314 On the Main and Drainage -Waters at Rothamsted. 
days will always result in more drainage and less evaporation 
than the same quantity of rain distributed over a month ; and 
to this cause some of the variations in drainage with a similar 
rainfall, to be found in Table XXI., are plainly due. In the 
Rothamsted drain-gauges, however, containing as they do a 
heavy soil in its natural condition of consolidation, and with 
its clay subsoil untouched, the differences in the amount of 
drainage due to irregularities in the distribution of the rain are 
far smaller than those which have been observed by other persons 
employing small percolators more or less loosely filled with 
porous soil, and generally with growing turf upon the surface. 
Turning now to the figures of the table, we see that the amount 
of evaporation from the soil during the whole drainage-year 
appears, with two marked exceptions, to be a fairly constant 
quantity. The whole range of evaporation in ten years is from 
14'279 to 19"686 inches; but excluding these extremes, the 
variation in eight years is only from 16'861 to 18'635 inches. 
That in years of very different rainfall the soil has evaporated 
almost the same amounts of water, is strikingly shown by many of 
the results, as for instance those for 1873-4, 1872-3, and 1876-7. 
The large amount of evaporation credited to the year 1870-1, 
is probably due to an error in estimation of the kind already 
noticed. The drain-gauges were constructed during the ex- 
ceedingly dry summer of 1870, and the blocks of soil to be 
included in the gauge having been isolated by trenches cut 
round them, were necessarily more or less exposed to air on 
all their sides ; the soil was thus dried to an unusual extent, and 
when rain commenced a considerable amount was consumed in 
wetting the soil before drainage took place. Under these cir- 
cumstances it is clear that a part of the rain credited to evapo- 
ration was really retained by the dry soil, the true evaporation 
was consequently distinctly below the estimated quantity. 
The extremely low evaporation of the year 1879—80, and the 
rather low result for 1878-9, are chiefly owing to the abnormal 
evaporations calculated for the winters of these years, the cause 
of which will be presently considered. 
The amounts of evaporation for the summer half of the year 
also display considerable uniformity, notwithstanding great 
variations in the rainfall. During ten summers the smallest 
evaporation has been 10*710, and the largest 13*483 * inches. 
* This calculated figure is undoubtedly rather hipfher than the true evapora- 
tion. The summer six months succeeded a dry March and concluded with 
a wet September, nearly 1 inch of raui falling in the last three days of this 
month. The drainage is tlius from two causes somewhat below the amount 
naturally belonging to the rainfall, and the calculated evaporation is to the same 
extent too high. 
