382 



NA TURE 



\_Fcb. 2 1, I i 



some respects, yet differed from it in some important points. 

 Tlie "low sun-bands " appeared weal: rather than strong, partly 

 perhaps by contrast with the great intensity of the rainband, and 

 the rainband itself was easily divided into lines, of which eight are 

 recorded in my notebook as being seen with a one-prism 

 spectroscope. The band between /' and F, observed by Mr. 

 Lnckycr, Mas also seen here, and was found to be one ascribed 

 to aqueous vapour, W.L. 504. A spectrum almost in all respects 

 similar to that observed here can be seen by any one who will 

 examine the absorption jjroduced by a small cloud passing over 

 the sun as seen with the spectroscope, having a lens in front of 

 the slit. The contrast with the bright spectrum of the sun sliows 

 the general absorption in the red very clearly, and if the sun be 

 near the horizon the other bands will be, in most cases, fairly well 

 seen. 



It is worth noting that we have had an unusually early and 

 heavy monsoon, ushered in by a remarkable thunderstorm and 

 folbwed by a period wlien the spectrum sho\\-ed an abnormal 

 freedom from vapour, the rainband at times being quite invisible. 

 During this latter period we have had beautiful rosy after-glows, 

 the .sunlight beingapparently reflected from thin, almost invisible, 

 cirrus clouds. 



If the presence of dust can be proved, these phenomena, 

 as I previously indicated, can be readily explained in accord- 

 ance with the facts so beautifully illu-trated by Mr. John 

 Aitken (Trans. R.S.E., vol. xxx. p. 337), for the dust particles 

 would condense moisture in the upper parts of the air, and we 

 would have a light haze, such as was observed here, not suffi- 

 ciently dense to cause actual clouds, but deep enough to give the 

 special absorption effects, while the dust it- elf would assist in 

 producing the general .absorption. 



Against the idea of Java dust, however, have to be set a 

 number of facts of which the f:)llowing are a few : — The maxi- 

 mum phase of greenness was on the same day (September 10), 

 all over Ceylon and South India, and as far west as long. 64° 

 (at sea). The green sun was not seen at Rangoon nor at the 

 Andaman Islands, though at the latter place the sounds of 

 the eruption were heard. The first rain that fell here after- 

 wards was subjected to careful microscopic analysis, and showed 

 no trace of volcanic du^t. The phenomenon reappeared on 

 .September 22. 



For my own part I think there is strong evidence that the 

 influence of the Jav.an eruption was an electrical one, and that 

 that was not necessarily propagated by the actual transference 

 of matter. Mr. Whymper's very interesting letter is of coarse 

 by no means conclusive as regards the effects of dus', for it is, I 

 believe, regarded as virtually proved that the mere existence of 

 dust in large quantities in volcanic ejecta proves the presence of 

 an abundance of water vapour. C. MiCHiE .Smith 



P.S. — There is a misprint in my letter to Sir William Thom- 

 son which, as I have seen it twice quoted, ought to be corrected. 

 It is in vol. xxix. p. 55, line 8, which should read : "After the 

 electricity had gone to negative." CM. S. 



The Christian College, Madras, January 23 



Since the end of October, when I first observed an unusual 

 red glow for a considerable time after sunset, I have been a 

 close observer of the atmospheric phenomena so fully described 

 by your correspondents. For some time past they have ap- 

 peared with little of their former brilliancy, until the evening of 

 the 7th inst., when there was a remarkably fine display, equal- 

 ling in many respects those of December. Of this 1 shall par- 

 ticularly mention but one feature which I had seen three times 

 previously, but never displayed with such intensity and clearness 

 of definition. At 5.30, when the after-glow was at its maxi- 

 nmm, a lovely crimson arc appeared opposite it in the eastern 

 horizon, in every respect .as described by Mr. Divers in his 

 letter dated from Japan, which appeared in Nature of January 

 24 (j). 2S3). I may remark that I have observed here, from 

 November 10 to this date, but latterly with much diminished 

 intensity, every one of the phenomena he so giaiahically de- 

 scribe.s. A. C. 



Roscommon, February 1 1 



"The Indians of Guiana" 



In the notice of Mr. Ini Thurn's work on the Indians of 

 Guiana, in the current volume of N.-\TURE (p. 305), Mr. Tylor 

 writes: "What is still more curious is that the rude method Of 



making thread by rolling palm or grass fibre into a twist with 

 the palm of the hand on the thigh may be commonly seen in 

 Guiana, although the use of the spindle for spinning cotton is 

 also usual." As such a fact appears to be curious to so eminent 

 an anthropologist as Mr. Tylor, it may be of interest to some of 

 your readers to learn that this mode of twisting fibres is still by 

 no means uncommon in India, though spinning must there have 

 been familiar to the natives for unnumbered generations. I have 

 frequently seen Hindus of various castes twist a mass of jute- 

 fibre into a compact and firm rope of considerable length, 

 between the palm of the hand and the inside of the thigh, and 

 by the same means they will frequently produce long pieces of 

 strongly coherent twine when the need for it arises. From my 

 experience, which, though confined to a small geographical area, 

 comprehended an acquaintance with both Hindus and Moham- 

 medans imported into the tea-districts from almost every part of 

 British India, I should suppose that this custom of twisting fibres 

 into rope and twine is universal throughout the country, though 

 doubtless it is resorted to rather as a makeshift than as a regular 

 mode of manufacturing twisted cords. That such a means 

 should be resorted to by the wild tribes of the north-eastern 

 frontier is by no means strange, though these have acquired not 

 a little skill in spinning and weaving cotton, but that so primitive 

 a method should still prevail amongst peoples so highly cultured 

 as the Hindus and Mohammedans of India often struck me as 

 remarkable. 



While noticing Mr. Tylor's interesting article, I cannot refrain 

 from questioning the justice of the supposition that pile-dwellings 

 on the land are due to the "survival of the once purposeful 

 habit of building them in the water." That in New Guinea 

 such is the case there can be little doubt, as Dumont d' Urville 

 and Mr. Wallace, as well as Prof. Moseley, have remarked. 

 And that Mr. Im Thurn's supposition with regard to the natives 

 of Guiana is also correct there can hardly be a doubt. But 

 these two cases scarcely seem to me sufficient upon which to 

 generalise, even when added to Prof. Moseley's pretty and 

 ingenious view as to the origin of the Swiss chalet. As has 

 been pointed out to me by my friend Mr. W. E. Jones, 

 F. R.I.B.A., Lecturer on Architecture in the Bristol University, 

 a somewhat similar development of single-storied into two-storied 

 dwellings is to be traced in the stone buildings as well as in the 

 less substantial dwellings of Western Asia, between the twentieth 

 and the twelfth centuries B.C., and though of course it is not 

 impossible, it certainly seems improbable that a race of ancient 

 lake-dwellers should have perpetuated on sandy plains a practice 

 which must altogether have ceased to be useful long before it 

 reached a region so far removed from its original home. And in- 

 deed it seems to me that pile-dwellings may be observed in 

 localities in which it is scarcely possible that the practice could 

 have originated in lake-dwellings, or in any dwellings of any 

 sort erected in water, whether fresh or salt. I allude more par- 

 ticularly to the r.aised dwellings of the Nagas, Kukis, Cacharis, 

 Khasias, and other hill-tribes of the north-eastern frontier ol 

 India, in the midst of which I lived for several years. That 

 these people should ever have dwelt so near the sea that they 

 acquired the habit of erecting pile-dwellings therein seems to 

 me highly improbable when it is remembered that their racial 

 and linguistic affinities place them undoubtedly in that great 

 Mongolian group of which the Thibetans and Burmese are ex- 

 amples ; and that therefore they may be regarded as immigrants 

 from more Eastern Asia, rather than as tribes which have been 

 gradually driven back from the Bay of Bengal by the encroach- 

 ing civilisation of the Hindus. Nor does it .seem probable that 

 their pile-dwellings were originally erected in lakes amongst the 

 hills, for in fact the lakes nowhere exist. There are indeed ex- 

 tensive li/:eeh or marshes, which during the rainy season some- 

 times contain a good deal of water. But these bheels are, during 

 at least a portion if not the whole of the year, so pregnant with 

 fever and ague that I cannot believe that they were ever em- 

 ployed, as were the lakes of Switzerland and Italy, for the 

 protection of the habitations of man. Yet these north-eastern 

 frontier tribes for the most part build their houses upon piles. 

 These are generally of bamboo, and so of course are very perish- 

 able, but occasionally small timber is employed. The floor or 

 platform (of coarse bamboo matting) is seldom raised more than 

 from twenty-four to thirty inches above the ground, though, if my 

 memory serves me, I have occasionally seen it raised as much as 

 between six and seven feet. Beneath this platform a good deal 

 of lumber generally accumulates, and the poultry and pigs fre- 

 quently congregate for shelter, but I think I never saw an 



