144 Proceedings of the Royal Irish Acaelemy. 



cooling the boiled water previous to its addition, and immersing tlie 

 apparatus in a bath, cooled by a mixture of ice and salt. The results- 

 in this case were variable, the introduction of 75 cub. cents, of -water 

 requiring from 0-9 to 2*4 of the hyposulphite. I found this to be due 

 to the length of time the apparatus was allowed to stand in the bath, 

 after the introduction of the water. The longer the time, the less the 

 quantity of the re-agent required to destroy the blue tint produced. 

 To make certain that such a change did occur, I brought the liquid to 

 the yellow tint, and then added a few drops of indigo, so as to produce 

 a distinct green colour. On allowing the apparatus to stand in the 

 bath, this gradually disappeared. I tried this several times, adding 

 vaiiable quantities of indigo in excess ; but in all cases (within cer- 

 tain limits), when allowed to stand in the cold water, the green colour 

 was gradually replaced by the yellow tint, just as if a quantity of 

 hyposulphite had been added. 



I attempted to remove these errors by a modification in the method 

 of procedure. 



The burette used for the indigo was replaced by a lai-ger one (100 

 cub. cents, capacity). After placing in the apparatus a quantity of 

 indigo solution, expelling the air, and bringing to the neutral point 

 as usual, a known excess of the hyposulphite was added. Saturated 

 water, as before described, was then added from the large burette, so 

 as to oxidize the excess of hyposulphite, and just tinge the liquid 

 green. From the volume required, by using the formula before given, 

 the volume of oxygen contained in tliis can be found ; hence the equi- 

 valence of the excess of hyposulphite in terms of oxygen. The same 

 operation is performed with the water under experiment acting upon 

 the same excess of hyposulphite. The relation between the volume 

 used and that of the saturated water gives the amount of oxygen in 

 the liquid. Any errors resulting from change in the solution would 

 thus be eliminated, and the calculations simplified. On trying this 

 method, I did not find the results any more satisfactory than the origi- 

 nal method, although performed with the greatest care : they gave 

 quantities in excess of that given by the Gasometric method. Thus,, 

 in 2"420 litres of water, 



Cub. Cents. 

 Yolumetric modification gave, . . . . 15-133 

 Gasometric method 12-650 



Excess, 2-483 



Although tliis Yolumetric method possesses the great merit of exceed- 

 ing rapidity, yet the many precautions necessary to be taken greatly 

 detracts from the value and reliability of the results. It seems to be 

 better suited for the detennination of oxygen in small rather than in 

 large quantities of a liquid, si;ch as are desirable in water estimations. 

 It is also valuable as a means of showing whether it would be desii'a- 

 ble to estimate the amoiint of oxygen present in a water, by the more 



