CORRESPONDING SOCIETIES. 441 



A more difficult problem, and one up to the present only partially solved, 

 lies in the most suitable anethod of exposing a rain gauge so that it shall give 

 as true as possible an account of the amount of rain falling. The earlier attempts 

 to solve this problem suffered from an incomplete appreciation of the real nature 

 of the difficulty. Numerous experiments, among the most elaborate of which 

 were those carried out at Rotherham from 1868 to 1890,^ confirmed the previously, 

 observed fact that with practically no exception, rain gauges exposed at a con- 

 siderable height above the ground caught an amount of rain diminishing with 

 elevation. An extraordinary volume of discussion took place lOS to the reason 

 of this, and it was many years before it was at all generally recognized that 

 as a matter of fact the variations observed depended not upon the amount of 

 rain falling in the vicinity of the gauge, but solely upon the proportion of that 

 amount which the gauge was capable of intercepting. Much attention was 

 devoted to quantitative observation of the variations, largely wasted labour, 

 for subsequent experience has shown that different conditions of exposure to 

 wind produce entirely different variations, the diminution of catch depending 

 entirely upon wind and little, if at all, upon mere elevation. 



The practical outcome was, however, eminently useful, the experiments 

 leading directly to the adoption of one foot above ground as the standard height 

 for the top of the funnel of a rain gauge. This rule, once decided upon, was 

 rapidly brought into general use, and at the present day a very large percentage 

 indeed of the rain gauges in use in the liritish Isles are placed at one foot above 

 ground. It is to be observed that the British practice in this respect is at 

 variance with the Continental, exposure at 1-5 metre being recommended in most 

 countries of Europe. So far as I am aware, the only valid objection which has 

 been urged to the exposure of gauges at one foot is that in case of deep snow the 

 instrument may be completely buried. Whilst the risk of this occurring on 

 rare occasions is no doubt obviated by elevating the gauge, it appears to have 

 been overlooked that the loss of catch occasioned by elevating a gauge is, as 

 a rule, far greater in snow than in rain, and except possibly in localities where 

 deep snow frequently falls, or in which drifts are of common occurrence, the 

 remedy is worse than the disease. 



The advantage of placing gauges with the rim comparatively close to the 

 ground lies entirely in the fact that when a gauge is freely exposed to the 

 sweep of the wind, eddies are apt to be set up round the funnel, which prevent 

 raindrops from entering. Close to the ground wind movement is at a minimvun, 

 and the rain falls more nearly vertically than at greater elevations. In a 

 sufficiently sheltered site, such as a walled garden, no diminution of catch 

 will be observed, even if the gauge is raised to several feet above the gromid, 

 the requisite wind check being provided. The greater relative loss in elevated 

 gauges during snow is due to the greater facility with which snow-flakes are 

 carried by wind than are rain-drops. In confirmation of the fact that it is 

 over-exposure and not mere elevation which causes the diminution of catch with 

 height, it will be found that a gauge placed on a flat house roof, provided with 

 parapets to break the wind, will indicate as much as one on the ground, whereas 

 if there are no parapets the catch will be reduced by from 20 to 40 per cent. 

 The formation of wind-eddies is very much aggravated if a gauge, in addition 

 to being exposed to wind, is placed on a site sloping downwards in the direction 

 of the prevailing wind. Land sloping to the windward is highly detrimental 

 even if a considerable distance from the gauge whilst the latter stands upon 

 level ground, unless some effective wind-screen intervenes. Thus a gauge should 

 not be placed on a hillock or on a terrace. A sloping house-roof is an extremely 

 bad position. 



I have frequently seen rain gauges placed on the ridges of high hills for 

 the purpose of ascertaining the rainfall at the summit. Such gauges, owing 

 to the sweep of wind, almost invariably indicate smaller falls than the valley 

 gauges on both sides, although the actual rainfall at the top is, as a rule, greater. 

 The frequency of snow in these exposed situations adds to the inaccuracy. 



The defect in the catch of rain in the case of gauges exposed under con- 

 ditions such as those described, increases with increase in the average wind 



» See British Rainfall, 1868 to 1890. 



