488 



NATURE 



[June 23,. 1910 



in the bed-rooms and bath-rooms of the Euston Hotel, the 

 supply coming from an artesian well. 



Ihere are two natural sources of extremely pure water 

 with which I am acquainted, and I cannot believe that the 

 water from the second source receives any colour or 

 appearance of blue through the reflection of light by fine 

 particles in suspension. The first is in the Fairy Loch 

 beside Loch Lomond, situated on a little promontory south 

 of Tarbert. It is difficult to see that there is any colour 

 in the water except at a point where it wells up from a 

 fissure in the rock and passes over a vein of milk-white 

 quartzite which crosses the bottom ; here it exhibits a 

 beautiful blue colour. 



The second source is the Wells of Dee, situated in the 

 Larig (Learg Gruamach) at the foot of Ben Macdhui, and 

 between it and Braeriach, about half-way between Dee- 

 side and Speyside. It is a small tarn or pool with a 

 bottom like that of three miles of the pass — nothing but 

 large pieces of splintered red granitic rock. It stands 

 about 2700 feet above the sea. The water, according to 

 my recollection, passes down underneath broken rocks in 

 a narrow rift in the mountain side, and is derived from 

 the melting of snow on its northern slope near the summit, 

 which is 1598 feet (1500 feet by my aneroid) higher and 

 above all vegetation. The pool is too small to be shown, 

 but the stream which runs out of it appears on the 

 Ordnance map (Sheet 64, i-inch scale), springing from the 

 highest point of the pass. Some small lochs on the oppo- 

 site side of the pass, about i^ miles further south, are 

 also called Wells of Dee, and are the principal source of 

 the river of that name. The bottom of the small pool is 

 visible everywhere, and its apparent colour varies in pro- 

 portion to its depth, being dull red near the sides, to a 

 brownish-purple where it is apparently deepest. The pure 

 blue colour of the water was only seen on putting a white 

 object, such as a piece of porcelain, into it. The effect of 

 the blue colour of the water on the light reflected from 

 the red rock at the bottom is to give it a purple tint. 



It is evident that the blue is wholly due to the absorp- 

 tion of rays of complementary colour, because if it were 

 not the reflection of blue rays by suspended fine particles 

 would be seen against a dark ground on looking into the 

 water. As a matter of fact, the water when undisturbed 

 on the surface was not visible ; it was very difficult to 

 form any idea of its depth, everything on the bottom 

 being sharply defined. These observations were made 

 under a diffused and subdued light in a very clear atmo- 

 sphere, the light being of uniform intensity over the whole 

 sky, which was entirely covered with small greyish clouds, 

 no direct sunlight or blue sky being anywhere visible. A 

 fact adverse to the view that the blue could be reflected 

 light is that the light which escapes reflection has a reddish- 

 golden colour. In a hazy atmosphere when the sun is 

 low and we look towards it, we see the golden colour ; in 

 the_ opposite direction we see the blue opalescence. The 

 vohite light from the sky traversed the water in two 

 directions to the bottom, and then, by reflection, back 

 again, and it is safe to say that these two opposite colours 

 would neutralise each other. W. N. Hartley. 



Royal College of Science, Dublin. 



The Temperature Conditions in Clouds. 



As one of those who expressed doubt as to the possi- 

 bilit}' of the existence of the temperature conditions in a 

 cloud described by Prof. Rotch at Winnipeg, I have been 

 greatly interested by the letters of Dr. Aitken and Mr. 

 Palmer (Nature, November 18, 1909; June 2, 1910), but 

 the examples which they quote do not present the same 

 difficulty as Rotch 's result, nor do they explain it. 



The increase of temperature at or above the upper sur- 

 face of clouds, which Dr. Aitken mentions, has been 

 frequently observed in kite ascents at various places, while 

 the two examples given by Palmer are (i) alto-cumulus, 

 a wave cloud of the Helmholtz type formed at a surface 

 of discontinuity : the temperature decreases upwards in the 

 cloud itself; (2) alto-stratus, a shallow cloud formed also 

 at a surface of discontinuity; here, too, the temperature 

 decreases upwards in the actual cloud. In neither case do 

 we attribute the temperature pecul'iaritv to the clouds, but 

 regard the clouds rather as the result' of the temperature 

 conditions. 



NO. 2 12 I, VOL'. 83] 



Rotch, however, found that in a cumulus cloud, 2 km 

 thick, the temperature increased from the base upwards 

 by more than 5° C, and the increase was most rapid in 

 the lower part of the cloud. 



Dr. Aitken suggests that the sun, shining on the upper 

 part of a cloud already formed and warming it, would 

 account for the phenomenon, or at least for inversions 

 near the upper surface of a cloud ; but jf the sun raised 

 the temperature of the upper part of the cloud, that part 

 would be no longer in equilibrium with its surroundings, 

 and would rise upwards. Its temperature would, in con- 

 sequence, fall under ordinary conditions until equilibrium 

 with the surrounding atmosphere again supervened. The 

 sunshine could only result in an actual increase of tempera- 

 ture if there existed already above the cloud an atmospheric 

 layer of higher temperature than that in the cloud itself. 



Now if Rotch 's cloud were formed by convection currents 

 according to the generall}^ accepted ideas, the summit 

 would be initially at least 10° C. colder than the base, and 

 consequently its temperature must have been raised 15° C. 

 to bring about the observed state of affairs. It is not easy 

 to imagine how this could be don^' without dissipating the 

 cloud, because it is unlikely that a cloud 2 km. higl: 

 would be formed by convection currents without the uppe 

 parts losing some of the water-vapour which they originally 

 contained, and in the present instance evaporation would 

 begin before the temperature had risen 10° C. Moreover, 

 the ascent took place about 9 a.m. in May; while assum- 

 ing that 35 per cent, of the incident sunshine is absorbe 

 by the cloud (Abbott and Fowle found 65 per cent, re 

 fleeted) and that no loss of heat by radiation occurred, i' 

 would take a twelve-hour day near the equator to raise 

 the mean temperature of a hemispherical cloud 2 km. high 

 by 9° C. It appears certain, therefore, that solar radiation 

 incident on the cloud cannot account for the phenomenon. 



The only reasonable explanation I can put forward is 

 that air below and above an inversion surface is lifted 

 bodily upwards sufficiently far for condensation to take 

 place on both sides. The balloon ascent must have been 

 made in a region of convergence, and the mechanism by 

 which the conditions were produced appears to have con- 

 sisted of a cold, damp easterl}' wind penetrating beneath 

 a warm upper current from a more southerly point. 



Cambridge, June 6. E. Gold. 



The Fertilising Influence of Sunlight. 



With reference to Dr. Russell's remarks on this subject 

 in Nature of April 28, I should like to remark (i) that 

 my point was not so much that toluene removed toxic 

 material from the soil as that it rendered it insoluble. 

 The question of washing out material from the soil wjtt 

 not raised by me. (2) and (3) Dr. Russell seems to beg 

 the question by taking " fertility " and " bacterial 

 activity " as synonymous. He has not, so far as I can 

 find, proved that the addition to partially sterilised soil of 

 an aqueous extract — or of a portion- — of an untreated soil 

 increases crop production (in contradistinction -to soil 

 fertility as indicated by bacterial activity and ammonii 

 production). If such is found to be the case, it wou: 

 certainly require further experiment before it could 

 explained on the toxic theory. 



With regard to water cultures, in one experiment th 

 water was boiled e%'ery two days in some of the culture 

 while in others it was not boi'ed. At the termination 1 

 the experiment — two days after the last boiling — thj 

 bacterial contents were found to be (per c.c.) : — 



J 2500 \ ^gj^j, _ 2 ,00 

 I 21CO I -^ 



In unboiled cultures 

 In boiled cultures 



I •'I°!-Mean= 375 

 \^ 400 / •^' -' 



The quantity of material precipitated by potassium sulphatt 

 from the two solutions was (per million) : — 



Unboiled solution 30 



Boiled solution ... ... ... ... ... 30 



It would appear that if this substance had been producec 

 by the bacteria there ought to have been at least sevee 

 times as much produced in the unboiled as in the boilet 

 solution, since the bacterial content of the latter' was nevo 

 more than one-seventh of the former, and must have beet 

 for most of the time almost nil. 



