2 24 



NA TURE 



{Jan. 3, 1884 



have taken place at a height of about fifty English 

 miles. 



Does it not strike you that the glow was observed at 

 earlier periods the more we advance towards the east — 

 the source of the late Java eruptions ? li. Brauner 



Bohemian University, Prague, December 18, 1883 



The late splendid sunsets which have so vividly 

 attracted the attention of men of science and of the 

 general public were so remarkable and of so long a dura- 

 tion in the clear atmosphere of the Castilian tableland, 

 where sunsets are usually dull, that they have not failed 

 to impress observers with the notion that they were due 

 to other causes than those of common atmospheric re- 

 fraction and reflection. 



When the phenomena had already lasted four or five 

 days, I read Mr. Symons' letter, published in the Times 

 of the 1st inst., and I thought that possibly evidence 

 might be obtained towards the confirmation of this theory 

 if the sediment of fresh fallen snow was thoroughly in- 

 vestigated ; for if the dust of Krakatoa was really reflect- 

 ing in the higher regions of the atmosphere the sun's 

 rays, some of it must necessarily be descending towards 

 the earth. 



Luckily on the 7th of this month, and when the pheno- 

 menon was at its height, and had already lasted for about 

 eight days, there was a fall of snow at Madrid, of which 1 

 naturally profited, submitting it to a thorough investiga- 

 tion, the results of which, I think, will throw some light 

 on so remarkable a phenomenon. 



The snow analysed was obtained from what had fallen 

 on some zinc plates before the exposed windows to the 

 north of my house, which is situated at the extreme north 

 end of the town, where there are no buildings facing it, and 

 also from what my friend Dr. Francisco Ouisoga gathered 

 from the windows of his house, situated about a mile to 

 the south-east of mine ; and in both the same substances 

 were found. 



The snow yielded about a litre of water, which, when 

 the sediment had collected, was decanted, and the solid 

 part dried at a temperature below that of boiling water. 

 The dry powder was then tested for magnetism and 

 it was found to be extremely magnetic. It was then 

 incinerated on platinum foil to a bright red heat so as to 

 destroy organic substances, and the remaining dust was 

 then submitted to microscopical investigation. The 

 greater part of it is made up of what probably is the 

 natural dust of the atmosphere of Madrid ; of particles of 

 mica, generally brown, and similar to that of the Guadar- 

 rama range, and in various states of decomposition, 

 splinters of quartz and felspar, the greater part uf it ortho- 

 clase, some small fragments of tourmaline, magnetic iron, 

 and fragments of diatoms. Besides these mineral sub- 

 stances, which may probably be traced to the rocks 

 forming the vicinity of the capital, some others were 

 found for the presence of which it is difficult to account. 

 The most remarkable are small particles of a foliated 

 mineral of a yellowish colour, perceptibly dichroic, and 

 which between crossed Nichols is extinguished when the 

 cleavage traces are parallel to the principal section of the 

 polarising Nichol ; the interference colours being of 

 bright blue, and red, and yellow colours. Treated by 

 boiling hydrochloric acid for twenty minutes, not a trace of 

 action was perceived. These characters are all referable 

 to a rhombic pyro.xene, and judging from its dichroism 

 this substance may be taken for a hypersthene, which 

 has besides a most striking resemblance to volcanic 

 hypersthene. In addition to this mineral, small particles 

 are found which appear to be referable to common 

 pyro.xene of a yellowish colour, of active action in 

 polarised light, and the extinction not taking place 

 parallel to what seem to be the edges of the prism. 

 Besides these minerals some corpuscles are found of 

 hardly any action on polarised light, and sometimes full 



of globular concretions and other kinds of microliths, 

 which, if seen in products of a volcanic region, I would 

 not hesitate in considering of volcanic origin. 



These are the principal substances which an investiga- 

 tion of the sediment of the snow which fell in Madrid on 

 the 7th inst. have revealed, and though 1 am far from 

 asserting that what appears to be foreign to the atmo- 

 sphere of this part of the world is referable to the dust of 

 Krakatoa, if further analyses in other parts of the world 

 should show these same substances floating in the atmo- 

 sphere, there would be powerful reasons for inferring that 

 the gorgeous sunsets of the past months have been 

 brought about in conseciuence of that stupendous display 

 of the volcanic forces of our globe. 



It is already a remarkable coincidence that hypersthene 

 should have been found both by MM. Daubrde and 

 Renard in their respective analyses of the ashes collected 

 in the vicinity of Krakatoa. JOSEPH M.\CPHERSON 



Madrid, December 22, 1883 



Complying with the request contained in your "Notes" 

 of December 13 (p. 157), I would say that the appear- 

 ances, already fully described by so many of jour corre- 

 spondents, commenced here on December i. On that 

 day I made an entry in my note-book as follows : — "Per- 

 fectly calm at sunset, with a light haze of a rose tint 

 rolling away from overhead towards the west-south-west 

 horizon. The colours of the sky were a very pale green, red, 

 gold, and pink ; and, as the light faded away, the south- 

 west was one mass of deep rich red. The crescent moon 

 (a little over eighteen days old) in the refractor was of a 

 pale green colour, and the bright limb seemed to extend 

 to an extraordinary distance round the dark body. Baro- 

 meter falling." 



Again : " December 2. — Sky clouded over by i p.m. 

 Sunset, as seen between breaks in the clouds, was again 

 of a deep rich red. Barometer steady." 



" December 3. — Rainy and very dull. Barometer 

 steady." 



" December 4 — Sunset, as seen through the clouds 

 along the horizon, was again of a deep red colour, gradu- 

 ally shading off into a pale rose tint towards the zenith. 

 The moon, Fomalhaut, and Vega seemed to float in a 

 pale rose sea ; whilst thin fleecy clouds as they drifted 

 across the moon's face were of a beautiful pale green. 

 This appearance— as did that on the ist — lasted for about 

 an hour and a quarter after sunset ; the rest of the sky 

 being covered with clouds, some faintly reflecting the 

 various tints. Barometer falling." 



I should not omit to mention that the stmrises were 

 also, more or less, of similar character. Since the 4th 

 we have had very bad weather ; gales from both north 

 and south, heavy rains, and snow. Yet the sky, when 

 occasionally glimpsed at sunset, seems to bear traces of 

 the same appearances. W. E. J. 



Constantinople, December 21, 18S3 



In addition to the remarkable sunsets which have led 

 to such a large amount of correspondence in N.ATURE 

 and elsewhere, there is another and possibly a related 

 phenomenon to which my attention has been directed 

 during the last few weeks. From country friends I learn 

 that the nights, in the absence of the moon, and even 

 when cloudy, have been remarkably light for the time of 

 year. I cannot profess to have witnessed this phenome- 

 non myself, living as I do in the midst of London, where 

 the perpetual glare of gas renders any satisfactory esti- 

 mate of the atmospheric luminosity quite hopeless. It 

 would be interesting, however, to learn whether other 

 observers more favourably located have noticed this 

 efl'ect. It occurred to me that the phenomenon might 

 perhaps be connected with the volcanic dust theory of 

 the sunsets, being, in fact, a result of the slight phos- 

 phorescence of this dust. Whether the latter exhibits 

 any degree of phosphorescence could be readily deter- 



