476 



NATURE 



[Marc/1,14, 1889 



look at the darkness of space. The blue colour, according to 

 his view, is supposed to be caused by reflection from minute 

 particles, which can reflect chiefly the blue rays by reason of 

 their small size. Experiments on highly attenuated vapours 

 during condensation to cloudy matter were the basis of this 

 reasoning. It always seemed to me that if a view be seen 

 through a turbid medium which reflects chiefly the blue rays, it 

 would not appear blue, but the complementary colour, yello\y ; 

 and therefore this theory could not account for the blue of dis- 

 tance. In fact, when a light mist hangs over the surface of the 

 earth, and the rays of the sun are transmitted in a direction 

 approaching the horizontal, the result is that the sun and all 

 objects lying in the direction looking towards it appear yellow, 

 while the mist in the opposite direction appears blue, and only 

 translucent, not transparent. The blue of the sky, if caused by 

 such a similar action of floating particles, would not be seen 

 when the sun was overhead, nor could it be seen by looking in 

 the direction of the fun. 



The blue would not be transparent and in character similar to 

 the blue of a clear distance, in which the outlines of mountains 

 and rocks are perfectly distinct and sharp, the shadows being of 

 an intensely deep blue, and the most distant objects the deepest 

 in colour. In 1880, Messrs. Hautefeuille and Chappuis liquefied 

 ozone, and found that its colour was indigo bhie (Com/>ies rendtts, 

 xcv. p. 522). On December 12, 1880, M. Chappuis presented 

 the Academy of Sciences of Paris with a paper on the visible 

 spectrum of ozone. He recognized the most easily visible 

 of the absorption-bands of ozone in the solar spectrum, and in 

 consequence he stated that a theory of the blue colour of the 

 sky could not be established without taking into account the 

 presence of ozone in the atmosphere, for the luminous rays 

 which reach us will of necessity be coloured blue by their trans- 

 mission through the ozone contained in the atmosphere. And 

 since ozone is an important constituent of the upper atmosphere, 

 its blue colour certainly plays an important part in the colour of 

 the sky. In March 1881, quantitative experiments made by me 

 were published to show how much of blueness could be com- 

 municated to layers of gas of different thicknesses when given 

 volumes of ozone are present. I showed that ozone is a normal 

 constituent in the upper atmosphere, that it is commonly present 

 in fresh air, and I accounted for its abundance during the pre- 

 valence of westerly and south-westerly winds. It was likewise 

 shown that it was impossible to pass rays of light through as 

 much as 5 miles of air without the rays being coloured sky-blue by 

 the ozone commonly present, and that the blue of objects viewed 

 on a clear day at greater distances up to 35 or 50 miles must be 

 almost entirely the blueness of ozone in the air. The quantity 

 of ozone giving a full sky-blue tint in a tube only 2 feet in length 

 is i\ milligrammes in each square centimetre of sectional area of 

 the tube. It is necessary to mention that a theory of the blue 

 of the sky was propounded by M, Lallemand (" Sur la Polarisa- 

 tion et la Fluorescence de 1' Atmosphere," Coviptes rendtis, Ixxv. 

 p. 707, 1872) after his observations had been found inconsistent 

 with all previous explanations. If the coloration be due to 

 leflection from minute particles of floating matter, or if it be due 

 to white light being transmitted through a blue gas, the b'.ue 

 portion of the sky should be polarized quite as much as white 

 light coming from the same direction in the heavens. But the 

 experiments of M. Lallemand prove that this is not so. Upon 

 these experiments he bases his theory that the blue colour of the 

 atmosphere is due to a blue fluorescence like that seen in acid 

 solutions of sulphate of quinine — that is to say, caused by a 

 change of refrangibility in the ultra-violet rays. 



Angstrom first threw out the idea of fluorescence being a pro- 

 perty of certain gases in the atmosphere. To possess this 

 property the gas mu.;t be capable of absorbing either in part or 

 entirely the ultra-violet and violet rays, and of emitting them 

 with a lowered refrangibility and without being polarized. 

 Ozone possesses the property of absorption in the highest degree 

 in the ultra-violet region, and I have now to announce that 

 strongly ozonized oxygen is highly fluorescent when seen in a 

 glass bottle two inches in diameter illuminated by an electric 

 spark passing between cadmium electrodes. The colour of the 

 fluorescence is a beautiful steel blue. This fluorescence has not 

 lieen observed in other gases, but it is in the highest degree 

 probable that oxygen is fluorescent, though this has yet to be 

 proved. There can be, however, little doubt that the colour of 

 the sky is caused in part by the fluorescence of ozone, and also 

 (o some extent by the transmission of rays through the blue gas. 

 The blue of distance is doubtless to be attributed more to trans- 



mission than the blue of the sky, though it is quite conceivable 

 that fluorescence also here comes into play. Whatever other 

 cause concurs in the production of the blue of the heavens, it has 

 certainly been established by M. Chappuis that the properties of 

 ozone participate in its production. 



In August 1884, a very short note was sent by me to Nature 

 concerning the red solar halo seen at Zermatt and on the Riffel- 

 berg with great distinctness. I recorded the occurrence of a 

 dark band in the spectrum, slightly more refrangible than D, 

 which was seen to vary in intensity ; a second band a little less 

 refrangible than D was also observed. On account of the 

 altitude at which the observations were made, viz. 9000 feet, 

 and the state of the weather at the time, these bands were 

 considered to be due to some constituent of dry air. 



The subject of the telluric rays has become of increased 

 interest since M. Cornu has studied the dark lines in the neigh- 

 bourhood of D, but unfortunately the rays absorbed to which I 

 refer are both a little more and a little less refrangible than those 

 figured on his map of this region. If we accept the number 5890 

 tenth-metrets as approximately representing the mean value of 

 the lines D^ D"^, the narrow bands observed by me have wave- 

 lengths about (l) 5950 and (2) 5770 at their darkest parts, as 

 far as one can ascertain with a hand spectroscope giving excellent 

 definition but small dispersion. They are very variable, being 

 dependent on the state of the v\ eather, and are more distinct and 

 broader when viewed with the sun on the horizon. In London 

 during the dry calm weather of June and July 1884, they were 

 very strong, but variable in different parts of the sky. 



The less refrangible band, or broad line as it usually appears, 

 below D, is generally over-lapped by a band belonging to water 

 vapour, the chief "rain-band." On this account observations 

 at an elevation of 10,000 feet or so during perfectly dry weather 

 were considered of interest. The bands were observed against 

 the blue sky on several occasions, but they were also at other 

 times entirely absent or barely visible. There is some liability 

 to a group of iron, barium, and other solar lines being mistaken 

 for the more refrangible band when it is not decidedly strong. 

 Chappuis observed bands in the blue sky coincident with ozone 

 bands, and I have on that account always expected to find some 

 indication of the spectrum of ozone in the upper atmosphere, 

 but the reason why there must always be a difficulty in obtaining 

 evidence of any absorption due to this substance arises from the 

 strongest visible band of ozone, with wave-length 6095 to 5935. 

 being masked by the band of water vapour ; and secondly, 

 because the total amount of white light so preponderates as to 

 overpower the effect of absorption — that is to say, the rays 

 absorbed are only a small fraction of those transmitted, so that 

 the bands are faint and the colour due to absorption is either not 

 seen or seen only with difficulty. Owing to this fact we cannot 

 distinguish the blue colour of the clouds when the sunlight is 

 bright ; but when the sky is completely over-clouded with 

 cumuli a faintly bluish tint is given to the cloud- shadows, even 

 at the zenith. Near the horizon not only are the bright parts 

 of the clouds blue, but their shadows have a rich blue tint.^ 

 The blueness varies somewhat : at times it may be seen to 

 shift about in the sky ; it has been observed, for instance, to 

 pass over from south-west to north-east. The second but less 

 conspicuous band of ozone absorbs rays with wave-length 577° 

 to 5600. Both bands have been observed in a dry atmosphere 

 at elevations varying fiom 6000 to 10,000 feet, both in the blue 

 of the sky and against white clouds. The measurements, very 

 imperfectly made under difficulties, showed them to have wave- 

 lengths about (i) 5950, (2) 5770, in the centre of the dark 

 portion, while, according to Chappuis, the bands of ozone are — 



(i) 6095 to 5935 ... mean, 6010 

 (2) 5770 to 5600 ... mean, 5680. 



On Angstrom's chart, a dark band, diminishing in depth 

 towards the ea>t, extends from 5785 to 5680, which is classed 

 among the raies atviospheriques ; this is similar to the band 

 observed by me when viewing the sun or bright clouds near the 

 horizon, and is similar to the second ozone band. 



The work of i'rof Piazzi-Smyth, " Madeira Spectroscopic," 

 does not give that portion of the spectrum which would serve for 

 comparison. The "low-sun band," 5, comes very near to band 

 (2), wave-length 5770, while the "rain-band" comes very near 

 (i), 6095. On several subsequent occasions the two bands were 



' Prof. Pickering has proved that sunlight as it reaches us is blue, which 

 must be the case if it has passed thn u^h a blue medium (Proc. American 

 Academy of Sciences, 1880, p, 236). 



