products are maintained for the use of scientific and technical visitors. The educational side of the Institute's work has also been reorganized. The galleries are at present closed for redecoration and rearrangement of exhibits on a plan designed to attract public attention and utilize fully their educational value. This scheme has been drawn up in co-operation with educational authorities, teaching bodies, and chambers of commerce. Special features are the attraction of attention by realistic dioramas and models, supplemented by photographs, diagrams, etc., and the arrangement of the exhibits systematically, starting with world position and geographical features, and coming later to the economic products and the life of the people. Through the support of the New Zealand Government, the New Zealand Gallery has been practically completed on these lines. The publication of a monthly Bulletin to assist teachers is under consideration. A further development is the holding of exhibitions of the work of the British School of Rome Scholars, and also of artists of the Empire. The Hydroplane of the Hamilton Rice Expedition, 1924-25. We have been informed by Colonel Sir Henry Lyons, F.R.S., Director of the Science Museum, South Kensington, that the Curtiss Sea-gull hydroplane, used on Dr. Hamilton Rice's last expedition to the Amazon valley, and presented by him to the Museum, is now on view in Gallery VIII. A full account of the work of the hydroplane was given in the Journal for July last (pp. 27-43). CORRESPONDENCE Houses in Eastern Tibet WHEN reading back numbers of the Fournal I noticed some remarks by a former President on Tibetan houses which do not in every case apply to the architecture in regions between 28° to 34° N. and Long. 98° to 104° E. The houses in the above districts are not all of the Egyptian type, although flat roofs are the rule. The structures in the Min and Wi valleys show indications of being an exception. The higher buildings, palaces, temples, and towers are here, as in western Tibet, of the Egyptian pattern mentioned, but in the ordinary dwellings the walls are built up with little if any difference in the thickness. In Chantui both types may be seen side by side. I suspect that the former may withstand the awful earthquakes better than the latter. Certainly, in Hor Drangu, during the earthquake in 1923, when almost every building collapsed, very high walls of mud, broad at the base and tapering towards the top, remained standing. The earthquakes in this region occur about every ten or twenty years and practically level all buildings. The epicentres seem to be moving north-west in the Kanzè direction. The first serious one within thirty years was at T'ai Ning, 60 English miles from Tatsienlu; then about twenty years ago a similarly serious one shook Dawo, 35 miles from T'ai Ning; and in 1923 another disastrous shock destroyed Hor Drangu and all the settlements to within 5 miles of Dawo. Hor Drangu is 40 miles from Dawo. In this last earthquake a French priest in a Mission station near the former settlement was killed in the ruins, with about 50 of his converts and adherents. The houses are built of clay or stone. The inside mass of roofs, etc., rests on wooden pillars with stone blocks and foundations, and are easily shaken into a confused heap-that is, they collapse inside the stone or earthwork walls. But the strength of the "shake" of 1923 may be indicated by the fact that small shrubs were in places thrown out of position. In such places the earth was broken up into large squares and resembled a gigantic draught-board. The use of clay and earth is very common along both the west and the north-west roads. The palaces of the princes in Litang, Batang, Hor Drangu, Chuwo, and Kanzè, if not Derge, are, if I remember correctly, built of clay or material other than stones. In some regions the clay is preferred, and it is not uncommon to see large squares or oblongs of this material in walls of stone. The clay seems to last as long as the stone. At Weichow the ruins of an imposing wall dates, according to tradition, as far back as 250 A.D.; but taking a hint from Du Halde, we might look upon it as the Eastern palace of the Tu Fan kings, and (approximately) refer it to 700 A.D. In this region also we find high slender towers-perhaps 100 feet-built of clay or a material resembling loess. A similar material abounds between Dawo and Kanzè.* I very much doubt if the flat-roofed stone or clay dwellings originated on the Tibetan steppes. A tent-like ridge and not a walled quadrangle would accord better with climatic conditions. The flat-roofed house seems to have been in use for thousands of years in the contiguous valleys; and to-day the wall and house builders are natives of Mao Chow and the low Min regions which at the dawn of Chinese history composed the "Kingdom of the Ti." Judging also from numerous chipped stone implements found in the Tibetan valleys and on passes and plateaus up to 14,000 feet, we should conclude that the house builders came from the valleys. Although these structures differ much in detail, a land with a scanty rainfall, without timber, but with abundance of clay and stone, may be suggested as the original control. Certain architectural peculiarities, as well as details in the dress of Tibetans and frontier tribes, would suggest a Han Dynasty influence. A comparison of Tibetan architecture, also, with imitation pillars, etc., on the Kia Ling caves might tend to confirm it. But while the Tibetans credit China and Nii Uyag + with giving them works of art and mathematics, laws and specimens of workmanship came from the north via "Hor and Yugera." The Hor still exist, vigorous and numerous, to-day around Kanzè and Drangu (see 'Jaeschke,' p. 511). The Hor may be the remains of the Yüch Chĩ, who in their historic flight from the Huns disappeared in a southerly direction, instead of overrunning the kingdom of Græco-Bactria (Ta Hsia or Ta Ha) like their more enterprising compatriots.‡ The limit of house dwelling by Tibetans may be about 12,500 feet, which is the limit also of any but the hardiest of cereals. On the Litang plain, apart from the lamaseries and castles of the former princes, there are a few score of well-built dwellings of the non-Egyptian type. But we must explain the owners as retainers and other agents of the princes who came from much lower altitudes. The nomads who swarm over the Litang plain are tent dwellers. In June last year these were in sufficient numbers to control 60,000 sheep, 40,000 yaks, and * Kanzè may be 180 English miles in a north-westerly direction from Tatsienlu. † The region half circling Tatsienlu is now known as Minyag. + Mentioned in Szema Chien's History of Ta Wau or Khôqand. 2000 horses. The number of savage dogs was not stated, and my vividly impressed imagination refrained from guessing. "Jhuming" J. HUSTON EDGAR. With reference to Mr. Mill's paper in the April number of the Royal Geographical Society's Fournal, I gather that he attributes the system of Jhuming" to the fact that the soil is too poor to produce yearly crops; but is this the case? This system of cultivation extends eastward from Assam across northern Burma to the borders of China and northward to the edge of Tibet. In the Kachin and Black Maru countries of the 'Nmai Kha (eastern branch of the river Irrawaddy) the land is cultivated once only in five to eight years, the explanation given to me being that after one crop has been reaped, the ground becomes so choked with weeds, chiefly sedges and grasses, that it is easier to prepare a new piece of ground than to clear the old piece. As the jungle grows the trees and bushes gradually kill out the weeds, and when this is accomplished the ground is again ready for "jhuming," or, as it is called on the Burma side, for "Toungyah" clearing. It would be interesting to have a sketch-map showing over what area of country this system of cultivation is in vogue; to the best of my recollection, it extends from the mountains forming the boundary of Burma and Assam as far as the valley of the Salween. I do not remember having met with it in the hills between Assam and Bhutan, nor in China or the Shan States. Then, again, regarding the ethnography of these people. According to Kachin legends, the original home of man was Tibet. When the country became overpopulated the Creator sent for the headmen of the various tribes and told them they must migrate down the valleys of the following rivers, viz. Brahmaputra, Mali Kha, 'Nmai Kha, Salween, Mekong, Red River, and Yangtze Kiang, and to each headman he gave a separate language written on strips of parchment. The Kachins maintain that their forefathers lost their way in the snows, and dying of hunger boiled their strips of parchment to make soup. Hence they lost the art of writing and have never been able to regain it. The main lines of migration down the above-mentioned rivers have been cut across by other migrations, such as that of the Shans to the Hukong Valley and the Chinpaw Kachins as far west as Sadiya in Assam, where they call themselves Singphos. I could make myself perfectly understood by the latter tribe, though they remarked from my accent that I must have learnt the language in Burma. ELDRED POTTINGER, Lieut.-Colonel (retired). r Martin Conway (18 (1910); the eastern and north-eastern areas Lt.-Col. H. Wood, R.reas from the maps of the Survey of India, usband. The work o |