1867.] Proceedings of tlie Asiatic Society, 67 



small amount by boiling. It is superior in tbis respect to wbat can 

 usually be obtained for the supply of towns. The influence of the 

 tides during the hot season was considered in the first communication : 

 the tidal water increases the amount of common salt, but does not 

 very greatly increase the hardness. 



" As regards organic matter, numerous observations have been made 

 since the last communication was laid before the Society, partly to 

 meet objections raised against the former results, which objections 

 however may now be considered as withdrawn. 



" Further examination of the various waters by oxidation by 

 permanganate of potash has not increased the author's opinion of its 

 value, and two tables are given which it is believed will justify this 

 unfavourable opinion. The first exhibited the very rapid change 

 which takes place in the deoxidating power of many waters both 

 river and tank, this power diminishing within two days to one half, 

 one third, or even less, of its original amount. This is not noticed in 

 the London Reports, probably because the samples having been taken 

 from the street mains, the water is at least two or three days old, after 

 which it changes much more slowly. The oxidation test appears to 

 indicate only certain kinds of impurities, — probably products of 

 fermentation or putrefaction, or even of living vegetable organisms, and 

 it is doubtful if it gives much important information of the quality of 

 these, as the second table shews that General's Tank water (considered 

 the best for drinking in Calcutta) equals in deoxidating power the 

 the water of the salt marsh to the east of the town ; and that the water 

 of the Circular canal, which receives the greater part of the sewerage 

 of Calcutta, requires no more oxygen than that of the best tanks. 



" The determination of organic matter by weight is the most trust- 

 worthy. Care has been taken in all the recent analyses to proceed to 

 the evaporation without delay, but continued observation has also 

 shewn that the results formerly given cannot have been far wrong. 

 The quantity of organic matter in the river water for the months of 

 January and November has in no case exceeded 15 grains per 100,000 

 grains, or 1.05 grains per gallon. A table is given of the results at 

 all seasons, which distinctly shews the influence of the tides, the quan- 

 tity of organic matter during flood tide being from one and half to 

 two and half times greater than during ebb tide. Yet the highest 



