34<5 



NATURE 



{Feb. 7, 1889 



had concluded that, while under ordinary circumstances 

 aqueous vapour preferentially stops by absorption the 

 more refrangible blue rays, under certain circumstances 

 somewhat analogous to the conditions under which the 

 blue and red solutions of gold were obtained by Faraday, 

 aqueous molecules exist which stop the red rays and 

 transmit the blue. Such so-called " red molecules " were 

 found to be larger than the " bkie molecules." 



Combining this with the facts which indicated an 

 unusual extension of volcanic dust and vapour into the 

 air by Krakatao, as well as the succession of dates, the 

 occurrence of white suns in the Indian Ocean and blue 

 suns at a distance and at noon, and the initially rapid 

 progress of the coloured suns and twilight glows round 

 the equator, and their gradual spread to the extra-tropics 

 (at first only faintly realized through lack of sufficient 

 data), he constructed the hypothesis that all the optical 

 effects witnessed in England in November and December 

 1883 were, like those which preceded them, nearer the 

 equator, traceable to the products of the August eruption 

 of Krakatao, carried thither by the upper currents of the 

 atmosphere. Although, in one or two minor details, 

 such as a supposed south-to north line of coloured suns 

 over India, the vastly greater mass of evidence ultimately 

 collected by the Committee enabled them to arrive at 

 a more correct conclusion, yet, in its main features, such 

 as the east to- west current along the equator, and the 

 concatenation of such at first apparently unconnected 

 phenomena as a sunset in London and a volcanic 

 eruption in Java, the work of the optical section of the 

 Committee has practically resulted in filling in the 

 framework sketched out by Mr. Lockyer. As time pro- 

 gressed, fresh data came pouring in, which not merely 

 testified to the universality of the phenomena over the 

 north and south temperate zones, but helped to fill up the 

 gaps which necessarily occurred over the oceans and near 

 the equator. 



In spite of all these links in the chain of circumstantial 

 evidence, many persons still continued to doubt the con- 

 nection of the extra-tropical twilight glows with the 

 analogous appearances in the tropics. In the case of the 

 blue and green suns, the evidence even at first, was too 

 strong to allow much doubt that they were in some way 

 or other connected with the eruption. Yet even so, the 

 rate at which they travelled (from 70 to 80 miles an 

 hour) along the equator was too much for some persons, 

 whose powers of imagination could hardly grasp the 

 enormous scale on which the operations were conducted ; 

 while in the case of the extra-tropics, all sorts of queer 

 and gratuitous hypotheses were put forward to account 

 for what they beheld from their own windows. 



Now, a very cursory glance at the general data and 

 evidence, as well as at the maps given in Section III., 

 will, we think, convince the most sceptical that the grand 

 series of optical appearances which were first seen in the 

 neighbourhood of Krakatao on the day of its great erup- 

 tion, extended themselves, at first rapidly in longitude, 

 and then slowly in latitude, until they finally embraced 

 the whole earth. It will also show that their arrival in 

 Euroje was but a mere incident in their spread over 

 a region fifty times as large. All this, however, has had 

 to be put forward in detail. 



Other points which have had to be described or 

 discussed were — 



(i) The proximate cause of the abnormal twilights, and 

 an explanation, as far as was possible, of the way in 

 which they differed from ordinary twilights, both in 

 quality and intensity. 



(2) The coloured suns, large corona round the sun and 

 moon, and the sky haze or eruption cloud which evidently 

 caused them. 



(3) Then came the geographical distribution, the height 

 and duration of the glows, a list of analogous phenomena 

 en former occasions, opidions put forward to account for 



the present series, and finally a general analysis of their 

 connection with the eruptions of Krakatao in detail, each 

 of which demanded a separate section. 



To give some idea of the principal facts and conclusions 

 of this part of the work, we will commence with the 

 abnormal twilights, considered as local phenomena. 



The phases of ordinary twilights have been investigated 

 with much attention by Kepler, Le Mairan, Ur. Hellmann, 

 and Dr. von Bezold,^ of whom the last discussed them 

 with wonderful clearness in 1863, and showed that certain 

 sequences of colour and intensity take place normally, 

 which have apparently been entirely overlooked until the 

 present series brought the subject again into notice. 



Thus the normal sunset consists chiefly of a series of 

 bands of colour parallel to the horizon in the west in the 

 order, from below upwards — red, orange, yellow, green, 

 blue, together with a purplish glow in the east over the 

 earth's shadow, called the " counter-glow." As the earth's 

 shadow moves upwards towards the zenith, and passes 

 invisibly across it, a reddish or purplish glow suddenly 

 appears above the coloured layers in the west, in a spot 

 which previously appeared of a peculiarly bright whitish 

 colour. This purple glow is substantially the "primary 

 glow," or, more definitely, " crste piirpiirlicht.'" It is 

 peculiar in appearing above the horizontal colours, and 

 in not extending far on either side of a vertical plane 

 through the sun and the spectator. As this glow sinks 

 on the horizon and spreads out laterally, it forms the first 

 red sunset. After its disappearance, under fa7>ourable 

 conditions, a second edition of twilight colours analogous 

 to the first commences, with a similar bright spot [dciin- 

 vierungschein) out of which a second purple light appears 

 to be suddenly developed, and sinks on to the horizon 

 as the secondary or " after-glow." 



These are the normal phases of a complete sunset 

 according to Dr. von Bezold, and the present series only 

 appear to be abnormal in exhibiting certain peculiar 

 yellow and greenish tints, a less-defined boundary of the 

 earth's shadow, together with a much greater brilliancy, 

 extension, and duration of the first, and particularly of the 

 second, purple glows. The horizontal layers were less 

 conspicuous than usual, and the abnormal extension of 

 the purple light made it appear as though there was an 

 inversion of the usual order of tints from below upwards. 



In order to explain these and other peculiarities which 

 we have not scope to describe, Mr. Russell starts with the 

 observed fact of a sky haze which, in the tropics, tended 

 to transmit blue or green rays in preference to red, and 

 assuming that all the usual elements which are included 

 under the term " optical diffusion " were present, viz. 

 diffraction, refraction, and reflection, describes what 

 should be the effects (i) assuming a haze composed of 

 opaque particles, and (2) one composed of very thin re- 

 flecting plates into which condition a large proportion of 

 the pumice ejected from Krakatao is shown to have been 

 transformed. His conclusion is that the distinctive 

 features of the KrakataT) glows were due mainly to re- 

 flection from these fine lamina:, of rays already tinted in a 

 certain order, by diffraction through the dust of the haze 

 layer and the lower atmosphere, as well as by the selective 

 absorption which ordinarily takes place in the more 

 humid horizontal layers near the earth's surface. The 

 direct as well as diffuse reflection by the plates and 

 opaque dust, which lay, as Mr. Archibald has shown in. 

 Section IV., at a height of from 50,000 to 100,000 feet, of 

 rays tinted in succession as both the direct and reflected 

 twilight boundaries followed the descending sun, and the 

 peculiar transmissive quality of the stratum for the more 

 refrangible rays, appear to afford a reasonable explanation 

 of the peculiar silvery glare, the unusual colouring, and the 

 unusual extension of the purple glows. 



It is admitted that diffraction played an important 

 part, as it does in ordinary sunsets (Lommel, for 



' P''gS. ^nn., BJ. cxxiii. (i863\ pp. 24C-76. 



