August 9, 19 17] 



NATURE 



467 



During two years, half the time is devoted to 

 research and half to preparation for a professional 

 <legree. At the end of the period these students 

 •enter the research department of the firm. 



The Mellon Institute of Industrial Research, 

 attached to the University of Pittsburg", was 

 erected by Mellon Brothers, bankers, to provide 

 manufacturers with the use of a well-equipped 

 laboratory and trained staff at less cost than 

 the establishment of a works laboratory. Any 

 manufacturer requiring- a subject investigated can 

 endow a fellowship for one or more years, paying 

 from looL to 400Z. and also the cost of any 

 special apparatus. The building cost 50,000/., 

 and the equipment 16,000/. A staff of seven men 

 of high attainments supervises the researches. The 

 director of the institute selects the fellows, usually 

 men with a doctor's degree. Seventy-five fellows 

 have been appointed in five years. The total 

 amount spent in salaries and maintenance is 

 30,000/. a year. 



A National Research Council has recently been 

 appointed by the Academy of Sciences, at the 

 request of President Wilson, to co-ordinate the 

 scientific research work of the country. 



The report, of which this is a very brief ac- 

 count, is extremely comprehensive, and should be 

 read by all interested in the industrial progress of 

 this country. The author draws some general 

 conclusions, and suggests the establishment of an 

 Imperial Industrial Research Laboratory, say, in 

 the Midlands, controlled by a board largely com- 

 posed of manufacturers. 



RAINFALL AND GUNFIRE. 



A /f .A.XGOT, the eminent director of the French 

 •i-^** Meteorological Service, has made a valu- 

 able and authoritative contribution, published in 

 the Journal of the French Academy of .A^griculture 

 for May, to the literature of a well-worn con- 

 troversy. The alleged connection between rainfall 

 and gunfire, in favour of which so many champions 

 sprang up during the wet p>eriods of 1914-16, has 

 recently lost favour as a subject for argument, 

 owing, no doubt, to the coincidence of the spring 

 drought of 1917 with the .\Ilied offensive on the 

 Western front; but so short is the public memory, 

 especially for negative evidence, that the incidence 

 of 3 in. of rain during a recent summer afternoon 

 in North-West London has proved sufficient to 

 disinter the bone of contention. The mental atti- 

 tude of the public towards a theory of this 

 nature is of great psychological interest : there is 

 little doubt that, should we experience this summer 

 a repetition of the weather of July, 1888, when 

 snow fell in London, followed by a recur- 

 rence of that of August, 191 1, when the 

 thermometer touched 100° F. at Greenwich, both 

 phenomena would generally be attributed to the 

 war. j 



Accordingly M. Angot's paper reaches us at an 

 opportune moment. After dealing briefly with the 

 historical aspect of the question, and alluding to 

 the work of M. Le Maout — who, not content with 

 ihaving established a connection between the bom- 

 XO. 2493, VOL. 99] 



bardments of the Crimean War and the rainfall of 

 India, the United States, Nicaragua, and Barbados, 

 went on to ascribe the diurnal variation of the 

 barometer to the striking of public clocks and the 

 ringing of church-bells — M. Angot proceeds to 

 consider the physical changes which could be 

 effected by the discharge of artillery, and could 

 at the same time be held responsible for the causa- 

 tion, increase, or acceleration of rainfall. 



The first proposition is that a succession of 

 violent explosions might result in the displacement 

 of masses of cold air at certain heights, which, com- 

 ing under the influence of the upper winds and 

 encountering layers of warmer, saturated air, 

 could give rise to precipitation which would not 

 otherw^ise have occurred : in this connection the 

 author points out that in order to obtain a rainfall 

 of so small an order as i mm, (o'o4 in.), even if 

 one were to take two equal masses of saturated 

 air, the one at a temjjerature of 0° C, the other 

 at 20° C. (an extreme case, of course), it would 

 be necessary to effect a rapid and thorough inter- 

 mingling of the two throughout a layer of air 6850 

 metres in thickness. In M. Angot's opinion, the 

 mixing of layers of air may be the cause of cloud- 

 formatioi\ or of slight drizzle at the earth's surface, 

 but can never be responsible for considerable pre- 

 cipitation. 



In the case of the second proposition — that 

 water-vapour resulting from chemical reaction of 

 the explosives might take effect — it is asserted that 

 in order to produce the same amount of rainfall 

 (i mm.) as in the previous proposition the employ- 

 ment of no fewer than 21,750 tons of melinite per 

 square mile would be necessitated — that, indeed, 

 only on the supposition that all the hydrogen in 

 the explosive became water-vaf>our which con- 

 densed immediately in its entirety and, so to 

 speak, on the spot. 



In the third and last instance, the possibility of 

 electrical action being brought into play is con- 

 sidered in some detail. We know that super- 

 saturated air {i.e. air which contains more water- 

 vapour than it normally should be able to hold for 

 the existing temperature) is a physical possibility, 

 in the absence of dust-particles or other matter 

 which may form nuclei for condensation. The 

 necessary medium may be supplied by the action 

 of ozone, of ultra-violet rays, by any cause, in 

 fact, which can set up ionisation of the atmo- 

 sohere ; under this last category may be classed 

 the detonation of high explosives, inasmuch as 

 highly ionised gases result therefrom. The lower 

 regions of the atmosphere, however, which alone 

 are the seat of explosive activity on a large scale, 

 always harbour large numbers of both ions and 

 dust-particles, and cannot, therefore, be subject to 

 sup>ersaturation ; while it has yet to be shown that 

 the addition of quantities of ions or of dust- 

 particles to a stratum of atmosphere nearly, but 

 not quite, saturated can bring about premature con- 

 densation. .Assuming for the moment the possi- 

 bility of such a hypothesis, we must consider that 

 no outp>ouring of ions or dust-particles 

 can do more than accelerate a precipitation 

 which would be necessitated sooner or later 



