Of ordinary spots it is merely necessary to remark, that they are all cases in our favour. Thus, in Satyrus hyperanthus we have "the ordinary round spots ... changed into lanceolate markings"; this takes place also in C. davus. The other cases of aberration do not concern us.
When, however, we come to the cases in which a species has two or more permanent forms, it is necessary to show that they are in all cases founded on structure lines. The patterns, as shown in [Plate V., Figs. 1-13], are always arranged structurally, and the fact that not only are intermediate forms known, as in Araschnia porima, [Plate V., Fig. 6], but that the various forms are convertible into one another, would in itself be sufficient to show that in these cases there is no departure from the general law. In Grapta interrogationis, [Plate V., Figs. 8-10], we see in the central figure one large spot above the median nervure, in the left-hand form this is surmounted by another spot above the lowest sub-costal branch, and in the right-hand figure this latter spot is very indistinct. We have here a perfect gradation, and the same may be said of the colouration of the lower wings. Take again the three forms of Papilio Ajax in the same plate, [Figs. 11-13], and we have again only modifications of the same type.
In local varieties, as in seasonal forms, we have again nothing more than developments of a given type, as is well shown in [Plates IV.] & [V.], Figs. [13-18] & [1-13].
When, however, we come to mimetic forms, whether they mimic plants, as in [Plate I.], or other species, as in [Plates II.] & [III.], a difficulty does seem to arise.
The leaf butterfly (Kallima inachus), [Plate I.], offers no trouble when we view the upper surface only with its orange bands, but its under surface, so marvellously like a dead leaf that even holes and microscopic fungi are suggested, does seem very like a case in which structure lines are ignored. Take, for instance, the mark which corresponds to the mid-ribs, running from the tail to the apex of the upper wing; it does not correspond to any structure line of the insect. But if we take allied and even very different species and genera of Indian and Malayan butterflies, we shall find every possible intermediate form between this perfect mimicry and a total lack of such characters. To cite the most recent authority, the various species of the Genera Discophora, Amathusia, Zeuxidia, Thaumantis, Precis, &c., figured so accurately in Distant's Rhopalocera Malayana, will give all the steps.
In the cases of true mimicry, as in [Figs. 1-3, Plates II.] & [III.], where insects as different as sheep from cats copy one another, we find that of course structure lines are followed, though the pattern is vastly changed. The Papilio merope, [Fig. 1, Plate II.], which mimics Danais niavius, [Fig. 3], does so by suppressing the tail appendage, changing the creamy yellow to white—a very easy change, constantly seen in our own Pieridæ—and diffusing the black. A similar case is seen in [Figs. 4-5, Plate III.], where a normally white butterfly (Panopœa hirta) mimics a normally dark one of quite a different section. Here again the change is not beyond our power of explanation. Where a Papilio like merope mimics a brown species like Danais niavius, we have a still greater change in colour, but not in structural pattern.
If we ascribe to these insects the small dose of intelligence we believe them to possess, we can readily see how the sense of need has developed such forms.
Local varieties present no difficulty under such explanation. The paramount necessity for protection has given the Hebridran species the grey colour of the rocks, and the desert species their sandy hue.
Plate VII.