I40 



NA TURE 



[June 6, 1901 



observer and the source of the scattered light which reaches him, 

 by the absorption of the short waves by interposed water vapour 

 and by the admixture of white light reflected from the larger 

 particles. 



The scattering to which we have been referring is evidently 

 different from what we ordinarily mean by reflection ; the latter 

 assumes that the reflecting surfaces have an area large as com- 

 pared with >3 ; whereas scattering assumes that the volume of 

 the particle must be small as compared with \''. 



Such is in outline the theory and the main facts in regard to 

 the cause of blue sky light ; but there are several secondary 

 features which must be now considered. The sky is bluer in 

 the zenith than elsewhere, evidently because the path traversed 

 by the scattered light is here the shortest, so that it suffers less 

 admixture with white light and less absorption of blue light. 

 Conversely it should be less blue near the horizon, and when 

 the sun is low may take on a red or orange tint, as we know 

 is the case. The light from the zenith is most intense when the 

 sun is nearest it, as at true noon, and its blue is least pure at the 

 hottest part of the day, on account of the maximum amount of 

 large particles of dust and vapour constituting the haze existing 

 at this time. 



Arago discovered that there is a point, about 15° above the 

 point diametrically opposite the sun (the antisolar point), where 

 the polarisation is zero ; between this and the horizon 

 the polarisation is horizontal. Babinet discovered a similar 

 point above the sun, and Brewster found one below it. Between 

 the neutral points discovered by Babinet and by Brewster the 

 polarisation is horizontal ; below Brewster's point and above 

 Babinet's it is vertical. For a little way on each side of the 

 neutral points the plane of polarisation is inclined at about 45° 

 10 the vertical. This seemed to indicate that superposed upon 

 the polarisation resulting from the scattering of direct sunlight 

 is a horizontal polarisation due to some secondary cause. It 

 was soon suggested that the horizontal polarisation is due to 

 a secondary scattering of the light coming from the lower layers 

 of the atmosphere, and this has generally, but not universally, 

 been accepted as the most probable explanation. Other neutral 

 points have been observed under rare conditions. 



The positions of the neutral points, the amount of polarisation, 

 the position of the point of maximum polarisation, as well 

 as the colour of the sky, are intimately connected with other 

 meteorological phenomena, but as yet the observations have been 

 so meagre, made under such dissimilar conditions and by such 

 various forms of apparatus, that it is nearly impossible to tell 

 what is the true connection. 



Cornu says — in words of which the following is a translation — 

 " In a general way, the amount of polarised sky light is connected 

 in so direct a manner with the condition of the atmosphere that 

 I have been led to think that it is characteristic of the state of 

 the atmosphere. The greatest clearness of the sky corresponds 

 to the greatest amount of polarisation ; cirrus and fog decrease 

 the amount, and even completely destroy the polarisation when 

 the sky is overcast .... \Vhat is particularly interesting is that 

 the least change in the state of the atmosphere is plainly shown 

 by the polarimeter several hours before other precursory phen- 

 omena (barometric variation, halos and various other optical 

 phenomena) have begun to indicate a change. 



" Under these conditions it would be useful to carry out these 

 observations in a methodical manner, and to compare the 

 polarimetric variations with other elements characteristic of the 

 atmospheric condition. . . . The amount of polarisation increases 

 as the sun sinks below the horizon until it reaches a certain 

 maximum, after which the polarisation rapidly disappears. 

 The law of this increase of polarisation with the time is very 

 important, for it appears to me to give the vertical distribution 

 of fog in the atmosphere ; indeed, if the increase is rapid the 

 lower layers are foggy and the upper ones transparent ; if the 

 increase is slow, the atmosphere is more homogeneous." 



In short, the more fog or cloud there is present the less the 

 amount of polarisation and the less pure is the blue of the sky. 



The most extensive series of observations are those of 

 Kubenson and of Brewster on the polarisation, and of Crova 

 and Abney on the colour of the light from the sky. The first 

 limited himself to observations made in fairly clear weather, and 

 the second directed his attention principally to the determination 

 of the positions of the various neutral points. Rubenson and 

 most other observers have laid special stress upon the intensity 

 of the polarisation at its maximum point in the vertical circle 

 Ihriiugh the sun. This is undoubtedly the point where observ- 



NO, 1649; VOL. 64] 



ations can be most easily taken, and those so obtained must be 

 of great meteorological value ; but the interpretation of them is 

 rendered difficult by the variation in the length of the path of 

 the scattered light at different times of the day. At sunrise and 

 sunset the point observed is the zenith, and the path is a minimum ; 

 while at noon, if the observer be in the tropics, the point ob- 

 served may be on the horizon, and the length of the path a 

 maximum. For other positions on the surface of the earth the 

 variation in length of path is less than this. 



On the other hand, unless we observe a point of maximum 

 polarisation the observations will be vitiated by every error in 

 determining the position, with respect to the sun, of the point 

 observed. Though other objections may be urged, it has 

 occurred to me that for meteorological prediction the most 

 valuable data would be obtained from continuous observations of 

 the amount of the polarisation of the light from points of the 

 sky on the horizon and 90° distant from the sun. These are 

 points of maximum polarisation ; these observations will give a 

 kind of integration of the atmospheric conditions over a large 

 area, and the length of path being the same at all times the 

 observations should all be comparable, except for the varying 

 angle of illumination of the surface of the earth, which, unless 

 the nature of the surface differs greatly in different directions, I 

 think would hardly affect the results appreciably, except, 

 perhaps, when the sun is near the horizon. No one, to my 

 knowledge, has carried out such a series of observations, hence 

 the suggestion is advanced with great hesitation. 



Since the colour of the sky is independent of the angular 

 distance of the point observed from the sun, being a function of 

 only the state of the atmosphere and the thickness of the stratum 

 observed, there is but little choice in the altitude of the point 

 where we make the colour observations. But since the blue is 

 a maximum in the zenith this is rather to be preferred, for a 

 slight error in the position of the point observed will here 

 produce the least effect. 



Whatever point or points are observed, the fact remains that 

 careful observations on the colour and the polarisation of the 

 light from the sky will give us data determining the amount and 

 size of the particles floating in the air, be they dust or water, 

 and, as any change in the state of the atmosphere will affect 

 these quantities, such observations should be of ever-increasing 

 importance to meteorology. First, however, we must have a 

 long series of observations taken at different places and under 

 all conditions, with exact meteorological data obtained at the 

 same time and place, together with a description of the nature 

 of the surrounding country. When these have been obtained 

 it should be not very difficult to find means of using future 

 observations with great success. 



UNIVERSITY AND EDUCATIONAL 

 INTELLIGENCE. 



The Senate of the University of Dublin has decided to confer 

 the honorary degree of D. Sc. upon Prof. W. Burnside, F. R.S., 

 and Mr. W. E. Wilson, F.R.S. 



On Tuesday, June 18, Lord Avebury will open an exhibition 

 of students' practical work, executed in connection with the 

 technological examinations of the City and Guilds of London 

 Institute, at the hall cf the University of London. 



The Report of the Council of the City and Guilds of London 

 Institute upon the work of the Institute during last year refers 

 to a number of noteworthy matters. The Institute has been 

 incorporated by Royal Charter, but the general constitution re- 

 mains unchanged. The Central Technical College has become 

 a School in the Faculty of Engineering of the University of 

 London. The Departmental Committee appointed to consider 

 "the best means for coordinating the technological work of 

 the Board of Education with that at present carried on by other 

 educational organisations " has had several meetings, and it is 

 hoped that arrangements may be made for the more intimate 

 association of the work of the Institute's Technological Ex- 

 aminations Department with that of the Board, by which the 

 overlapping of examinations may be avoided and the instruction 

 provided by county councils and technical schools may be 

 brought into closer relationship with the Board of Education 

 and the Institute. Referring to the extrance examination and 

 the teaching of science in secondary schools, the Council re- 

 marks : " The Central Technical College is the only college of 



