The caterpillar figured by Harris is of the brown form, so we see that even at this early date something was known of the life history of this moth and the variation of the caterpillar. Since that date and up to 1894 only very few larvæ appear to have been found in our islands. Barrett states that it is doubtful if more than twenty had then been recorded. In 1895 caterpillars were obtained in Cornwall (four) and in Kent (two). Then for five years little or nothing was reported about this stage, although the moth seems to have occurred in varying numbers each year. In 1901, August and September, over one hundred were reported, rather more than half of which were taken from a hedgerow, overgrown with _C. sepium_, in {30} Northumberland; twenty-six were obtained on the bindweed growing on Lancashire sandhills, thirteen or fourteen in Essex, and others in Bedfordshire, Kent, Hants, Dorset, and Devon. Mr. Bell-Marley obtained thirty eggs, September, 1897, and although these were kept in a cold room, thirteen caterpillars hatched, September 21. They were supplied with _Convolvulus arvensis_ and _C. soldanella_, and seemed to relish one as much as the other. Seven died during the first three moults. The bindweeds being nearly over, seedlings were raised by forcing, but before these were ready the larvæ had been on short commons, and just immediately before the seedlings came to hand, had been twenty-four hours without food. On these tender seedlings and some endive the remaining larvæ, six in number, attained full growth in December. Two subsequently died in the first half of that month, and the others went under the soil. Only one, however, managed to assume the chrysalis state. A small caterpillar, about one week old, described by Paymaster-in-Chief G. F. Mathew ("Notes on Lepidoptera from the Mediterranean," _Entom._, xxxi. 115), was 1¾ inch long, pale glaucous green in colour, and thickly covered with raised white dots; oblique side stripes white, bordered above with dark green. On September 26, 1897, this caterpillar, which had been found on September 18, was nearly full grown, and the writer goes on to state that when gathering bindweed he obtained either eggs or tiny caterpillars at the same time, and he eventually found that he had eight of them altogether. They fed up rapidly, as a caterpillar, hatched about September 27, had gone down on October 18. Owing to accident, four produced deformed chrysalids, but each chrysalis resulting from the others was perfect and healthy on February 15, 1898. The large reddish-brown chrysalis is figured on Plate 11, and it will be noted that the "tongue" case forms a curious bent projection not unlike the handle of a pitcher. To give some idea of the irregular way in which this migratory species visits our islands, it will suffice to note the records only since 1894. Previous to that year it was common, more or less generally, in 1846, 1868, 1875, 1885, and 1887. [Illustration]