1890 ] 251 



CLASSIFICATION. 



BT A. F. GEIFFITH, M.A. 



It has become rather the fashion of late years among men of 

 science to depreciate the value of classification in comparison v/ith 

 the study of morphology and the like. But it must be remembered 

 that careful classification lies at the root of all accurate knowledge of 

 any branch of Zoology, so that no labour can be thrown away which 

 is expended in obtaining correctness in this fundamental portion of 

 science. 



The visible outcome of all such labour is a catalogue, and it is 

 the object of this paper to ascertain certain principles applicable to 

 Natural History catalogues in general ; which principles, if correctly 

 ascertained, may with great advantage be applied to obtain a standard 

 catalogue of Lepidoptera, and put an end to the annoying, and in 

 many cases frivolous, changes with which we unfortunate collectors 

 have so frequently to put up in the nomenclature and arrangement 

 of our insects. 



Botanists speak of the " Natural System " of classification as opposed to earlier 

 systems or wants of system, and know the value to their science of the great advance 

 in knoveledge, which is crystallized in that expression. We Lepidopterists can 

 scarcely be said to have known any other than a Natural System. But how is a 

 Natural System to be reduced to meet the iron necessities of a catalogue ? It will 

 be noticed immediately on looking at a list that any species is thereby connected 

 with two others only ; namely, that immediately preceding, and that immediately 

 following. If, therefore, it were thought necessary to indicate the relationship of 

 one species with any other than these two, it would be necessary to use some system 

 of cross lines, reference letters, or the like, which would infallibly break into the 

 order of the catalogue, and render it less useful for all other practical purposes. 

 The same remark holds good with regard to the representation of relationships of 

 genera and other larger groups. Now, do these relationships in Nature conform to 

 the rule which the catalogue maker, under the stress of this difficulty, would fain 

 lay down for them ? The answer is clearly, No. Each collector will readily bring 

 to mind species and genera which show close relationship to more than two others. 

 This being so, it is as absolutely impossible to represent accurately by a catalogue 

 the whole of a Natural System as it is to represent on a plane the relative positions 

 of points on a non-developable surface. Hence we learn that it is idle to look to 

 natural affinities alone to fix the order of the standard catalogue. If some recognised 

 authority were to-day to attempt to formulate such a catalogue from such materials 

 only, it would be open to any one to-morrow to point to affinities which the first 

 authority had perforce to overlook for the purpose of his list, to make up his mind 

 that these neglected affinities were more important than those relied on by the first 

 (for the closeness and remoteness of such affinities must, to a very great extent, be a 

 matter of opinion j who is to decide whether most reliance is to be placed in any 



