June 17, 1920] 



NATURE 



493 



these conditions should be observed, nor that the 

 observation of surface air temperature was an 

 exceedingly difficult matter. 



When an Army Meteorological Service uas 

 established in 191 5 it was a small unit which had 

 to justify its existence, and in the course of ex- 

 ploring fields of usefulness it found the artillery 

 ready and anxious for improved meteorological 

 information. The shell from a high-velocity gun 

 may rise to a height of 20,000 ft. or more, and 

 surface conditions may be a very misleading 

 guide. But to ask the gunner to use detailed 

 observations of wind and of atmospheric pressure 

 and temperature at different heights up to 

 20,000 ft. under active service conditions, and 

 without previous training, was useless. The 

 meteorologist, having found a sphere of useful- 

 ness, had to put his information in a form in 

 which it could be used with the existing artil- 

 lery organisation. It is already suggested in some 

 quarters that the meteorological service adequately 

 met the artillery's requirements during the war 

 without any peace-time organisation, and that 

 therefore it is unnecessary now to keep any close 

 liaison between the gunner and the meteorologist. 

 In the writer's view this is a great mistake. The 

 meteorological service was able to help the gunner 

 by doing work which the gunner could have done 

 if he had received the proper training, and it 

 is necessary that he should do this work for 

 himself in order to make the best use for his 

 particular gun of data supplied to him by the 

 meteorologist. 



The artillery organisation for meteorological 

 corrections consisted in the supply to gunners of 

 tables of variations in line and range produced by 

 winds constant in velocity and direction at all 

 heights and of variations in range produced by 

 changes in surface temperature and pressure, 

 based on the assumption that the ratio between 

 the actual air density and that assumed in the 

 construction of the range table was the same at 

 all heights. It was a fairly obvious first step to 

 suggest that the gunner should be supplied with 

 a fictitious wind such that, when used with the 

 usual table of variations, the proper correction 

 was applied for the cumulative effect on the pro- 

 jectile before reaching the target of a wind vary- 

 ing with height. Such a wind came to be called 

 the equivalent constant wind, or the ballistic 

 wind, and methods of estimating it were investi- 

 gated simultaneously by the Meteorological 

 Section, R.E., and the Anti-Aircraft Experimental 

 Section (A.A.E.S.) of the Munitions Inventions 

 Department (M.I.D.). At first the investigations 

 were entirely independent and from different points 

 of view, but later they were continued in close co- 

 operation under the sympathetic guidance of the 

 Ordnance Committee. 



An initial difficulty of great importance was 

 that the ballistic wind is not the same for any two 

 trajectories, even for the same wind distribution. 

 But, fortunately, the height to which a projectile 

 rises when fired on the flat is nearly the same for 



NO. 2642, VOL. 105] 



all projectiles which have the same time of flight, 

 and also the length of time which such projectiles 

 spend in any particular stratum of the atmosphere 

 is nearly the same. Thus, though the range of 

 a high-velocity gun may be double that of a 

 howitzer for the same time of flight, yet the pro- 

 jectiles in each case rise to nearly the same height, 

 and are affected by the same winds for nearly the 

 same length of time. To a first approximation, 

 therefore, the ballistic wind is the same for every 

 projectile having the same time of flight, and if 

 a selection of such winds for different times of 

 flight is given to the gunner, he can choose the 

 one most nearly suited to the conditions under 

 which a shoot is taking place. 



As a first approximation, in the calculation of 

 the ballistic wind it was assumed that the atmo- 

 sphere was stratified into several layers, and that 

 in each layer the wind was constant in velocity 

 and direction, though varying from layer to layer. 

 It was further assumed that the effect of the wind 

 in any layer on a projectile was proportional to 

 the time spent by the projectile in that layer and 

 to the density of the air. "Weighting factors" 

 for the portion of the total displacement of the 

 projectile caused by unit wind in any layer were 

 thus determined. Subsequent mathematical 

 analysis showed that the "weighting factors" 

 varied materially for each different trajectory, and 

 also differed for winds across and along the line 

 of fire. Considerable refinements were introduced 

 for the analysis of experimental shoots on which 

 the construction of range tables was based. The 

 researches of the A.A.E.S., M.I.D., though prin- 

 cipally directed towards anti-aircraft gunnery, in- 

 cluded careful and detailed discussions of varia- 

 tions in the trajectory of a shell produced by vary- 

 ing wind and density, and made the careful 

 analysis of such experimental shoots possible. For 

 a considerable period, however, the facilities 

 afforded by the field meteorological service in the 

 different theatres of war made possible much 

 greater accuracy of correction than had been 

 aimed at in the experimental shoots from which 

 range tables were compiled. Ultimately average 

 weighting factors, deduced from theoretical 

 factors computed in a large number of cases, were 

 adopted for different times of flight, and the 

 method of constructing the ballistic wind for use 

 in the field became standardised. 



The second step was the introduction of the 

 idea of ballistic density — a fictitious density such 

 that when used with the usual tables of variations 

 the proper correction is applied for the effect of 

 an abnormal vertical distribution of density. For- 

 tunately, the pressure and temperature which 

 practically determine the air's density may be 

 considered separately. If it is assumed that the 

 vertical temperature distribution is known and 

 remains unchanged while changes in pressure are 

 registered at the surface of the earth, it is easily 

 shown that there are proportional changes in 

 pressure, and therefore in density, at all heights. 

 Thus the surface barometer reading affords a real, 



