Feb. 7, 1889] 



NATURE 



345 



drawer. The lower part of the cord is attached to a 

 cross-bar which unites the bottom of the two rings. 

 When the cord is stretched under the action of the 

 sounding-weight, the upper arm of the lever falls into the 

 prolongation of the cord, in a vertical position, and the 

 lower arm keeps the drawer close. When the action of 

 the sounding-weight ceases, the drawer opens under 

 the action of the antagonistic spring. The process is 

 the same as in the large apparatus, but in a slightly 

 different form, permitting a much smaller and lighter 

 order, which is more suitable to the superposition in 

 series. To avoid the lower apparatus interfering by their 

 additional weight with the opening of the higher ones, 

 when the action of the sounding-weight ceases, each 

 apparatus and the corresponding cord were exactly 

 counterbalanced by a floater of glass in the form of a 

 vial, inclosed in a net and attached to the cord imme- 

 diately below the apparatus which is wished to be free. 

 Under these conditions the apparatus all open simul- 

 taneously, immediately that the sounding-weight touches 

 the bottom, and also close all together at the moment when 

 the cord is drawn up and the sounding-weight begins 

 again to act on them. The simultaneous exposure of 

 several plates at various depths is thus obtained, and it is 

 possible to follow the decreasing action of the light with 

 thedepthinone and the same experiment, all circumstances 

 being equal." 



The sensitive plates used in these experiments were 

 those of extra rapid gelatine-bromide of M. Lumi^re, at 

 Lyons. They were protected by a varnish from the action 

 of the sea-water. The duration of the exposure and that 

 of the development were both ten minutes, as in the 

 preceding experiments. 



They were carried on in a locality presenting a depth 

 of about 500 metres, so that the purity of the water and 

 the limit of the light were not influenced by the nearness 

 to the bottom. The place chosen was 1300 to 1500 

 metres from the Cape of Mont Boron, separating the 

 strand of Villefranche from the Gulf of Nice. 



Amongst the series of plates obtained the following 

 succeeded well and are particularly instructive. 



5^m'i- ^.—Between 1.15 and 1.25, April 7, 1886, the 

 sun being about 60^ above the horizon. The sky was 

 very clear and the sun brilliant ; a moderate breeze from 

 the east made little waves. 



Plate I. Exposed at 430 metres: no trace of luminous 

 impression. 



Plate 2. Exposed at 390 to 393 metres : a very faint 

 trace, but yet a clear one. 



Plate 3. Exposed at 350 metres : a still faint 

 impression. 



Plate 4, Exposed at 310 metres : a strong impression. 



Plate 5. Exposed at 270 metres : a very strong , 

 impression. ' 



Plate 6. Exposed at 230 metres : completely blackened, 

 as were the following. 



The limit of light is, then, very exactly towards 400 

 metres in April, in the middle of the day, in fine weather. 

 This is as complete a confirmation as possible of the 

 conclusion arrived at in the preceding experiments. 



5^/-/Vj ^.—Between 8.20 and 8.30, April 5, 1886. Sky 

 veiled by a uniform layer of white clouds thick enough 

 for the sun to project no shadow. Moderate breeze from 

 the east. 



Plates I, of 450 metres, and 2, of 415, have no 

 impression. 



Plate 3, of 350 metres, presents a very slight impression, 

 a little less strong than that of plate 3 (390 metres) of 

 Series A. 



Plate 4, of 315, is of the same force as plate 3 of 

 Series A. 



Plate 5, failed by accident. 



Plate 6, of 245, and the following, are completely 

 blackened. 



Series C— Between 6.5 and 6.15, April 8. The setting 

 sun was hidden by a bank of black clouds. The rest of 

 the sky was pretty clear, v.'ith some little cirro-strati with 

 a faint white light. The light was altogether faint, and 

 like that which is generally found when the sun has just 

 set. The surface of the sea was little agitated, with a 

 slight breeze from the west. 



Plates I, of 400 metres, 2, of 340 metres, and 3, of 3CX3 

 metres, have no trace of impression. 



Plate 4, of 260 metres, is of almost the same force as 

 plate 3 of Series A. 



Plate 5, of 220 metres, similar to plate 4 of Series A. 



Plate 6, of 180 metres, like plate 5 of Series A. 



Plate 7 and the following, completely blackened. 



The limit in this last series may be placed with all 

 probability at 290 to 295 metres from the surface. 



We see from these experiments that the layers situated 

 at 300 metres are lighted every day, not for a very short 

 time, but all the time that the sun is above the horizon ; 

 at 350 metres the light penetrates at least during eight 

 hours daily. 



According to the tables that M. Holatschek has drawn 

 up for the latitude of Vienna, especially after the photo- 

 chemical experiments of MM. Bunsen and Roscoe, the 

 actinic intensity of the light of the blue sky would be on 

 April 21, 33 at 8.30 in the morning ; 3807, at noon ; and 

 14" 1 8 at 6 p.m. ; that of the sky and the sun at once would 

 be on an average, in April, 75 at 8.30 a.m., 133 at noon, 

 and 15 at 6 p.m. 



According to these figures, the depth that the actinic 

 rays attain in the sea after the setting of the sun is very 

 remarkable. The Commission wait, however, to have 

 more numerous experimental proofs to try to calculate 

 a formula of absorption of which they have to determine 

 the constant for sea-water. 



5. New Experiments on the Effect of the Seasons on the 

 Limit pf Penetration of Daylight in the Lake of Geneva. 



The results of these, which require to be confirmed under 

 better conditions, seem to indicate that the difference of 

 transparency between the waters of the lake in winter 

 and in summer is greater than one would have thought. 

 On the other hand, the want of agreement with the 

 experiments of Series 3 leads one to think that the dis- 

 tribution in extent and depth of the layers of troubled 

 water brought by the afifluents of the lake is subject to 

 variations dif^cult to foresee and to appreciate. 



The Report concludes by referring to certain improve- 

 ments in the apparatus. 



THE REPORT OF THE KRAKATAO 

 COMMITTEE OF THE ROYAL SOCIETY. 



PART IV. of this Report, on the optical phenomena, 

 by the Hon. F. A. Rollo Russell and Mr. E. 

 Douglas Archibald, comprises 311 pages of letterpress, 

 or the major portion of the work. Owing to the enor- 

 mous mass of material which went on accumulating 

 for nearly four years from the date of the eruption, as 

 well as the complexity surrounding the optical pheno- 

 mena, which in some cases were entirely novel, and in 

 others differed both in quality as well as intensity from 

 their normal analogues, it was plainly a work of some 

 considerable difficulty to decide how best to arrange and 

 discuss the data, as well as to avoid arriving at hasty 

 conclusions from the first indications of appearances 

 which continued in part right up to the beginning of 

 1887. When the wonderful sunsets appeared in this 

 country, the idea of their being connected with the erup- 

 tion of Krakatab was first suggested and traced out with 

 remarkable clearness by Mr. J. Norman Lockyer, F.R.S., 

 in his article in the Times (December 8, 1883). 



From the results of certain experiments by himself, hg 



