This insect is extremely variable, but I do not think it likely that any of the numerous forms will prove sufficiently constant to be regarded as distinct species. The male varies in the size and number of the black spots, many of which are often absent; in the extent of the purple sheen which is sometimes absent from the hind-wings, sometimes partially absent from the fore-wings, and sometimes extends over the whole of both pairs of wings; also in the colour of such sheen, which often inclines towards blue. Some specimens are much paler than others, and so far as my experience goes, these are chiefly found at considerable elevations; in such specimens, the ground colouring inclines towards yellow or orange, and the purple sheen is very brilliant, and extends over the whole of the wings. The female of this form is proportionately paler. Other specimens have the hind-wings almost black with no purple sheen, whilst in others the purple sheen remains. Another form has the usual markings, but the hind-wings are deep orange-brown, without purple sheen, which is also absent from the outer portions of the fore-wings. One female in my collection is dull brown, with yellow markings between the two rows of black spots. The under side is still more variable. One very striking form has only the basal portions of the fore-wings yellow, the rest of the ground colour is pale bluish-grey, and the spots black. On the hind-wings there are a number of black spots near the base; then an irregular band of black, and then a double row of marginal spots. An almost unlimited number of varieties appears to connect this form with one, in which all the markings on the hind-wings are nearly obsolete. The specimens of this insect taken in each district appear to exhibit differences from those taken elsewhere, but specimens also differ from the same district, so that at present we are unable to detect any well-marked local variation, or topomorphism, as it has been termed. It is consequently highly desirable that collectors should endeavour to obtain specimens from as many localities as possible, so that the nature of the variation of this butterfly may be better understood. Mr. Fereday states[57] that after carefully examining a patch of _Donatia {119}novæzealandiæ_, a plant he had noticed much frequented by this butterfly, he succeeded in finding a larva which there could be little doubt would have given rise to this insect, had it lived. The following is taken from his description: The caterpillar is shaped like a wood-louse, hairy, and pale green. There is a series of conical purplish spots down the back, edged first with white, and then with dull red. On the sides there is a series of pale pinkish oblique stripes, blended with dull red towards the spiracles.