oo 



^o 



NAIURE 



[Jzuy 25, i8;8 



well-known to the natives under the title " Buddha's rays." It 

 has also, I believe, been noticed in Cashmere. The pheno- 

 menon, which is very striking indeed under favourable condi- 

 tions, is confined to the mountain region in the central parts of 

 the above-named island, and is never, as far as I am aware, 

 seen in the low country. It was therefore with considerable 

 interest that I learnt that it is well known to the French, and 

 had been seen by Mr. Lockyer at sea. In May, 1876, in a 

 paper on "Remarkable Atmospheric Phenomena in Ceylon," 

 read before the Physical Society and published in the Proceed- 

 ings, I offered an explanation of this phenomenon, in accord- 

 ance with the conditions under which it appeared. As this ex- 

 planation is very brief, may I ask your permission to reproduce 

 it verbatim 1 



"Not unfrequently in the mountain districts broad beams 

 apparently of bluish light may be seen extending from the 

 zenith downwards, converging and narrowing as they approach 

 the horizon. This ray-like appearance is very similar to that 

 seen before sunrise ; only the point from which the rays pro- 

 ceed is opposite the sun : the rays themselves are very broad and 

 blue in colour ; and the spaces between them have the ordinary 

 illumination of the rest of the sky. If we suppose in this in- 

 stance that the lower strata of air are colder than the upper (a 

 condition of the atmosphere which not unfrequently occurs in a 

 tropical mountain district like that of Ceylon, where large cur- 

 rents of heated air sweeping up a valley cross another valley 

 nearly at right angles and at a considerable elevation above it), 

 the refraction spoken of in the case of Adam's Peak will be 

 downwards instead of upwards. If, too, the observer be below, 

 the veil of darkness will appear to him like a very elongated 

 triangle, apex downwards, or broad ray, through which the blue 

 sky beyond may be seen free from the palish illumination of the 

 atmosphere, whilst on either side the ordinary illuminated sky 

 Avill be seen. If now we suppose several isolated masses of 

 cloud to partially obscure the sun, as was the case when I 

 witnessed the phenomenon, we may have several corresponding 

 inverted veils of darkness, like blue rays in the sky, all appa- 

 rently converging towards the same point below the horizon. 

 This apparent convergence of the beams is merely an effect of 

 distance, as in the case of parallel rays of light from the rising 

 or setting sun, the blue rays being practically parallel bands in 

 the atmosphere devoid of illumination. It will be evident that 

 conical-shaped clouds are not necessaiy to produce this effect. 

 Isolated clouds of any massive form would be sufficient to throw 

 the bands of shadow through the illuminated atmosphere, and 

 refraction and perspective would do the rest. The above pheno- 

 menon is called by the Singhalese * Buddha's rays ; ' and though, 

 according to Sir Emerson Tennent, it is very varied in character 

 and appears in different parts of the sky, yet I have only seen it 

 Avhen the sun was low at evening and when the rays converged 

 to a point, apparently directly opposite the sun ; and I do not 

 think it possible for the phenomenon to be seen in any other 

 position." 



It will be seen that I thought it necessary to suppose a pecu- 

 liar state of the atmosphere, in order that the shadows them- 

 selves might be sharply refracted downwards ; indeed, I think, 

 if the phenomenon can be produced under ordinary conditions, 

 it would be much less rare in the island than it is. It may be 

 mentioned that the sharpness and definition of the rays was most 

 decided — far more so than the sunbeams and shadows are at sun- 

 rise or sunset — and that the position from which I observed the 

 ' phenomenon was a ridge 4, 500 feet high, overlooking a deep valley 

 lying north and south, into which another valley running east 

 and west opened at a high elevation, so that the warm air coming 

 np the latter from the country below would overlie the cooler 

 air of the first valley. I am not certain whether mountains a 

 few miles to the west had any part in the production of the 

 phenomenon, hut the sun was certainly setting behind dense but 

 isolated masses of cloud. "With the above exception, it appears 

 that my explanation is identical with that of Mr. Norman 

 Lockyer and Prof. Briicke. 



The reference in the above quotation to Adam's Peak is to 

 the shadow which this isolated and cone-shaped mountain throws 

 up into the sky at sunrise. The summit of this peak is 7, 200 

 feet high, and overlooks the low country for fifty miles or more 

 to the west. The shadow thrown by it at the moment of sun- 

 rise appears to lie "horizontally over the land and sea for a 

 distance of seventy or eighty miles. As the sun rises, the 

 shadow rises and takes the form of a gigantic pyramid rising in 

 the air and approaching the mountain, until at last it appears to 

 fall over on the summit. The explanation which I offered of this 



was in principle identical with that of the mirage, the rays 

 grazing the summit and sides of the mountain being refracted 

 upwards, and thus leaving a portion of the atmosphere, similar 

 in shape to the upper part of the Peak, unilluminated. This re- 

 fraction is extremely likely to occur, as the difference in tem- 

 perature between the air in the low country and on the top of 

 the Peak is at least from 30° to 50° F. during the night and 

 morning. The sudden vanishing of the shadow is, no doubt, 

 due t3 the rays reaching the critical angle when the sun has 

 risen to a certain height, and total internal reflection ceasing to 

 take place. It may be mentioned that in both cases the shadows 

 appear to be blue, but this, I imagine, is only the effect of con- 

 trast. This theory will, I believe, also explain the phenomenon 

 of the spectre of the Brocken, which appears, I understand, 

 under similar but less pronounced conditions. 



The above appear to be very good instances of the illumina- 

 tion which the particles of air are capable of receiving, for it is 

 evident that neither phenomenon would be visible if the atmo- 

 sphere were perfectly pure. 



Another phenomenon I shall be glad to refer to, as bearing 

 on the subject of a perfectly clear sky at sunset. During the 

 dry season on the West Coast of Ceylon, when the air is so 

 free from moisture that even in houses close by the sea-shore the 

 backs of books and sheets of paper curl up as if they had been 

 placed before a fire, I have noticed, in looking westward over 

 the sea immediately after sunset, that the sky has presented the 

 appearance of an almost perfect spectrum extending from the 

 horizon upwards for a very considerable distance towards the 

 zenith (as far as I remember to within 20° of the zenith), a fiery 

 red being lowest, succeeded by orange, yellow, green, and blue. 

 With the exception of the red the colours were of the most trans- 

 parent character possible ; such as it seemed to me at the time it 

 would be quite impossible to reproduce except with the prism or 

 or diffraction lines. 



The explanation which I would offer of this seems to bea 

 directly oh the question of the colour of the sky, and as it has not 



I believe, been suggested, I shall be glad if you will allow it to 

 appear. It may, however, be an old explanation, but I am not 

 aware that it is. When a ray of sunlight passes obliquely 

 through a thick plate of glass the issuing ray is found to have 

 been dispersed during its passage, although on account of the 

 overlapping of the spectra produced by the pencil it appears to 

 the eye as wh'te light, one edge where there is no overlapping 

 being slightly red, and the other slightly blue. Now the atmo- 

 sphere of the earth is such a parallel plate as this, and the solar 

 rays falling on it very obliquely are dispersed into their com- 

 ponent colours, the air being the dispersive medium. These 

 colours cannot be observed (i) unless the obliquity is very 

 great, for the colours overlap and white light is the result ; (2) 

 unless there is very little moisture in the air to absorb the more 

 delicate colours and so prevent the dispersion effect being distin- 

 guished. In the case I refer to both these conditions were present 

 in a remarkable degree. A reference to the figure will show 

 at once the explanation offered for the spectral appearance of the 

 sky. Let A A' be the surface of the earth, B b' the upper limit of 

 the atmosphere, D d' the upper limit of the more refracting 

 layers of the atmosphere, a the point of observation, A x the 

 horizon. 



A solar ray striking the atmosphere at P will be refracted and 

 dispersed, the blue ray striking the earth at c, and the red at c'. 

 When these rays reach the surface of a sea that is nearly calm 

 they will be reflected in various directions according to the 

 angle at which they strike the wave surfaces. Some will be 

 reflected nearly vertically and will be lost in the clear sky, 

 others will be reflected less vertically and will be subjected to the 

 refracting influence of the lower strata of the air, the blue being 

 refracted most and the red least. Thus the figure c F A may 



