in the Twenty Years before 1895 57 



the carbonic acid of the neutral carbonates being driven out 

 by the concentration of the chloride of magnesium, which 

 was obvious; but he thought that what he had obtained 

 was mainly the free and half-bound carbonic acid. He found 

 that carbonic acid came off during the whole of the distillation, 

 and his view was that it was loosely bound to the chloride 

 of magnesium. In the autumn of 1872, I made some experi- 

 ments on the absorption of carbonic acid by saline solutions. 

 These were not very conclusive, but they seemed to point 

 to the sulphates as being capable of retaining carbonic acid. 

 Acting on this indication, I thought it could do no harm to 

 remove the sulphates before boiling out the carbonic acid, 

 and an essential feature in the method which I used was the 

 addition to the volume (225 c.c.) of sea-water of 10 c.c. of 

 saturated chloride of barium solution, which was more than 

 sufficient to precipitate all the sulphates. The effect of this 

 addition was to accelerate greatly the evolution of the carbonic 

 acid, and to produce very smooth and regular boiling. 



When Jacobsen's "Pomerania" results were published in 

 1874, and were received on board the "Challenger," I compared 

 my results up to date with those which he had obtained in 

 northern waters, and I was concerned to note that the carbonic 

 acid which he found was almost exactly double what I found. 

 It seemed unlikely that this could be due to locality, and I 

 very thoroughly examined my method, and found nothing 

 to find fault with. The well-known risk in such an operation 

 was that, from decomposition of neutral carbonates by the 

 chloride of magnesium, the carbonic acid obtained would 

 be too high. As mine were all much lower than Jacobs, 

 they were probably freer from that source of error than hi- 

 lt must be remembered that what both Jacobs< n .m<l I wished 

 to determine was the carbonic acil whi< h i^ not bound to the 

 base as neutral carbonate. It was only the free or I 

 bound carbonic acid that could be held to form part of the 



>sphere offered to the inh.ibit.mts of the water, win 

 was our first business to analyze. As rh mivts. both of us 

 -were aware that the total carbonic acid could be easily and 

 accurately determined by aci<lit\in- tin- \\.IUT and ( 



