52 'WHAT IS A SPECIES?' 



urgent ncccssit)' fur llic study of structure in other groups, 

 render them better instruments of zooloLrical education. 

 Init altliough the Lepidojjtera are inferior in this respect, 

 aUhough they lack the unique interest of the I Iymenoj)tera 

 and the social Neuroptera, and cannot claim any of the 

 respect due to venerable age like the Ai)tera, C)rthoptera 

 and NeurojUera — in spile of their many demerits they 

 stand at the head, not only of all insects, but of the whole 

 organic world, as the registers of subtle and elusive 

 change — ever going on, )et never seen, — by means of 

 which forms are slowly becoming different from what 

 the)- have been in the past. It is the existence of a com- 

 plex pattern composed of several colours, which renders 

 butterflies and to a less extent moths such a remarkably 

 delicate record of change. As we trace the representative 

 individuals of a community of butterflies over an)- wide 

 range, the trained eye, and often the inexperienced eye, 

 can detect differences which are not seen to anything like 

 the same extent in the individuals of other Orders with 

 corresponding ranges. If the wings of Ilymenoptera, 

 Diptcra, or Orthoptera possessed the same elaborate 

 l)atterns as the Lepidoptera, we cannot doubt that they 

 too would exhibit the same differences in various parts of 

 their areas. The continual chancre which we fmd as 

 we study the distribution of Lepidopterous forms in space, 

 is undoubtedly a measure of the speed with which evolu- 

 tion has occurred in time. Raijidit)- of change is essential 

 if it is to keep its adjustment with nicety to the fleeting 

 details of distribution.' Hence we may confidently 



' 1 1 is to l)c observed ihat I speak of the dchiils as fleeting. The 

 general area of distribution is doubtless extremely ancient in most cases. 

 Thus, although a species of Hdiionius, <fcc., may liave originated within 

 the South American Irojiics, and never have wandered beyond them, the 

 complex shape of its actual area of distribution at anyone time cannot be 

 regarded as fixed or ancient. Yet in many a species the variation of the 

 constituent individuals is adjusted with precision to the geographical details 

 of the existing range. 



I\Ir. Roland Trimen, on reading the above footnote, writes to me 

 January 24, 1904 : — ' Your note reminds me of the recent appearance on 

 the Natal coast of several conspicuous Kast- African butterflies, lid. : Picris 

 spilUri, CrcTiis rosa, and Godartia ivakefieldii^ all of which are shown to 

 have not only extended their range to a point where they were previously 



