Lerrs anp Buaxe—Estimation of Oxygen in Solution. 459 
owing to absorption of atmospheric oxygen by the upper layer of 
the water as it flows into the wide part of the vessel is avoided. 
The stopper is now inserted, and the excess of water drained 
off. The stopper is next removed and about 7 c.c. of the water 
taken out by a pipette and thrown away. 5c.c.' of the (clear) 
ferrous sulphate solution are then carefully measured off in an 
accurately graduated pipette and allowed to flow into the separating 
vessel, the nozzle of the pipette being just dipped below the surface 
of the water, where the ferrous sulphate owing to its higher density 
flows in a stream to the bottom of the water. Strong ammonia 
solution is then cautiously poured on to the top layer of the 
water until the vessel is just full, when the stopper is inserted ; 
but even if a drop or two of the ammonia is lost when inserting 
the stopper no harm is done, as the ammonia floats on the surface 
and does not appreciably mix with the contents of the vessel. A 
little dexterity is necessary in inserting the stopper so that no air 
bubble is enclosed. Within the separating vessel there is now a 
layer of ferrous sulphate below, next the water, and above all the 
ammonia. ‘These are mixed together by inverting the vessel once 
or twice by a swinging motion, when a greenish turbid mixture 
results, containing ferrous hydrate partly in solution and partly in 
suspension, which rapidly darkens as it absorbs the dissolved 
oxygen. In all the authors’ experiments an interval of fifteen 
minutes was allowed for the purpose. Experience has shown it to 
be sufficient, but it is possible that a shorter period would suffice. 
The next operation consists in inverting the vessel, still stoppered, 
and then filling or nearly filling its tube or lower extremity (now, 
however, its upper extremity) with the mixture of sulphuric acid 
and water (fig. 2). The stop-cock is then opened, when the acid 
flows downwards into the alkaline mixture, and in the course of a 
few minutes dissolves the iron hydrates to a clear solution.’ 
Should an air bubble remain in the boring of the stop-cock, it 
prevents this flow, but a few taps causes it to be displaced and to 
_ rise through the acid. 
1 Of course these quantities would have to be be varied if a vessel were used either 
much smaller or much larger than the one described. 
2 In some of the test experiments with distilled water saturated with air this did 
not occur until the mixture was heated. But with all samples of natural waters, sew- . 
age effluents, &c., no heating was necessary. 
