SOUTH AUSTRALIAN WATER SUPPLY. 635 



chemical analyses, and further point to the desirability of pui-ifica- 

 tion, and to the possibility of the elimination of all bacteria from 

 our waters as drawn from the service pipes. It is hardly necessary 

 for me to point ont that in many cases where bacteria do not pro- 

 duce any epidemic they may cause complications and render other 

 diseases more dangerous, or they may in other cases, by producing 

 small disorders of the system, pave the way for more serious 

 illnesses. 



I have to acknowledge the courtesy of Mr. MoncriefF, the 

 Engineer-in-Chief, in supplying me with reports and samples of 

 meter. 



Appendix. 



ON METHODS OF ANALYSIS USED. 



" Total solids" — Includes all residue left after drying to constant 

 weight in a large water bath at 100° C. 



" Chlorine " — Determined volume trically, using potassium chro- 

 mate as indicator. 



"Carbonic acid radicle, CO ^ " — By titrating a known volume 

 of water with iVth normal hydrochloric acid, using cochineal as 

 indicator and 0060 as factor for CO3 as bi-carbonates. 



" Ammonia free" — Distilling and nesslerising. 



"■Ammonia albumenoid" — Distilling with alkaline permaganate 

 and nesslerising. 



" Oxygen absorbed in fifteen minutes," " O.xygen absorbed in 

 four hours" — Tidy &. Fiankland's process. 



" Nitrogen as nitrates " — By copper zinc couple with addition 

 of a little sulphuric acid, pouring olf after twelve hours, precipita- 

 ting with caustic soda, allowing to settle, decanting and nesslerising, 

 allowance being made for free ammonia and for ammonia in 

 re-agents. 



"Comparative degree organic impurity" — Calculated from the 

 formula given by Mr. John Muter (" Analyst," June, 1883). 



" Opacity" — 1-0 indicates an opacity equal to that produced by 

 one part of kaolin in 100,000 parts of water. The standard 

 solution of kaolin is made by shaking powdered white kaolin with 

 distilled water, allowing to stand twenty-four hours, and decanting 

 off the unsettled portion, which is diluted to the required strength, 

 0"1' grain per litre being convenient for use. This and the color 

 Test are best made together on the same sample. 



"Color" — lO'OO indicates a color equal to 0-0001 ammonia 

 (NH3) in 100c. c. of water nesslerised ; the tests being made with 

 a solution of caramel in 50 per cent, alcohol, a known volume of 

 which is equal to 0-OOUl nesslerised ammonia in lOOc.c. of water. 



Although these tests for color and opacity are not of great 

 scientific accuracy, they are of easy and cheap application, and are 

 decidedly better than such terms as " slightly yellow," a " trace 

 opaque," &c. Strongly colored or opaque waters should be quanti- 

 tatively diluted before testing. 



