RELATIONS TO OTHER SCIENCES 77 



theory. The significance of the arrangement of atoms in the molecule 

 in determining chemical reaction was emphasized and developed by 

 Kekule", but it was not until 1874 that the space diagrams of mole- 

 cules of van't Hoff and Le Bel marked a full appreciation of the 

 possibilities of structure in explaining the differences of isomeric 

 forms. 



All of these physical and chemical developments of the atomic 

 theory have been in accordance with a general method of scientific 

 procedure which may be called the method of mechanical models. 

 According to this method, an attempt is made to conceive a certain 

 mechanism by which the various phenomena sought to be explained 

 may be imagined to be brought about. 



Such a theory of atoms, for example, if perfect, would exhibit all 

 the properties of atoms as direct consequences of the assumed struc- 

 ture. This cannot, however, be taken as proof that the assumption 

 is real, though for the purpose of our thinking such a theory would 

 have all the value of reality, since all consequences deduced from it 

 would conform to the facts of observation. And this suggests wherein 

 the great value of such a theory lies, not alone in the large number 

 of observations which it correlates and brings under a few general 

 principles, but in that it suggests the application of experiments 

 and tests of its sufficiency, thereby enlarging and making more pre- 

 cise our knowledge. 



Perhaps the most remarkable instance of the application of this 

 method was Maxwell's development of a mechanical model to illus- 

 trate the reactions in the electromagnetic field. Working from this 

 model he developed the equations of the field, which later he deduced 

 in a more general way. And Hertz speaking of them says, "We can- 

 not study this wonderful theory without at times feeling as if an 

 independent life and a reason of its own dwelt in these mathematical 

 formulae; as if they were wiser than we were, wiser even than their 

 discoverer; as if they gave out more than had been put into them." 



On which Boltzmann's comment is, "I should like to add to these 

 words of Hertz only this, that Maxwell's formulae are simple conse- 

 quences from his mechanical models; and Hertz's enthusiastic praise 

 is due in the first place, not to Maxwell's analysis, but to his acute 

 penetration in the discovery of mechanical analogies." Such an 

 example well illustrates the importance of the method. 



But of recent years, the influence of quite a different method has 

 been strongly marked in chemical research. A method in which 

 certain general laws are established and then applied to particular 

 cases by a process of mathematical reasoning, deducing conclusions 

 quite independently of the particular details of the operation by 

 which they are brought about. This method is well illustrated in 

 Professor J. J. Thomson's work on the application of dynamics to 



