BY R. M. JOHKSTOX, F.L.S. 39 



where the definition of a species is based upon the observatiou 

 of a large number of specimens from different localites. I do 

 not use the words species as the type of a group of allied 

 organisms which have a rigidly determinate numljer of 

 immutable characteristics in common ; for the characters 

 which, as a whole, are relatively constant in those sections 

 which we group under a specific name are themselves variable, 

 and are frequently to be found interlapping other groups 

 of merely relative constant characters, but which we yet 

 acknowledge as belonging to a distinct species. 



The type of a group termed species is fixed upon mainly to 

 define the maximum of relatively constant characteristics 

 around which all the individual varieties may cluster, and 

 which shall serve to distinguish the type species from a closely 

 allied group of a similar character. Indeed, we may picture 

 species as the nodes of an irregularly moniliform series, 

 whose extremities are in some cases sharp and distinct, and 

 in other cases mere constrictions, whei'e the extreme indi- 

 viduals of each node or group meet, and can hardly be dis- 

 tinguished from each other. But even when we clearly 

 understand, and agree with each other as regards the pi'in- 

 ciples which determine classification, it is often perplexing to 

 fix upon characters whereupon to erect the standard of a 

 species or variety, for it is well known in practice that 

 characters are seized upon rather from stability and associa- 

 tion with certain other charactei's than from absolute 

 differences in particular features. Gwyn Jeffreys thus 

 defines the degrees of difference which should determine 

 species : — " They constitute more or less extensive groups of 

 individuals which resemble each other as well as their parents 

 and oft'spring to the same extent as we observe in the case of 

 our own kind. These groups to deserve the name of species 

 must be distinct from others : because, if any of them are so 

 intimately blended together by intermediate links, so as to 

 make the line of separation too critical, the test fails, and a 

 subordinate group, or what is called a ' variety,' is the result. 

 For this reason it is indispensably necessary to compare as 

 great a number of individuals as possible, and especially a 

 series of different ages and sizes, commencing ah ovo, as well 

 as specimens collected from various localities." And again, 

 he states in respect of what are termed varieties, that " the 

 characters by which they usually differ from species consist of 

 size, comparative proportions of different parts, colour, and 

 degree of sculpture ; " and he remarks that such differences 

 "originate in some peculiarity of climate, situation, composi- 

 tion of soil or water which they inhabit, the nature or supply 

 of food, and various other conditions." These latter, he 

 adds, may be " permanent or local." When permanent he 



