50 Papers from the Department of Marine Biology. 



An experiment designed to illustrate this was made as follows: 

 5 c.c. of sea-water were placed in the hydrogen electrode and a stream 

 of pure hydrogen of about 100 c.c. per minute passed through it. 

 About 15 minutes were required for saturation of the water with 

 hydrogen (so that the P H was the same, 8.2, as with stagnant hydrogen, 

 but with the electrode raised out of the water), and this time and 10 

 minutes more were counted out of the experiment. The water now 

 contained 0.2 c.c. of CO 2 in the form of carbonates and bicarbonates, 

 calculated from the P H of 8.3. In 10 minutes, with the passage of 

 1,000 c.c. of H2 in small bubbles, the P H rose to 8.4, indicating that 

 0.007 c.c. of CC>2 had been removed. The rate of removal constantly 

 decreased in an asymptotic curve, and at the end of 2 hours a P H of 9 

 was reached, indicating that a total of 0.05 c.c. of C0 2 had been 

 removed. During the next 10 minutes the P H rose to 9.02, indicating 

 the further loss of 0.0014 c.c. of CO 2 . At the end of 270 minutes the 

 P H had reached 9.18 and was rising at the rate of 0.01 in 10 minutes, 

 indicating a total loss of 0.0616 c.c. and a final rate of 0.0007 c.c. per 

 10 minutes. 



There was still enough C(>2 in the sea-water to convert most of 

 the excess base into normal carbonate. This shows the hopelessness 

 of reaching equilibrium at very low COz tensions, and we seldom 

 attempted anything lower than 0.0002 atmosphere, and then made 

 tests to see that equilibrium was attained within the limit of error of 

 measurement. The buffer mixtures were intended to be C0 2 -free, and 

 similar experiments on them showed that if traces of CC>2 were in them 

 they were not removed by hours of bubbling hydrogen through them. 



We know of no attempts to determine the asymptote or limit that is 

 approached in the above experiment. If it is an NaOH solution that is 

 approached, we should be able to remove any traces of CC>2 from the 

 buffer mixture (theoretically with ideal apparatus). The weak acid 

 in the buffer should help displace CC>2. Perhaps the reason we could 

 not reach a higher P H by bubbling H2 through the buffer mixture was 

 that the traces of C02 were not sufficient to measurably change the 

 P H . The electrodes for comparing the colors of the buffer mixtures 

 and sea-water of the same P H were made of the same bore as the sealed 

 tubes and so that they could be placed in the colorimeter after determin- 

 ing the P H electrometrically. The essential form of these is shown in 

 figure 13. There are no ground joints or rubber connections exposed 

 to the air. The hydrogen is allowed to escape through the trap at the 

 top, which is filled with some of the same solution as in the electrode. 

 Electrolytic connection with the calomel electrode is made through the 

 ground joint at the bottom submerged in an intermediate vessel of 

 saturated KC1 solution. 



Some of the electrometric titrations were made with the dipping 

 electrode shown in figure 14, which is rinsed with the solution by 



