The caterpillar is brown above with a tinge of red or purple, and freckled with darker; there are three pale lines along the back, but only the central one is distinct, and this is more or less interrupted by clusters of darker freckles; there is a darker stripe composed of freckles on the sides, and below this is a pale brownish stripe; head, brown, plate on the first ring of the body blackish with pale lines upon it. It feeds when young in sallow catkins, and later on low-growing plants, also leaves of sallow and seeds of wych-elm. Early stages are figured on Plate 6. The moth is out in September and early October. It is widely distributed, and generally common, throughout England and Wales, Scotland to Moray, and Ireland. Its range abroad extends to Amurland and Japan. NOTE.--It may be stated here that the present species, together with _aurago_, _lutea_, _fulvago_, _gilvago_, and _ocellaris_, are referred to _Cosmia_, Ochs. and Treit., by Hampson (_Cat. Lep. Phal._ vi. 497). [Illustration] 2 Pl. 10. 1. ORANGE SALLOW. 6-8. THE SALLOW. 2, 3. BARRED SALLOW. 9, 10. DUSKY-LEMON SALLOW. 4, 5. PINK-BARRED SALLOW. 11. PALE-LEMON SALLOW. 12. ORANGE UPPER-WING. [Illustration] 2 Pl. 11. 1, 2. RED-HEADED CHESTNUT MOTH. 7-10. DARK CHESTNUT. 3-6. CHESTNUT MOTH. 11, 12. DOTTED CHESTNUT. {21} THE DUSKY-LEMON SALLOW (_Mellinia_ (_Xanthia_) _gilvago_). Two examples of this species are shown on Plate 10, Figs. 9 and 10. The purplish-brown mottling or clouding and greyish suffusion of the fore wings is much denser in some specimens than in others. Often the suffusion is quite absent, and the purplish brown is only seen as spots. Again, in an almost unicolorous form the ground colour is of a pale orange tint, the cross markings and outlines of the reniform are as in the type, and the series of blackish points on the submarginal line, usually present in the type, are more conspicuous, owing to absence of the other usual dark markings; this seems to be the _palleago_ of Hubner, which has been considered a distinct species; I think, however, that it is only a form of _gilvago_. The earliest recorded British specimen of this form was taken at Brighton in 1856, and it and others captured in the same district were then thought to be examples of _M. ocellaris_, but their true identity was established by Doubleday in 1859. Very few specimens of this form have been reported from other parts of England, but I have recently seen one that was taken at light in the Canterbury district, Kent, on October 3, 1907. In its typical form this species has an extensive range in England, spreading from Yorkshire to Surrey and Sussex. The earliest known British specimens were captured in the neighbourhood of Doncaster over sixty years ago, but its occurrence in Surrey seems not to have been noted until comparatively recent times.