sea-fisheries laboratory. 141 



7. — Estimation of the Carbonic Acid Gas in Solution. 



IOOcc. of the sea water was mixed with a slight excess of 

 barium chloride solution and warmed till the precipitate 

 had settled. This was then filtered and washed, and 

 repeatedly treated with known amounts of standard 

 hydrochloric acid. As the precipitate would consist of 

 barium carbonate and barium sulphate, it was hoped that 

 the hydrochloric acid would dissolve the former. The acid 

 solution was then boiled to drive off the carbonic acid, and the 

 excess of acid estimated by titration with standard alkali, 

 using aurine as indicator. The alkalinity, determined by 

 methods already described, was subtracted from the value thus 

 obtained, when the difference was supposed to represent 

 the carbonic acid directly dissolved in the water. 



This ought to be an accurate method, but probably it is 

 not, for so many processes have to be gone through that 

 the accumulated iron due to each probably amounts to 

 some considerable sum. For this reason it was only 

 attempted in one or two cases, and as the results seem open 

 to suspicion, they are not given. 



8. — Estimation of Lime. 



Owing to lack of time this was only done in a very few 

 cases. IOOcc. of the sea water were mixed with ammonium 

 chloride solution and an excess of ammonium oxalate, and 

 the mixture was warmed till the precipitate settled. It was 

 then filtered off, washed, dissolved in dilute sulphuric acid, 

 and the oxalic acid estimated by titration in the heat with 

 standard potassium permanganate. 



The permanganate solution was standardised against 

 weighed amounts of ferrous-ammonium sulphate. It was 

 then used at once. 

 m 



