August 3, 191 1] 



NATURE 



167 



mon whites. Then the scales are quite unlike plume-scales, 

 having neither fimbria; nor accessory disc, while the foot- 

 stalk is short and quill-like, instead of being long and 

 as in the plume-scales. They are, indeed, quite of 

 the type of the ordinary scales, except that they differ a 

 little in size and shape from the scales of their immediate 

 Hidings. The distribution of tracheae, or air-tubes, 

 to the site of the scent patch is noteworthy, and so also is 

 the fact that in ordinary circumstances the patch is 

 covered by the overlap of the forewing, which acts like a 

 sliding lid. It may reasonably be conjectured that this 

 arrangement ensures economy of the perfume. The pro- 

 duction of the scent is confined to a limited area, and its 

 escape is prevented under ordinary conditions by the over- 

 lapping edge of the forewing. When emission of the scent 

 is required, a slight separation of the fore and hind wings 

 givi - it exit. The special distribution of tracheae may be 

 a provision for pumping air into the patch from below, and 

 so supplying a vis a tergo to assist the escape of the 

 perfum._. 



Man; other butterflies possess similar patches of scent- 

 distributing scales. They are generally, though not always, 

 so placed as to be covered up in the ordinary position of 

 the insect. In some instances there are two patches on 

 each side, one on the upper surface of the hindwing, the 

 other on the under surface of the forewing, these being so 

 arranged that they exactly cover one another when the 

 butterfly holds its wings in the normal position. It is to 

 be observed that in these patches the rule is for the scent- 

 scales to be of the same general character as the ordinary 

 seals of the wing, though they may differ much from the 

 latter in shape, size, and arrangement. The patches seldom 

 contain plume-scales, and, when fitted with a sliding lid, I 

 felievi it would be correct to say that they never do. 



The question may be asked : is it not necessary that the 

 sceni should be economised in the case of the plume-scales 

 quite a< much as in the case of the definite patches? No 

 doubt it is ; and a little further consideration of the typical 

 plume-scale may show us how this is effected. The plume- 

 scales, it is true, being as a rule generally distributed over 

 the wing, cannot be shut down under a lid ; but they are 

 frequently scattered among ordinary scales which are a 

 good deal longer and larger, and which may act as cover- 

 ings to the individual scales, though there is no general 

 covering for the whole. Then again, as we have seen, the 

 plume-scale has an accessory disc and a long footstalk. 

 The disc with its internal chitinous structure may act as a 

 reservoir for the scent ; it will be remembered that in many 

 cases it appears to contain a convoluted tube. The foot- 

 stalk seems to be flexible, and it often shows one or more 

 sharp bends in its course. These bends may impede the 

 passage of the scent from the reservoir in the disc to the 

 lamina and fimbriae, and the butterfly may be able, by some 

 movement of its wings, to bring about a straightening of 

 the footstalk and a consequent liberation of the odour. At 

 any rate, it is probably significant that the apparatus of 

 accessory disc and long flexible footstalk belongs to the 

 jlume-scale alone. 



There is a handsome butterfly, common in Africa, which 

 I not far removed in affinity from our well-known " brim- 

 gone. ' This butterfly, which is known as Catopsilia 

 torella, has in the male a strong fragrant scent. Now 

 Catopsilia florella possesses on the hindwing a patch of 

 ipecial scales which is somewhat similar in aspect and 

 >osition to the scent-patch in the clouded yellow. But, 

 :uriously enough, the characteristic scent appears to pro- 

 ved, not from the patch on the hindwing, but from an- 

 ither structure altogether. This is a beautiful silky fringe 

 if long hair-like scales which are set along the edge of the 

 orewing on its underside, and are covered as a rule by the 

 iverlap of the hindwing. When the wings are parted and 

 he fringe spread out, a scent is exhaled which appears to 

 oth Dr. Longstaff and myself to be like that of jasmine 

 r tuberose. The portion of the wing which forms the seat 

 f this silky fringe, as well as that underlying the scent- 

 atch of the hindwings, is furnished, like the patch in the 

 louded yellow, with a plentiful supplv of air-tubes, pro- 

 eeding toit from the adjacent "vein." In the case of 

 he patch in Catopsilia florella. the ramifications of these 

 ir-tubes seem to form a fine polygonal network, each mesh 

 f which surrounds the socket of a scent-scale. A similar 



NO. 2179, VOL. 87] 



appearance may be seen in the scent-patch of one of the 

 l"eracoli (T. fausta), and probably in that of others. 



Sessile scent-patches, which may or may not be accom- 

 panied by silky fringes, occur in many other species of the 

 brimstone-like section of Pierines, and in several of these, 

 including both Asiatic and American species, a flowery 

 odour has been detected both by the late Mr. de Niceville 

 and by Dr. Longstaff. An observation which would be of 

 great interest, but which, so far as I am aware, has not yet 

 been made, would be to compare the odour diffused by the 

 fringes with that conveyed by the sessile patches in those 

 species where both these forms of apparatus occur together. 

 An accumulation of hair-like scales, no doubt serving as 

 scent-distributors, may also be seen in another genus of 

 Pierines (flismorphia) which is remote in affinity from 

 those butterflies which have just been considered. But 

 these odoriferous tufts or fringes are by no means confined 

 to the Pierines. In the very different group of Satyrines, 

 to which our common brown hedgerow butterflies belong, 

 the males of some species possess fringes or tufts which 

 are clearly similar in function to those of their distant 

 j cousins, the Whites. In one of these, an African species, 

 I found that the odour produced was like that of vanilla 

 chocolate. Another species of thi 3, a native of 



India, was named by Wood-Mason and • uaveo- 



lens, from its pleasant fragrance. The vanilla odour was 

 found by the same two observers in several other Indian 

 butterflies belonging to different groups. 



Some Satyrines have plume-scales which are not unlike 

 those of the Pierines, but differ in seldom, or perhaps never, 

 . possessing an accessory disc. At the utmost they may show 

 a slight dilatation of the articulating end of the footstalk. 

 I Plume-scales much like those of the common browns are 

 also found in the Fritillaries, which belong to the great 

 group of Nymphaline butterflies. 



There is yet another kind of scent-scale, specialised in 

 form. This is the well-known "battledore" scale, present 

 in the male of many of the small blue butterflies belonging 

 to the subfamily of Lycaenids. These battledore scales are 

 provided with apertures on their general surface which no 

 doubt serve, like the apertures of the fimbriae in the plume- 

 scales, for the escape of the odour into the outer air. The 

 ribs apparent in the " battledore " are in all probability 

 homologous with the longitudinal channels seen in the 

 Pierine plume-scale. These in Mylothris, as we have seen 

 on the screen, take the form of ribs as definitely marked as 

 those of the Lycaenids. 



So far, all the scents with which we have been con- 

 cerned are of a kind that is agreeable to our own senses. 

 But there is another sort of odour which is also commonly 

 present, especially in the butterflies of tropical and sub- 

 tropical regions, and which, instead of being pleasant to 

 the human sense, is disagreeable or even repulsive. The 

 Acraeas, which are mostly reddish or brownish butterflies 

 with small dark spots ; the Euploeas, large butterflies which 

 often show a brilliant purple gloss like that of our own 

 purple emperor ; the Papilios, of which a good example is 

 the black and yellow swallowtail butterfly of the Cam- 

 bridgeshire fens, have many of them an odour w-hich may 

 be called disgusting. Musty straw, stable litter, rabbit- 

 hutches, acetylene, bilge-water, these are some of the sub- 

 stances to which the odours of these unsavoury butterflies 

 have been compared. In some cases, as in the instance of 

 the agreeable perfumes, the seat of these evil-smelling 

 odours has been found in patches or tufts of specialised 

 scales or hairs ; in others the scent appears to be emitted 

 from the general wing-surface. But in no instance, so far 

 as I am aware, has any structure like a plume-scale been 

 found guilty of emitting anything but a pleasing fragrance. 

 A very remarkable difference between the scents pleasant 

 and the scents unpleasant is this : that the former kind 

 usually, though not invariably, is confined to the male sex ; 

 while the latter kind is common to both sexes, being often, 

 indeed, stronger in the female. 



It has, no doubt, occurred to you to ask : has the presence 

 of these scents any particular significance with regard to 

 the needs of their possessors; and if so, what? And why 

 should the agreeable scents be so commonly confined to the 

 one sex, while the repulsive odours are shared by both? 



The second question helps us to answer the first. With 

 regard to the scents of the disagreeable kind, which are 



