On the Bain and Drainage - Waters at Rothanistcd. 255 
Thus on an average of three years, with a mean rainfall of 
28-465 inches, we have G'lO lbs. of nitrogen as ammonia supplied 
to the soil per acre each year ; and in addition, on an av erage 
of two years, 0 74 lb, of nitrogen in the form of nitric acid ; 
giving a total of G"84 lbs. of nitrogen. If, however, we only 
regard the two years in which the nitric acid was actually 
determined, the total nitrogen becomes 7*29 lbs. per acre, 
equivalent to 46^ lbs. of ordinary nitrate of sodium. It must 
be recollected in dealing with these figures, that the analyses on 
which they are based were made at a time when many of the 
modern refinements in chemical methods were unknown. We 
shall by-and-by compare these results with those obtained by 
other chemists. 
The amount of ammonia supplied to the soil by rain does not of 
course represent the whole amount furnished by the atmosphere ; 
we have also to take into account the direct absorption by the soil 
itself. The quantity of ammonia annually absorbed from the 
atmosphere by a moist soil is doubtless considerable, but in the 
present state of our knowledge no estimate of the amount can be 
made ; we shall, however, refer to the point again further on. 
The next analyses of Rothamsted rain-water we have to notice 
are those made by Dr. E. Frankland, and published in the Sixth 
Report of the Rivers' Pollution Commission, 1874, p. 27. This 
series includes 71 samples of rain- and snow-water collected in 
the first large gauge between April 1869 and May 1870, and 
7 samples of dew and hoar-frost collected within the same period. 
The examination to which these waters were submitted was far 
more complete than in any of the preceding analyses, the work 
in fact stands in some respects unique among investigations of 
the kind hitherto published. Dr. Frankland determined the 
total solid matter dissolved in the water, and the quantity of 
carbon and nitrogen existing in the form of organic matter, 
besides making determinations of the ammonia, nitric acid, and 
chlorine present ; the hardness of the water was also determined. 
The methods of analysis employed by Dr. Frankland are de- 
scribed in the Appendix to the Report just mentioned, and also 
in the ' Journal of the Chemical Society,' 1868, p. 77. 
The waters examined by Dr. Frankland were either fair 
samples of individual rainfalls, or represented the water collected 
during some part of a fall. The samples by no means represent 
all the rainfalls of a year, or even of any month, and the analyses 
are therefore insufficient for calculating the total amount of 
nitrogen or chlorine furnished by the rain in the course of a 
year ; but they illustrate excellently the variations in the com- 
position of rain-water under various conditions, which indeed 
was a special object of the inquiry. 
