METHODS 37 



ing to the human mind should be chosen by our ear in col- 

 lecting sounds. For the sake of consistency in this ques- 

 tion of methods, I will even say that, although Pierre Bon- 

 nier's theory seems acceptable to me, it is possible that the 

 ear uses neither the method of resonators nor that of register- 

 ing sonorous forms, but some third method not yet dis- 

 covered which, though perfectly adapted to our auditive 

 mechanism, would probably be of little use in teaching / 

 acoustics in school. 



Second Example : Analysis of a Chemical Compound 

 into its Elements or Functions. 



In the study of a chemical compound the end proposed 

 may be either to describe it so completely as to be able to 

 recognize it afterwards wherever and whenever it is found, 

 or to characterize its behaviour toward this or that other 

 compound previously chosen. 



In the first case, any and every means of analysis will 

 surely be good if exact. You measure the density of the 

 body you study, its boiling and freezing temperature ; you 

 estimate as closely as possible its colour, taste and smell ; 

 you make its chemical analysis into its elements and estab- 

 lish how much carbon it contains, how much sulphur, how 

 much hydrogen, and so on. In a word, you gather together 

 all the characters you need ; each may be useful in a given 

 case, especially in pharmacies where the only need is to be 

 sure of the nature of a product. But if you only gather 

 up in this chance way all the descriptive characters of the 

 body you are studying, you will in no wise be fitted to 

 foresee how the body will behave in presence of sulphuric 

 acid, for example, for the analysis of the body from the 

 sulphuric-acid point of view has probably nothing in common 

 with its analysis from the human point of view. On the 

 other hand, a very simple process will enable you to resolve 



