XVI INTRODUCTION. 



to the color only a secondary importance, on account of lis fre- 

 quent variableness. However, there is nothing absolutely fixed 

 in nature ; the forms and the sculpture likewise vary within 

 certain limits. Therefore the descriptions can be only averages 

 deduced from a certain number of individuals. The description 

 should represent, as it were, the algebraical formula of the 

 species, or its ideal type. It is not required that this should tally 

 with the individual, but, on the contrary, that it should represent 

 the average of the characteristics of the ensemble of individuals. 

 But in practice, the description can never be so perfect, since it 

 is drawn from a certain number of individuals and not from the 

 ensemble of the individuals that represent the species. It is for 

 the reader to know how to seize the connection that exists between 

 the description and the heterogeneous individuals which he 

 may have beneath his eyes. In a word, my method of description 

 aims above all at generalizing, and requires that the reader should 

 generalize likewise. It can hardly suit the amateur inclined to 

 lose himself in a multiplicity of details, for whom the collection 

 takes the place o.f nature, and for whom the determination of an 

 individual is the final purpose of the study of a species. 



From the principles just laid down, it follows that in the extreme 

 subdivisions of genera, I have usually preferred the characteristics 

 taken from the form to those taken from the color. Undoubtedly 

 it is less convenient for the reader, for the natural method is 

 always less easy to follow than the empirical systems ; never- 

 theless, I think that it is preferable to proceed in that wise, fur 

 whatever may be done to seek the natural method, a large portion 

 of empiricism is sure to remain, as I shall endeavor to show in 

 the study concerning the filiation of the species. We cannot, 

 therefore, abstain too carefully from classification of empirical 

 elements. 1 



' It is necessary to observe on this head that no absolute rule can be 

 laid down as to the subordination of characteristics. To be sure, forms 

 varying less than colors, they offer, in general, characteristics more im- 

 portant than the latter; but there is, however, now and then a case in 

 which the colors are more fixed than certain forms, and assume a real 

 importance— for instance, as being the stamp peculiar to a certain geo- 

 graphical zone. Tims, the division Hypodynerus (semis Od/jnerus), which 

 depends greatly on the colors and facies, and which oomprises the most 

 divergent forms. In this case, the livery becomes the stamp of a fauna, 



