258 Mr. A. B. Northcote on the Water of 



the analyses in question I simply removed the bulb-tube and 

 evaporated its contents in a platinum dish, which was left over 

 oil of vitriol for the same time, and weighed under precisely the 

 same conditions before and after the evaporation. A litre of 

 April water yielded thus '0015 grm. of amnionic chloride, and 

 the same quantity of October water gave '0010 grm., correspond- 

 ing respectively to '000476 and '000317 grm. of ammonia 

 (NH 3 ); in July the amount obtained was inappreciable, and, 

 whilst the former residues gave good indications both with di- 

 platinic chloride and Nessler's test, Nessler's test alone gave a 

 faint reaction in this instance ■ and the same may be said also 

 of the water of February. The presence of nitrates was detected 

 in the usual way, by concentrating the water with addition of a 

 little disodic carbonate, and applying the ferrous-sulphate test, 

 the depth of colour which the different specimens yielded being 

 in accordance with the quantitative results which are given below, 

 and which were obtained by means of Puglr's process. The 

 water of February gave no reaction with the ferrous- sulphate 

 test, even when a litre was concentrated to a few cubic centi- 

 metres. This water also appeared to contain no trace of phos- 

 phates ; for neither the concentrated water nor the mixed ferric 

 and aluminic precipitate obtained from it yielded any reaction 

 with the diammonic-molybdate test; whilst, when similarly 

 treated, the ferric and aluminic precipitate from a litre of the 

 waters of July, April, and October gave indications, slight in the 

 case of the first, but quite distinct in those of the two latter 

 waters. The waters of July and October contained traces of 

 manganese, from which the others appeared to be perfectly free; 

 The determinations of the other constituents do not need re- 

 mark. After separation of silica and the precipitate by ammo- 

 nia, the calcium was estimated as carbonate by ignition of oxa- 

 late, and the magnesium as pyrophosphate ; the potassium was 

 separated from the mixed chlorides as chloroplatinate ; the chlo- 

 rine and sulphuric acid were weighed as salts of silver and ba- 

 rium. The atomic weights* employed in obtaining the numbers 

 which follow are those now in common use. 



I have placed the analyses made in 1866-67 in two Tables, 

 the first of which gives the constituents of the waters of ] 866, 

 whilst the second gives the comparative results obtained from 

 the two specimens of February 1867 ; of the latter, I have placed 

 the analysis of the water taken higher up the river first, by 

 which the influence of a further course of three miles, and of the 

 addition of Hawford Brook, is easily perceived. Although the 

 time, being one of flood, was perhaps not the best adapted for 

 ascertaining the change which the influx of Hawford Brook 

 * Odling's ' Practical Chemistry/ Tables, 1865. 



