314 Dr Smith on the Composition of the 



exactly to that indicated by the soap-test namely, 0*94 

 grain ; and if the magnesia, in like manner, be calculated 

 as carbonate of lime, it will turn out 0*26 grain, differing 

 very little from the soap-test result ; and in the case of 

 magnesia, it is to be expected that the ordinary analysis 

 will give under the truth. The quantity of pyrophosphate of 

 magnesia, afforded by a gallon of water (A), I have noted 

 at 0-30 grain. If this be made 0*32, it will correspond ex- 

 actly with the indication of the soap-test. This new applica- 

 tion of the soap-test promises to be of considerable utility.* 



IV.' — Composition of Don Water. 



Two specimens were examined. One (A) was taken on 

 the fourth of October, from the river about a mile from its 

 embouchure, quite beyond the reach of the tide. The other 

 (B) was collected near the Bridge of Alford (about 35 miles 

 up, counting the windings of the river), on the same day, 

 and about nine hours previous to the collection of specimen 

 A. The river was in its average condition. Specimen A 

 was tinged brown with vegetable matter, which also affected 

 the taste. Flocculi of organic matter were seen in it, and 

 brown specks of iron and organic matter, speedily subsiding 

 on rest. No air-bubbles emitted on exposure. By careful 

 evaporation, the solid matter was found to be in the pro- 

 portion of 8-64 grains per gallon, and of this there were 3*04 

 grains volatilized by ignition. The composition of the dry 

 matter was as follows : 



* The Culter Burn, a large tributary of the Dee, which drains the Loch of 

 Skene and surrounding country, I found to contain at its confluence with the 

 Dee (about 8 miles up), on the 26th of September, 6 grains of fixed salts per 

 gallon, with 2 or 3 grains of organic matter additional. Hardness 3°-l. 



