92 J. MURRAY 



one even as of generic rank. Had the early practice been followed, whereby species 

 of the genus Echiniscus were supposed to be sufficiently characterised if they differed 

 by one pair of spines or setoe from the related species, or if even a seta were replaced 

 by a spine or vice versa, a hoat of new species would have been added. I prefer a 

 definition of species in that genus which excludes such forms, and requires species to 

 show peculiarities in other points of structure than the spines (such as claws, or plates, 

 or texture), or to have the spinous armature conspicuously different from that of the 

 nearest relatives. 



No definition of species can be made which will be universally applicable. In 

 some groups it is necessary to distinguish species by characters so slight that they 

 would not be admitted in other groups. Species corresponds to no entity in Nature 

 it is a human convenience. We may seek to distinguish by it such races of any 

 organism which have through some degree of isolation and the tendency to vary 

 (under influences of which we are profoundly ignorant) acquired more or less distinct 

 marks by which they may be distinguished from the related races of the same stock, 

 and which we believe to be constant (in the sense that they continually recur, and 

 that we believe the recurrence to indicate community of origin). The amount of 

 distinctness considered necessary to constitute specific value gives an opportunity for 

 the personal element in naturalists to assert itself, and has, incidentally, given occasion 

 for much disputation. 



Every species throws some light on the biological history of the world, and so we 

 want to discover them all. Such cheap species as the ardent discoverer could so 

 easily manufacture out of the myriad forms of Echiniscus would yield the minimum 

 of light on biological problems. In so far as they are permanent, they are only 

 fortuitous side eddies in the stream of evolution, if such a term can be per- 

 mitted a popular application in science. I mean that the species having two 

 spines and those having three, and not otherwise differing, are not instances of 

 essential adaptation to the conditions of life of the animals, but that one is as 

 good as the other. Further, the probability is that many of these slight forms are 

 really not species in even the widest acceptation of the term, but merely individual 

 peculiarities. 



Among Echinisci there are forms known, such as E. granulatus and E. quadri- 

 spinosus, var. cribrosus. '[see Murray (12) and (18)] which possess several good charac- 

 teristics, besides their distinctive spines. In colonies of such species, which may be 

 confidently identified by the concurrence of many characters, it continually happens 

 that individuals occur in which one or more of the customary spines is lacking. 

 That it is frequently only one spine of a pair which is absent indicates that these are 

 :nly errors or peculiarities in the development of the individual. Yet by the accepted 

 practice one pair of spines or seta) makes a species. 



There is next to be considered the case where these slight peculiarities seem to be 

 constant. Admittedly when example after example turns up, having all the trifling 



