the River Severn at Worcester. 257 



tion to render the control by potassic permanganate of any 

 value. The colour of the residues on evaporation was a yellow- 

 ish brown, varying in strict accordance with the tint of the 

 water from which they were obtained ; they were dried at 140° 

 C, cooled over oil of vitriol, and weighed. Upon gentle igni- 

 tion they blackened, and after about two minutes burnt nearly 

 white ; they were then treated either with a saturated solution 

 of carbonic acid for twenty-four hours, redried at 140° C, and 

 weighed, or with a solution of diammonic carbonate for the same 

 time, and very gently ignited before weighing, — either treatment 

 being repeated until a constant weight was obtained. These 

 two methods of restoring the lost carbonic anhydride give closely 

 accordant results ; but the former is preferable on account of the 

 great difficulty of avoiding too strong an ignition. 



All the waters contained traces of ammonia, and, with the ex- 

 ception of that of February, nitrates and traces of phosphates. 

 The water of April was richest in these constituents, then that 

 of October, whilst after the drought of July they were found in 

 greatly diminished quantity, and after the floods of January and 

 February the water contained only a trace of ammonia. A re- 

 ference to a Table which follows, showing the rainfall about the 

 sources of the river, may serve to explain this, as the few days' 

 rain which will be seen to have occurred in April and October 

 would doubtless flush all the drains which conduct sewage to 

 the river, whilst in July they would probably have been for some 

 time dry, and in February well cleansed by the flood occasioned 

 by the long-continued thaw. Two of the waters were tested for 

 nitrites, but without success. The ammonia I found to be con- 

 veniently obtained by a modification of the distillatory process : 

 a litre of the water was placed in a flask of convenient size, which, 

 after the addition of baric hydrate, was closed by an india-rubber 

 cork pierced by two holes ; through one of these a straight tube 

 passed, the lower extremity of which was just closed by the liquid 

 within ; and in the other was inserted a Welter's tube with two 

 capacious bulbs, which terminated immediately below the cork. 

 In the latter tube four or five drops of pure chlorhydric acid 

 diluted with an equal quantity of distilled water were placed. 

 The liquid was then raised slowly to the boiling-point, and kept 

 gently boiling for a quarter of an hour. Care must of course 

 be taken to avoid rapid or long-continued ebullition, which 

 might cause ejectment of the contents of the bulbs, or render 

 them replete with condensed water. At the close of the opera- 

 tion the bulb-tube may either be removed immediately, or the 

 atmosphere of the flask first drawn through it, with the precau- 

 tion of adapting a wash-bottle containing chlorhydric acid to 

 the straight safety-tube to arrest any accidental ammonia. In 



