Glossina fusca and G. palpalis are prevalent throughout the Western Province of Ashanti, the latter being much the most common species. They both occur in dense jungle, but the type of bush most favourable for them appears to be the scrub and smaller forest which occurs in the dividing line of country between the dense jungle and typical “orchard” grass bush. Scrub of a height of three or four feet upwards appears to be sufficient to harbour these insects.

These notes are intended as a summary of the early investigations into the habits of Glossina morsitans carried out in connection with the Luangwa Sleeping Sickness Commission. The part of the valley that has been examined is that lying between 12° S. and 13° 50' S., but it has been practicable to make only one journey to the eastern side of the river. This journey was undertaken in September and consisted of an examination of the Rukususi and Rukusi Rivers from the Luangwa almost to the Nyasaland boundaries. Hence these remarks apply mainly to the district immediately surrounding Nawalia, the headquarters of the Commission. The observations have been made during the period from August 1911 to the beginning of March 1912.

The compilation of the following key has been a matter of no little difficulty, mainly owing to the close connection of the species in some of the groups, which sometimes makes it almost impossible to assign specific limits. The difficulty has in some cases been increased through the paucity of material, which prevents any adequate conception of the range of variability being obtained. This is particularly the case with some of the species coming from the Mediterranean region, which are very closely allied, and of which, as a rule, the British Museum possesses very few specimens. Names have only been sunk here as synonyms in those cases where there appeared to be no reasonable doubt, either after a comparison of the types, or of the descriptions, when these were sufficiently detailed. Eventually, therefore, it may be found that some forms which are here given specific rank will have to be regarded at most as varieties. Since so many figures of Anopheline wings, etc., have already appeared, it is not deemed necessary to add to their number. Some new records have been included, but on the other hand some old ones, which appeared to be questionable, have been omitted. As with the writer's previous papers, this key is merely intended to supplement the detailed descriptions which will be found in other works.

The following article has been written from notes of observations and experiments extending over nearly three years—April 1909 to December 1911. Its main object is an attempt to demonstrate that a true conception of the production of Auopheline mosquitos in a definite country or district can only be formed by a large amount of continual work throughout the seasons of the year. At the present time very numerous observations are still to be made before much generalising should be allowed concerning the breeding-grounds and bionomics of these insects.

The silkworms which are found in the largest quantities in Uganda belong to the genus Anaphe (family EUPTEROTIDAE), of which we have three species, Anaphe infracta, Wlsm., A. venata, Butl., and A. moloneyi, Druce. The other species of silkworms which have so far been observed in the Protectorate are Mimopacha gerstaeckeri, Dew., and Hypsoides milleti, de Juan, but these are comparatively rare. This paper deals only with Anaphe infracta.

Owing to the very limited time available during my stay in England, it has not seemed advisable to attempt any general account of so extensive a region as that covered by my recent journey through Tropical East Africa. This region extends practically from the Zambesi River to the Upper Nile, and is included in 12 degrees of longitude and no less than 22 parallels of latitude. It comprises all types of country from open plains to dense forests, and elevations from sealevel to 10,000 feet. This area includes, in addition to a large portion of the Eastern Tropical subregion, a small portion of the South African subregion and. in Uganda, of the Western Tropical subregion. The entomological study of so large an area involves a great variety of interesting questions and an even more bewildering diversity of insect life. The economic relations of these numerous species with man and his domestic animals present a large number of problems, in regard to many of which much additional evidence is required.

Mr. A. D. Michael has already recorded* the presence of mites of the genus Tarsonemus upon diseased sugar-cane from Barbados. He states that two species belonging to this genus were present in the material sent to him, and that the larger of the two species was certainly identical with one which Mr. Bancroft found doing serious damage to sugar-cane in Queensland. Mr. Michael proposed the name Tarsonemus bancrofti for this larger species. So far as I am aware, no description of this nominal species has been published and I am obliged to rely on Mr. Bancroft's published sketches† for information concerning it. Unfortunately his drawings are not executed in sufficient detail and I am not certain that his mite is the same species as the one which is dealt with in the present note; but as the figures of the Queensland mite differ appreciably from the Barbados specimens, it seems advisable to describe the latter under another name (T. spinipes). The species of the genus Tarsonemus often resemble one another very closely in structure, and they cannot be recognised with certainty unless a fully detailed account of their principal characters, accompanied by careful drawings, is given. Dr. Bancroft gives drawings of both sexes of his mite. He does not figure any spines on the third leg of the male, but he shows a lobeshaped expansion, similar to that of T. spinipes, on the inner side of the short fourth leg. The hairs of the body are not depicted. According to his drawings, the body of the female resembles that of T. spinipes in being very long and narrow, but is apparently much narrower at the anterior end. He represents the two terminal setae of the fourth leg of the female as being both very long and slender, the outer one being seemingly almost as long as the inner. The size of T. bancrofti is not stated, nor is the scale of enlargement of the figures given.

Q.—Length (12 specimens) 9·5 to 12 mm.; width of head 3·5 to 4·4 mm.; width of front at vertex just under 0·5 to just under 0·6 mm.; length of wing 7·75 to 9·8 mm.