i897. THE GENEALOGY OF THE SCIENCES. 401 



history ; for no classification can be rational unless it is founded on 

 that genealogical tree which the history of science ought to furnish. 

 The progress of science dates only from the middle of last century. 

 It was then only that the knowledge of mathematicians and 

 philosophers became grouped in distinct sciences. The history of 

 science then is recent, because, if need be, we are able to begin with 

 the encyclopaedists. And from our point of view, the more modern 

 part of this history interests us the most. 



A classification to be useful to a bibliographer should be com- 

 plete. The natural classification that we demand from history should 

 go further, in the dismemberment of subjects of study, than do the 

 sciences themselves. And it is modern history which tells us how 

 each science is subdivided into sections, embracing each subject of 

 particular study. The history of each science considered separately 

 can go into great detail; it can give numerous data, and as the annals 

 of the sciences are very precise, this history can give a classification 

 down to the minutest details desirable. Thus then each science may 

 be represented by a bundle of branching twigs, upon which we can 

 group with certainty the material to be classified. And in this way 

 we can collect everything that ought to be collected, and classify it, 

 without there being any particular difficulty in avoiding confusion. 



But all is not to do, for that which is done persists. Imperishable 

 monuments have been raised to science. Houzeau and Lancaster's 

 Bibliography of Astronomy is one of them. It is to be noted that 

 this Bibliography is based on history ; this is more than a series of 

 biographies, it is a history full of observation, a model. 



Others than Houzeau have written on similar lines. Each 

 science perhaps has its historians.^ The Germans and the English 

 have written more than one history of the mathematics, of physics, of 

 chemistry, and other sciences. It is not for us to criticise them ; but 

 one may remark that, though precious, they are of no use for our pur- 



cette histoire. Le systeme general des sciences et des arts est une espece de 

 labyrinthe, de chemin tortueux, 011 1' esprit s' engage sans trop connaitre la route 



qu' il doit tenir On pourrait former 1' arbre de nos connaissances en les divisant, 



soit en naturelles et en revelees, soit en utiles et agreables, soit en speculatives et 



pratiques Nous avons choisi une division qui nous a paru satisfaire tant a la 



fois le plus qu' il est possible a 1' ordre encyclopedique de nos connaissances et a leur 

 ordre genealogique. Nous devons cette division a un auteur celebre (Bacon) .... 

 nous avons pourtant cru y devoir faire quelques changements." Further on we read : 

 " Ces trois facultes (memoire, raison, imagination) forment d' abord les trois divisions 

 generales de notre systeme et les trois objets generaux des connaissances humaines, 

 I'Histoire qui se rapporte a la memoire ; la Philosophie qui est le fruit de la raison, 

 et les Beaux Arts que I'imagination fait naitre." That is how d'Alembert was 

 brought to put the history of minerals and plants by the side of civil history, and to 

 unite metaphysics with mathematics. But his errors concern us Httle ; he did the 

 best for his time. These extracts have not been made to criticise him, but to render 

 him his due. The idea of a classification of sciences based on such a genealogical 

 tree was clearly formulated in 1781. 



1 See Jahresberichte der Geschichtswissenschaft,— §70 of the last few volumes 

 especially. 



