SOUTH AUSTRALIAN WATER SUPPLY, 629 



in August. This may probably be ascribed to the drainage from 

 the manured land. The best example of quick natural jjurification 

 that has come under my notice, however, is given by the sewage 

 at Islington. Of the two analyses of this water in Table 1 one is 

 of the influent sewage after passing the strainer, the other of the 

 water flowing away in the creek below the farm ; but for the 

 salinity and high nitrates of the latter it would be difficult to 

 recognise it as a sewage water. The figures given for the Isling- 

 ton waters can only be taken as very approximate, as a long series 

 of analyses wordd be required to obtain the mean composition of 

 the influent and effluent, but the above are quite sufficient to show 

 that the land is most suitable in this respect for a sewage farm. 

 It is most desirable that frequent analyses be made of the intake 

 and effluent of the sewage waters, especially to learn whether the 

 salinity of the land is being increased by the continued irrigation, 

 the water running from the farm being at least five times as saline 

 as the water of the city supply. 



In Table 2 monthly analyses of the Onkaparinga water from 

 samples taken at Clarendon are given to illustrate the variation of 

 organic purity and salinity with the seasons. A. similar variation 

 occurs with most of our waters. The waters are generally of 

 greatest organic purity just before the heavy rains of winter set in; 

 they are then also most saline and hard. Immediately after the first 

 heaA^y rains they are organically most impure and least saline, and 

 it is owing to this last quality and the general scarcity of a suitable 

 supply that the purification of turbid water would be of such 

 innuense advantage here. The Beetaloo flood water is, for instance, 

 of about one-fourth the salinity of the ordinary flow, and the 

 Oukajiaringa the same, and such waters as the Bundaleer Creek, 

 and even the Rocky River, at Huddleston, might not be too saline 

 for use if collected during a heavy flood. After a lieavy rain this 

 year the Beetaloo reservoir was filled with water containing only 

 ten parts of saline matter per 100,000, the average quantity for the 

 three jDrevious years being from fifty or sixty parts. 



On looking over a long series of analyses of the same water 

 collected from reservoir and service pipe, it becomes evident that 

 the latter is decidedly organically purer than the former, the 

 passage through the pipes appearing to have a purifying action. 

 This may be due in part to a subsidence of some of the organic 

 matter in the mains, this being afterwards removed by flushing ; 

 but, as it occurs with the clearest waters as well as with the most 

 turbid, it is probably partly due to the iron reacting on the aerated 

 Avater ; examples of this greater purity of the service water are 

 given in Table 5, each analysis given being the average of twenty- 

 four analyses, extending over two years. 



On examining tabulated analyses of our principal potable waters 

 it appears that there are only three of them which can be almost 

 continuously classed as organically pure ; these are the Blue Lake 

 at Mount Gambler, the Port Augusta, and the Gawler supplies. 



