" TYPES " l\ NATL'KAL HISTORY. 201 



bases to choose tVom, but by this time we have universally agreed — 

 have we not ? — to accept some theory of evolution, and, consequently, 

 phylogeny is the only basis on which a satisfactory classification can 

 be built. It follows then, that if we are to have a (iasxificatinn as 

 opposed to a mere nomendatnre, the " recognition " theory of names 

 will not satisfy us, but we require groupings more or less compre- 

 hensive, the genus, the tribe, the family, etc. ; and further, these 

 groupings must be made on the facts of evolution as they gradually 

 come to light, one of the consequences of which is that generic, tribal 

 and family nomenclature must continue, at whatever inconvenience to 

 collectors, to be in a somewhat unfixed condition " until omniscience 

 is attained." So far, probably, my critics and I are at one, but at this 

 point we join issue. They would say (I hope I am not misrepresenting 

 them) : " Yes, this is all very well until you come to the species, but 

 at this point the ' recognition theory,' as you call it, becomes more 

 important, and unless we agree to call every individual of a species by 

 the natae given to the first-cajDtured, or rather the first-named 

 specimens known, and to condder tliis form the type of the species, fixity 

 of nomenclature will never be obtained ; and this must be carried out 

 even to the origmal spelling of the name, however incorrect or 

 grotesque, in utter defiance of orthography and grammar, and wi 

 mniproiiiise is possible." (I repeat that I hope I am not misrepresenting 

 my critics, and I ofier the most heartfelt apologies if I am doing so, 

 for such a position seems almost inconceivable when gravely attributed 

 to such logical and scientific workers, and I am almost afraid I shall 

 be accused of trying to give a comic turn to the discussion ; quite 

 honestly, nothnig is further from my intentions, I consider the matter 

 too important.) This theory would be less open to attack, or perhaps 

 might, in spite of certain inconsistencies, have escaped attack 

 altogether, if it had been universally agreed to disallow all varietal and 

 aberrational names, and to describe all such forms as i, ii, iii, or a, /i, 

 7, etc., but the moment varietal and aberrational names are admitted, 

 the species ceases to be the lowest term in use, and assumes towards 

 its varieties and aberrations a position somewhat, though not wholly, 

 analogous to that occupied by the genus towards the species, and 

 logically the same rules must be, as far as possible, applied. Let me 

 take a possible, though perhaps not very probable, instance. Suppose 

 that a pair of some new species is taken and named. Later, a number 

 of $ s are taken in some other locality, and also a few 3 ^, the latter, 

 the season being nearly over, in a worn condition. The ? s are 

 apparently identical with the first taken (such cases occur among 

 the Lyctenids), though the worn <? s difler somewhat from the original 

 specimen, which was fresh. On the strength of the apparent identity 

 of the ? s it is taken for granted that the two sets of specimens belong 

 to the same species and the original name is applied to all. For some 

 years the original locality remains unvisited, but numerous examples 

 corresponding with the second set of specimens are taken in various 

 localities, and the original 3 is then supposed to be an unusual form. 

 After some time, anatomical investigation, or the discovery of the 

 earlier stages, shows that the two are specifically distinct. What about 

 the name then ? It belongs of right to the original pair only, until 

 others of the same form are discovered, and the name of all other 

 specimens must be changed. As "convenience" and "fixity of 



