fall into two divisions. 

 The first of these is purely descriptive ; the second involves 

 the use of induction. Much confusion exists about the 

 extent and status of these sections — some writers holding, 

 like Kirchhoff, that all science is descriptive, others (and 

 this class includes many idealist philosophers) maintaining 

 that all science is inference, and invalid inference at that. 



Now description is a term with a wide application. There 

 are senses in which a statement that one sensation is different 

 from another is a description ; others in which a statement 

 about the shape of a leaf is one ; and others, again, in which 

 only a general proposition such as the law of attraction 

 under which the planets move about the sun can be called 

 a description. All these types of proposition are necessary 

 in a scientific review of the world, and each is more general 

 than the preceding. Without the particular propositions 

 no general descriptive proposition can be given any weight. 

 Without the general propositions no particular propositions 

 can be given any comprehensive significance. Some writers 

 do not make the distinction between these kinds of propo- 

 sition that may be called descriptions sufficiently clear. It 

 is, however, of the utmost importance in a theory of scientific 

 knowledge to keep them distinct, for they are quite different 

 in origin and logical structure. The particular propositions 

 must, of course, precede induction. But in themselves they 

 constitute only a very small part of science as we know it. 



Our point may be illustrated by a crude example, which 

 we shall not attempt to analyse into the primitive elements,, 

 empirical and logical, which it really involves. Suppose we 

 have a disk which can roll down an inclined plane, and that 

 we have some mechanism that enables us to record its position 

 every fifth second from the start. Suppose it is found that 



