On the Rain and Drainage - Waters at Rothainsted. 271 
The drain-gauges just described were constructed in Barn- 
field ; and the new large rain-gauge was afterwards constructed 
in their immediate neighbourhood. The soil at Rothamsted 
consists generally of a somewhat heavy loam, with a subsoil of 
clay, both mixed with flints, and lying on chalk, which, however, 
seldom comes very near the surface. In the present case the 
cultivated soil of the field was about 8 inches in depth ; this was 
succeeded by about 10 inches of friable clay, followed by subsoil 
of rather stiff clay. The whole of the experimental soil was 
above the chalk. The land had previously been under the 
ordinary arable culture of the farm. 
The gauges were constructed in the summer of 1870; since 
then a few alterations have been made. In November and 
December 1874, leakage from the outside being feared, the sides 
of the gauges were bared, the old walls coated with cement, and 
then thickened by an additional half-brick. Again, in February 
1879, the drainage from the 20-inch gauge appearing very 
excessive, one of the walls was bared in which a leak was sus- 
pected, and its external surface coated with cement. 
2. TJie Measured Drainage, and the Evaporation. — The amount 
of the monthly drainage through each of the three gauges, from 
September 1870 to the end of 1880, is shown in Table XIX.* 
The monthly drainages recorded in Table XIX. are in a few 
cases not those actually recorded, but a corrected figure. For 
instance, the recorded drainage from the 20-inch gauge in 
February 1879 was 6'734 inches : this amount was greatly 
above that of the other gauges, and there being some evidence 
of leakage from the outside, the record has been rejected, and the 
amount passing through the 40-inch gauge substituted in its 
place. Again, in the tremendous rainfall of August 2-3, 1879, 
some of the receivers overflowed, and one was floated and dis- 
connected by the rise of water in the chamber under the gauges. 
In this instance, and in some others of a somewhat similar 
kind, an estimated drainage deduced from a consideration of the 
amount of rainfall, and other facts of the case, has been adopted 
in place of the observed drainage, which was obviously incorrect. 
Another source of error has arisen from unequal drifts of snow 
on the surface of the three gauges. Thus, in April 1878, the 
recorded drainage through the 20-, 40-, and 60-inch gauges was 
respectively 2'249, 2*822, and 3"467 inches ; but as the receivers 
of the 20-inch gauge ran over slightly, while a small snow- 
drift had occurred on the surface of the 40-inch gauge, and a 
* A summary of the results for the first five harvest-years (Sept. 1 to 
Aug. 31 inclusive) was given at a meeting of the Institution of Oivil Engineers, 
Feb. 29, 1876, and is published with a few comments in the ' Minutes of Pro- 
ceedings ' of the Institution, vol. xlv., part iii. 
