April 7, 1892] 



NATURE 



541 



N' 



ON INSECT COLOURS} 

 II. 

 OW it is necessary to explain the "reversion effects" 

 of red, so frequently alluded to. I am tempted 

 to give a detailed account of the experiments made in 

 this connection, but the length to which this article 

 has already run warns me that I must be very brief 

 indeed : and I will therefore content myself with giving 

 simply the broad results.- All reds and pinks (always 

 omitting the last four in the table), are turned orange or 

 yellow instantly by acids. When nitric acid is used, 

 this tHect'is per?nanc}tt ; and whether the yellowed wing 

 be dried, or washed, the yellow is immovable. I have 

 kept such wings for five or six months, and they were as 

 yellow as possible at the end of that time. In the case 

 of all other acids,* X.\i& yellow is permanent only so long 

 as the wing is actually acid: directly the acid is 

 removed, the original red returns ; and thus a 

 wing may be alternately yellowed and "reverted" 

 time after time. This reversion to the original 

 red may be produced either by long exposure to the air, 

 allowing the last traces of acid to drain off; or instantly 

 by neutralizing the residual acid with a drop of ammonia, 

 ar by copious quashing. It must therefore be under- 

 stood that, with the exception of those cases in which 

 nitric acid has been used, the permanency of the artificial 

 yellow is entirely dcpendetit upon the presence of acid : 

 remove the acid, and the yellow vanishes. Accordingly, 

 I have suggested the following explanation. Let us de- 

 note the molecule of red pigment by X ; when any acid, 

 except nitric, is added, I assume that this forms with X 

 a so-called molecular compound: for instance, on treating 

 with hydrochloric acid, we should get the hydrochloride of 

 X, viz. X.(H.C1)„ ; and it is evident that these hydro- 

 chlorides, hydrosulphates, &c., of X are yellow, although 

 the original X is red. To all these facts, of course, there 

 are ample analogies known to chemistry. Next, for the re- 

 suscitation of red. We must suppose — what is certainly 

 to all appearance very clear— that these molecular com- 

 pounds are very unstable ^ an easily understandable fact ; 

 and that consequently the addition of even excess of water 

 is sufficient to decompose them, removing the acid mole- 

 cule, and thus restoring the pigment X to its original 

 condition. Far more rapidly does this resuscitation occur 

 if a drop of ammonia be used, this at once combining 

 with the acid and liberating the X molecule. In the case 

 of resuscitation produced by slow air drying, the action 

 apparently would be in some cases due to gradual 

 evaporation, or to some process of oxidation — anyhow 

 producing dissociation of the molecular salt of X. Finally, 

 in the case of nitric acid, it is clear that this acid does 

 not form a molecular compound, but, as we might expect, 

 exercises a permanently destructive action on the ori- 

 ginal pigment. Admitting that red has been developed 

 from yellow, it is not surprising that it may be easily re- 

 converted permanently into yellow by such a reagent as 

 nitric acid. Before quitting this topic, I may point out 

 that the cyanide reaction of the yellows is very sug- 

 gestive indeed as to the kind of process by which the red 

 pigment is developed from yellow. 



Now, as to the last four species noted in the table. In 

 these, I believe, the red is not developed from yellow at all, 

 but from its close analogue, chestnut. Up till very recently, 

 I supposed V. atalanta to be the only representative of 

 such development, and was rather surprised that yellow 

 should so commonly develop into red, and chestnut so 

 rarely. But recently I have found that Anartia avtalthea 

 is exactly identical in behaviour with V. atalanta, whilst 

 Heliconius amaryllis seems half-way between these spe- 



' Continued from p. 517. 



"^ A full account of these experiments will be fonnd in the Entomologisi, 

 Kxm. pp. 39-40 and 53-59. 



' I have used hydrochloric, sulphuric, acetic, pho.sphoric, hydrofluosilicic, 

 and oxalic acids, in these experiments. 



NO. I 171, VOL. 45] 



cies and the normal reds, but nearer to the former. The 



evidence on which this conclusion as to the nature of 

 the red in V. atalanta was founded is as follows. The 

 red of atalanta does not change to yellow, but to the 

 brown or chestnut normally present in V. cardui, or to a 

 more colourless tint. The change is not similar to that 

 of red to yellow, but is a solution effect : consequently no 

 reversion effect can be obtained ; and this alone is almost 

 decisive. It seems to me especially interesting that this 

 experimental conclusion as to the nature of atalanta red 

 is entirely corroborated by totally independent evidence 

 from the entomological side, since the connection of V. 

 atalanta and cardui is exceedingly close, and there are 

 transition forms between them.^ 



And now we come to the last colour — chestnut — for 

 which a very brief account will suffice, in addition to the 

 details already given in the table, and the incidental 

 remarks made during the discussion of yellow. It must 

 be understood, then, that the constitution of chestnut 

 appears to be very close indeed to that of yellow : like as 

 in yellow, we can distinguish several stages of solubility, 

 although deepening colour still less implies decreasing 

 solubility even than it does in yellow — a conclusion which 

 will be borne out by an examination of the table. Like 

 yellow, chestnut may develop into red, as has already 

 been explained ; and the brilliant copper colour of Lyccena 

 phla'as and virgaurea appears to occupy, both in its ex- 

 treme solubility and its relation to the main line of develop- 

 ment of the chestnut pigment, a position exactly analogous 

 to that occupied by the orange oiE. cardamines among the 

 yellows. The only further remark that I have to make 

 with reference to this colour concerns V. io and V. antiopa, 

 which I have specially marked as notable examples. In 

 these species the rich chocolate colour is very soluble, but 

 leaves a black wing instead of a white. If chestnut had 

 been developed from a white pigment, this would have 

 been a grave difficulty ; but it perfectly accords with the 

 view that the pigment has been developed, not from any 

 such white pigment, but in a previously unpigmented, 

 usually white, wing ; in these species it has been de- 

 veloped in a previously black wing. I have always 

 considered the behaviour of these two species to con- 

 siderably support my views of the nature of the chestnut 

 pigment, and indirectly of the yellow.-^ 



The main heads of the preceding pages may be very 

 briefly summarized as follows. Blacks and whites are not 

 pigment but absorption and reflection colours respectively. 

 The great majority of blues are also physical colours — 

 chiefly, if not entirely, interference colours ; and it is 

 doubtful if there be any pigment blues at all. Some 

 greens are also physical colours, very similar to the blues ; 

 the character of another group is somewhat ambiguous, 

 although probably these, too, are physical. A third group, 

 is pigmental, and probably derived from yellow. All 

 reds are pigmental, being developed chiefly from yellow, 

 but in a few cases from chestnut ; the former are char- 

 acterized by the reversion effect. The great majority of 

 yellows are pigmental, of various degrees of solubility 

 or insolubility ; but a few cannot at present be decisively 

 pronounced either physical or pigmental, and the same 

 remark applies to the chestnuts.^ 



In concluding this summary of my work, I must point 

 out that it is not put forward as in any sense of the word 

 final, even so far as it goes, but merely as a basis of 

 systematic inquiry, in various directions. Up to the 

 present, almost nothing at all has been known about the 

 behaviour or character of these colours ; now I will dare 



' There is similar evidence in the case of Anartia ainalthea. Two 

 specimens of this were sent me for experiment. One was marked with a 

 chestnut band, and one with a scarlet. This scarlet was at once changed 

 into the chesnut notmally present in the other. 



^ I am disappointed at having as yet come across no yellow species 

 analogous to V. io. But in this connection I may call attention to the 

 behaviour of theg-reett species of Cidaria, which are changed to a brownish- 

 grey. It is possible that these greens may be descended from yellow 

 developed on an originally dark wing. 



3^Cp. the instances of Vanessa io and antiofia. 



