8 Biographical Memoir of' Michel Jdansofi. 



glimpse of some portions of this sublime picture ; and the point 

 from which we might embrace the whole is still but a sort of 

 ideal object, which we may perhaps never attain at all, although 

 ft is our duty unremittingly to tend toward it, and although, by 

 continued labour, we may every day approach nearer to it. 



The most direct method would be to determine the functions 

 and influence of each organ, in order to calculate the effect 

 of its modifications ; then, forming the great divisions according 

 to the most important organs, and thus descending to the lower 

 divisions, we should have a scale, which, although formed in 

 advance, and almost independently of the observation of species, 

 would nevertheless be the real expression of the order of nature. 

 It is this principle which is named the subordination of the cha- 

 racters. It is perfectly rational and philosophical ; but its ap- 

 plication would suppose a degree of knowledge with regard to 

 the nature, functions, and influence of organs, which, at the pe- 

 riod when M. Adanson commenced his labours, was too far from 

 being attained, even in approximation, to admit of being em- 

 ployed ; and, perhaps, even the idea of it never presented itself 

 to his mind. 



He, therefore, had recourse to a method the reverse of this, 

 which may be called the empirical method, or that of experiment, 

 founded upon the actual comparison of species ; and, in order 

 to apply it, he devised a plan which is peculiar to himself, 

 and which cannot but be regarded as highly ingenious. Consi- 

 dering each organ separately, he formed of its different modifi- 

 cations a system of division in which he arranged all the beings 

 known. Repeating the same operation with relation to many 

 organs, he thus constructed a number of systems, all artificial, 

 and founded each upon a single organ arbitrarily chosen. 



It is evident that the beings which none of these systems 

 would separate, would be very intimately allied, since they would 

 resemble each other in all their organs. The affinity would be 

 somewhat less in those which some systems would not assimilate 

 in the same classes. Lastly, the most distant of all would be 

 those which would not come together in any system. 



This method would, therefore, afford a precise estimate of 

 the degree of affinity of beings, independent of the rational and 

 physiolo^cal knowledge of the influence of their organs ; but it 



