For the first time, as stated before, I attempted the rearing of a small number of Atlas larvæ in the open air on the ailantus tree, but had to remove the last two remaining larvæ in September; the others had all disappeared in consequence of the heavy and incessant rains. These larvæ were from eggs sent to me by one of my German correspondents. The pairing of the moths had taken place on the 17th of July, and the eggs had commenced to hatch on the 4th of August.

I had about eighty cocoons of another and larger race of Atlas imported from the Province of Kumaon, but only eight moths emerged at intervals from the 31st of July to the 30th of September. Not only did the moths emerge too late in the season, but there never was a chance of obtaining a pairing. In my report on Indian silkworms, published in the November number of the "Bulletin de la Societe d'Acclimatation," for the year 1881, compiled from the work of Mr. J. Geoghegan, I reproduce the first appendix of Captain Thomas Hutton to Mr. Geoghegan's work, in which are given the names of all the Indian silkworms known by him up to the year 1871.

Of Attacus atlas, Captain Hutton says: "It is common at 5,500 feet at Mussoorie, and in the Dehra Doon; it is also found in some of the deep warm glens of the outer hills. It is also common at Almorah, where the larva feeds almost exclusively upon the 'Kilmorah' bush or Berberis asiatica; while at Mussoorie it will not touch that plant, but feeds exclusively upon the large milky leaves of Falconeria insignis. The worm is, perhaps, more easily reared than any other of the wild bombycidæ."

I will now quote from letters received from one of my correspondents in Ceylon, a gentleman of great experience and knowledge in sericulture.

In a letter dated 24th August, 1881, my correspondent says: "The Atlas moth seems to be a near relation of the Cynthia, and would probably feed on the Ailantus. Here it feeds on the cinnamon and a great number of other trees of widely different species; but the tree on which I have kept it most successfully in a domestic state is the Milnea roxburghiana, a handsome tree, with dark-green ternate leaves, which keep fresh long after being detached from the tree. I do not think the cocoon can ever be reeled, as the thread usually breaks when it comes to the open end. I have tried to reel a great many Atlas cocoons, but always found the process too tedious and troublesome for practical use.

"The Mylitta (Tusser) is a more hardy species than the Atlas, and I have had no difficulty in domesticating it. Here it feeds on the cashew-nut tree, on the so-called almond of this country (Terminalia catappa), which is a large tree entirely different from the European almond, and on many other trees. Most of the trees whose leaves turn red when about to fall seem to suit it, but it is not confined to these. In the case of the Atlas moth, I discovered one thing which may be well worth knowing, and that was, that with cocoons brought to the seaside after the larvæ had been reared in the Central Provinces, in a temperature ten or twelve degrees colder, the moths emerged in from ten to twenty days after the formation of the cocoon. The duration of the pupa stage in this, and probably in other species, therefore, depends upon the temperature in which the larvæ have lived, as well as the degree of heat in which the cocoons are kept; and in transporting cocoons from India to Europe, I think it will be found that the moths are less liable to be prematurely forced out by the heat of the Red Sea when the larvæ have been reared in a warm climate than when they have been reared in a cold one.

"I do not agree with the opinion expressed in one of your reports, that the short duration of the larva stage, caused by a high temperature, has the effect of diminishing the size of the cocoons, because the Atlas and Tusser cocoons produced at the sea-level here are quite as large as those found in the Central Provinces at elevations of three thousand feet or more. According to the treatise on the "Silk Manufacture," in "Lardner's Cyclopedia," the Chinese are of opinion that one drachm of mulberry silkworms' eggs will produce 25 ounces of silk if the caterpillars attain maturity within twenty-five days; 20 ounces if the commencement of the cocoons be delayed until the twenty-eighth day; and only 10 ounces if it be delayed until between the thirtieth and fortieth day. If this is correct, a short-lived larva stage must, instead of causing small cocoons, produce just the contrary effect."

In another letter, dated November 25, 1881, my correspondent says: "I am sorry that you have not had better success in the rearing of your larvæ, but you should not despair. It is possible that the choice of an improper food-plant may have as much to do with failures as the coldness and dampness of the English climate. I lost many thousands of Atlas caterpillars before I found out the proper tree to keep them on in a domesticated state; and when I did attain partial success, I could not keep them for more than one generation, till I found the Milnea roxburghiana to be their proper food plant. I do not know the proper food-plant of the Mylitta (Tusser), but I have succeeded very well with it, as it is a more hardy species than the Atlas. Though a Bombyx be polyphagous in a state of nature, yet I think most species have a tree proper to themselves, on which they are more at home than on any other plant. I should like, if you could find out from some your correspondents in India, on what species of tree Mylitta cocoons are found in the largest numbers, and what is about the greatest number found on a single tree. The Mylitta is common enough here, but there does not seem to be any kind of tree here on which the cocoons are to be found in greater numbers than twos and threes; and there must be some tree in India on which the cocoons are to be found in much greater plenty, because they could not otherwise be collected in sufficient quantity for manufacturing purposes. The Atlas is here found on twenty or more different kinds of trees, but a hundred or a hundred and fifty cocoons or larvæ may be found on a single tree of Milnea roxburghiana, while they are to be found only singly, or in twos and threes, on any other tree that I know of. The Atlas and Mylitta seem to be respectively the Indian relations of the Cynthia and Pernyi. It is, therefore, probable that the Ailantus would be the most suitable European tree for the Atlas, and the oak for the Mylitta."

Attacus mylitta (Antheræa paphia).--I did not receive a single cocoon of this species for the season 1881. My stock consisted of seven cocoons, from the lot received from Calcutta at the end of February, 1880. Five were female, and two male cocoons; one of the latter died, thus reducing the number to six. The moths emerged as follows: One female on the 21st of June, one female on the 26th, one female on the 28th, one female on the 1st of July, and one male on the 3d of August; the latter emerging thirty-four days too late to be of any use for rearing purposes. The last female moth emerged, I think, about the end of September. These cocoons had hibernated twice, as has been the case with other Indian species. I had Indian cocoons which hibernated even three times.