ON THE RAINFALL IN THE BRITISH ISLES. 175 



they were purposely read only once a month, in order to ascertain if any 

 material difference would be detected. 



Gauges -whose mouths are level with, or near to, the surface of the ground 

 are always found to collect more or less of the soil surrounding them, which 

 is splashed up by heavy rain and coUeetcd in the gauge. Rain-gauges 

 are usually so constructed that very little, if any, rain which falls within 

 their receiving area shall splash out and escape. Hence it seems probable 

 that gauges nearly level with the surrounding soil will always collect the 

 true rainfall, jj?«s insplashing. As a means of eliminating this source 

 of error, another gauge, of the same pat- 

 tern as the others, was added to the series 

 in April 1865; it was sunk in a pit, the " ' 

 depth of which was equal to the height of 

 the gauge, and with sloping sides, as in fig. 3. 

 By this means the receiving surface was 

 exactly level with the ground, and insplashing was impossible. This gauge 

 will be called the pit-gauge. 



These, then, are the instruments wherewith, and the conditions under 

 which, the observations at Calne were taken. We may now introduce some 

 of the results, some only, for the observations are so voluminous and com- 

 plete as to be almost inexhaustible. 



Table I. contains the total amount collected in each gauge in each month, 

 from August 1863 to December 1867, both inclusive. 



Table II. contains those monthly totals converted into ratios, the amount 

 measured in the gauge wjiose receiving surface is one foot above the surface 

 being taken as unity. 



Table III. is exactly similar, except that the " pit-gauge" (fig. 3) is taken 

 as unity. 



Table IV. contains the total fall in each year and in the whole period ; 

 also these values reduced to the ratios of the 1-foot gauge. 



Table V. contains the 1-foot ratios, grouped according to months. 



Table YI. contains the mean monthly values deduced from Table V. 



Table VII. contains the mean monthly values deduced from Table III. 



Table VIII. contains the monthly (1-foot) ratios for 1864 and 1865, 

 grouped according to the mean temperature of the days in each month on 

 which rain fell. 



Table IX. is similar, but grouped according to the Jiygrometric condition 

 of the atmosphere on those days. 



Table X. is similar, but grouped according to the mean velocity of the 

 ivind*. 



It is hoped that this series of Tables is so complete as to render lengthened 

 remarks unnecessary. 



"We have already referred to the pit-gauge, we may now point out its 

 result, viz. that in the winter it catches very much less than the one sur- 

 rounded by grass ; and that though in summer both catch alike, on the 

 average of the year the pit-gauge is 3 per cent, less than the " level," and 

 agrees almost exactly with the one whose orifice is 2 inches abo^■e the 

 ground. This agreement of the "pit-gauge" and the "2-inch" seems 

 to show that the amount collected by gauges thus placed is correct. The 

 accompanying diagrams (Plate V.) show the mean monthly and annual 

 deficiencies therefrom of gauges at various heights above the ground. 



* In the formation of the last three Tables some calculations by Dr. Barter of Bath 

 have been very useful. 



