102 REPORT— 1876. 



variation of colour than is generally suspected ; but tliore are evidently other causes 

 at work, and one of these seems to be an influence depending strictly on locality, 

 whose nature we cannot yet understand, but whose effects are everywhere to be 

 seen when carefully searched for. 



Although the careful experiments of Sir John Lubbock have shown that insects 

 can distinguish colours — as might have been inferred from the brilliant colours of 

 the flowers which .are such an attraction to them — yet we can hardly believe that 

 their appreciation and love of distinctive colours is so refined as to guide and regu- 

 late their most powerful instinct — that of reproduction. We are therefore led to 

 seek some other cause for the varied colours that prevail among insects ; and as 

 this variety is most conspicuous among butterflies — a group perhaps better known 

 tlian any other — it offers the best means of studying the subject. The variety of 

 colour and markingamong these insects is something marvellous. There are probably 

 about ten thousand different kinds of butterflies nowlaioviTi, and about ban of these 

 are so distinct in colour and marking that they can be readily distinguished by this 

 means alone. Almost everj' conceivable tint and pattern is represented, and the 

 liues are often of such intense brilliance and purity as can be equalled by neither 

 birds nor flowers. 



Any help to a comprehension of the causes which may have concurred in bring- 

 ing about so much diversity and beauty must be of value ; and this is my excuse for 

 lajing before you the more important cases I have met with of a connexion between 

 colour and locality. 



Our first example is from tropical Africa, where we find two unrelated groups of 

 butterflies belonging to two very distinct families (Nymphalidre and Papilionidaa) 

 characterized by a pi-evailing blue-green colour not found in any other continent *. 

 Again, we have a group of African Pierida} which are white or pale yellow with a 

 marginal row of bead-like black spots ; and in the same country one of the Lycfenidne 

 (Leptena erasius) is coloured so exactly like these that it was at first described as 

 a species of Pieris. None of these four groups are known to be in any way speci- 

 ally protected, so that the resemblance cannot be due to protective mimicry. 



In South America we have far more strilring cases ; for in the three subfamilies 

 DanaincB, Acrames, and Helicominee, all of which are specially protected, we find 

 identical tints and patterns reproduced, often in the greatest detail, each peculiar 

 type of coloration being characteristic of distinct geographical subdivisions of the 

 continent. Nine very distinct genera are implicated in these parallel changes — 

 Lycorea, Ceratinia, Mechanitis, Ithomia, MeUiKea, Tithorea, Acraa, HeUconiiis, and 

 Eueides, gi-oups of three or fom- (or even five) of them appearing together in the 

 same livery in one district, while in an adjoining district most or all of them undergo 

 a simultaneous change of coloration or of marlnng. Thus in the genera Ithomia, 

 Mechanitis, and Heliconius we have species with yellow apical spots in Guiana, all 

 represented by allied species with white apical spots in South Brazil. In Mecha- 

 nitis, 3Ielin(ea, and Heliconius, and sometimes in Tithorea, the species of the 

 Southern Andes (Bolivia and Peru) are characterized by an orange and black livery, 

 while those of the Northern Andes (New Granada) are almost always orange-yellow 

 and black. Other changes of a like nature, which it would be tedious to enumerate, 

 but which are very striking when specimens are examined, occur in species of the 

 same groups inhabiting these same localities, as well as Central America and the 

 Antilles. Tlie resemblance thus produced between widely different insects is some- 

 times general, but often so close and minute that only a critical examination of 

 structure can detect the difference between them. Yet this can hardly be true 

 mimicry, because all are alike protected by the nauseous secretion which renders 

 them unpalatable to birds. 



In another series of genera {Catagramma, CaUithea, and Ai/rias), all belonging to 

 tlie Nymphalidse, we have the most vivid blue ground, with broad bands of orange- 

 crimson or a different tint of blue or purple, exactly reproduced in corresponding, 

 yet unrelated species, occui-iing in the same locality ; yet, as none of these gi-oups 

 are protected, this can hardly be true mimiciy. A few species of two other genera 



* Bo7nalcosoina and Euniphcne (Nymphalidre), PajMlio s:almo.ris and several species of 

 the iViVfws-group (Papilionid.-B). 



1 



