320 The Water Supply of London, [July, 



of ammonia. The amount ©f these substances present in rain which 

 falls at Eothampstead has been most carefully determined by a 

 laborious series of monthly analyses, made independently on the one 

 hand by Messrs. Lawes and Gilbert, and on the other by Professor 

 Way, and extending over two years. The results of these chemists 

 accord well, and they give as the average amount of nitrogen in the 

 forms of ammonia, nitrite of ammonia, and nitrate of ammonia, *0985 

 parts in 100,000 parts of rain-water. This must be deducted there- 

 fore from the calculated amount of previous sewage contamination 

 of any sample of water. It corresponds to 985 parts of previous 

 sewage contamination in 100,000 parts of the water. There is no 

 doubt that this reduction is too large, and therefore favourable to the 

 character of the water, since in most cases but a very small propor- 

 tion of the water of a river or spring falls as rain directly into the 

 stream; and Professor Way has proved that almost every trace of 

 the ammonia contained in rain-water is absorbed when the water 

 percolates through cultivated soils. Now, as three-fourths of the 

 combined nitrogen in rain-water is in the form of ammonia, it is 

 evident ftiat rain-water must be deprived of much of its original 

 nitrogenous contamination before it reaches such a river as the 

 Thames. The very small amount of combined nitrogen found in 

 natural waters of undoubted purity, such as that of Loch Katrine 

 for instance, also testifies to the liberality of the above allowance. 

 The water of Loch Katrine contains only one-third as much com- 

 bined nitrogen as that present in the rain falling at Eothampstead, 

 so that, starting from the base line of purity above proposed, the 

 water of Loch Katrine exhibits a negative previous sewage contami- 

 nation equal to 575 parts in 100,000 ; or, in other words, it would 

 require 575 parts of average London sewage to be added to, and 

 allowed to oxidize in each 100,000 parts of Loch Katrine water 

 before the purity of the latter would be reduced to the standard 

 with which I propose to compare the metropolitan waters. It is 

 necessary here to remark, however, that owing to the more copious 

 rains of the Highlands of Scotland, the rain-water of that district 

 probably contains less combined nitrogen than that which falls at 

 Eothampstead. The following table exhibits the amount of total 

 solid impurity contained in 100,000 parts of the metropolitan 

 waters in the months of February and April last ; it also shows the 

 amount of carbon contained in the organic matter forming part of 

 that solid impurity ; the amount of nitrogen in the salts of nitrous 

 and nitric acids ; the quantity of ammonia, and finally the amount 

 of previous sewage contamination. For the purpose of comparison 

 the results yielded by Loch Katrine water as supplied to Glasgow 

 in February last are appended. 



