On the Rain and Drainar/e- Waters at Rothamsted. 
Tablk XXIII. — Calculated Averagb Monthly Drainage from 
Soil 20, 40, and GO inches in depth, when Eainfall is of Normal 
Quantity. 
Rainfall. 
Average Average 
28 Years, lu Years, 
1853-«U. ' 1871-80. 
Ruin of 
10 Years 
under or 
above 
Uiiin of 
28 Years. 
Drainage, absuiuing normal Rain. 
Soil 
20 inches 
deep. 
Soil 
40 iiiclics 
deep. 
Soil 
60 inches 
ducp. 
! January .. 
[ Februaiy 
' Miuch 
April 
, May.. 
\ Juno 
July.. 
August .. 
September 
\ October .. 
I November 
December 
Whole Year . . 
Inches. 
2-590 
1-728 
1- 693 
2- 008 
2-329 
2-451 
2-704 
2-643 
2- 638 
3- 089 
2-345 
2-084 
Inches. 
802 
100 
505 
398 
224 
663 
280 
677 
123 
3-162 
3 094 
2-333 
Inches. 
1+0-212 
:+o-372 
'-0-098 
+0-390 
-0-105 
+ 0-212 
+ 0-576 
+0-034 
+0-485 
+0-073 
+0-749 
+0-249 
Inches. 
1-794 
1-029 
0-038 
0-420 
0-5:^7 
0-309 
0-314 
0-636 
0- 685 
1- 621 
1-409 
1-509 
Inches. 
2-082 
1-164 
0-773 
0-531 
0-606 
0-3-23 
0-342 
0-629 
0- 559 
1- 609 
1-492 
1-657 
Indies. 
1-818 
1-006 
0-683 
0-462 
0-538 
0-274 
0-228 
0-575 
0- 442 
1- 341 
1-250 
1-475 
28-302 31-451 +3-149 10-891 ' 11-767 
10-092 
We must now glance at the average amounts of evaporation 
during each month, which have been given in Table XXII. 
Looking first at the results obtained with the 20-inch drain- 
gauge we see that evaporation from the soil takes place to the 
smallest extent in December. In January it is slightly greater,, 
the smaller supply of radiant heat during the short days of 
December checking evaporation apparently more than the lower 
air temperature of January. In February evaporation remains 
little altered, but in March a decided rise commences, continuing- 
steadily up to July, when the maximum rate of evaporation is 
attained ; from this point a decline sets in till the minimum is 
once more reached in December. 
The amount of water evaporated per month is not always, how- 
ever, exactly shown by the mode of calculation we are forced to 
adopt. We have already frequently called attention to the 
error which must occur in our calculations when the soil is 
drier or wetter at the end of the experimental period than it 
was at the beginning. This source of error only occasionally 
affects the correctness of our calculations when the period in- 
question is a long one, or when the average of many similar 
periods can be obtained in which it is an equal chance whether 
the soil is drier or wetter at the end than at the beginning, but 
it certainly distinctly affects a portion of the monthly averages. 
The amounts of evaporation calculated for the mid-winter months 
will be practically free from errors of this description, if the 
average of a sufficient number of years be taken. From winter. 
