In titrating the non-volatile buffers the same procedure minus the indicator is followed (though it is possible to cut all of the quantities in half if necessary). The entire quantity of acid is added at first and the sea-water boiled about 15 minutes gently, in order to remove all of the Any in the water will be estimated as non-volat buffer and influence the result. The equivalent of about 1 c.c. of 0.1 NaOH is added to the boiling sea-water; the volume is quickly ai accurately restored with boiling distilled water, and the flask tight closed with a rubber stopper. The NaOH solution must be free fro and contain little silicate. The method of its preparation ai protection is given on page 45. The exact quantity of NaOH to 1 added can not be stated, but there should not be sufficient to cau precipitation of earthy hydroxides. It is best to have two series tubes of known P., one colored with thymolsulfophthalein and t1 other with either o-cresolsulfophthalein or a-naphtholsulfophthalein in other words, a total range of P. = 7.5 to 9 or greater, but only tl first series of tubes are necessary. Two nonsol test-tubes of exact the same bore as the standard tubes and with etched marks at t] volume to be titrated are provided, each with a rubber stopper with central perforation closed with a glass rod. The required quantity indicator is placed in each of the tubes and sea-water from the flag poured in (to the mark) and the stopper quickly inserted. Care mu be taken never to breathe toward the test-tube unless it is tight closed. The tube with thymolsulfophthalein is first examined ai compared with the standard tubes and its P. recorded. The glass r( is removed from time to time and small quantities of 0.1 n HC1 a run in from the burette. The tube is shaken and its P„ is recorde together with the burette reading.
This process is repeated until the limit of the indicator is near reached, when enough indicator is added to correct for the increas( volume and any error due to dilution of the indicator is recorded. Tl same process is repeated with the other indicator and the results a compared with figure 4.
If the excess base is known the total may be determined the P. by using the conversion table in figure 5. The three dia oral lines are for sea-waters containing 23, 24, and 25 excess ba4 respectively. The smallest possible difference in per liter that v have been able to determine by means of the standard tubes is 0.1 c.c but in order to do this it is necessary to have a finer gradation of tub than are ordinarily used. Such a degree of accuracy is only relativ since the absolute accuracy of the conversion table probably does n( exceed 1 per cent. The diagonals in the conversion table are drawn! straight lines, but this is merely an approximation. If they wei extended they would have to be curved or lose in accuracy. The sea of too complex a composition to admit of any simple mathematic relations. Empirical formulEe, such as the one developed by Fox f( estimating the total may be discovered, but they are only approx mations. Exact data may be obtained only by direct experiment, an
the result depends on the technique.
If the sea-water contains more than normal non-volatile buffer, the content will be less than that read from the table, and vice versa, but the quantitative side of this relation has not been thoroughly investigated. We may picture it in the following manner: Suppose the water is sealed in an air-tight container so that can not escape. If we add boric acid part of it will remain free and increase the number of H ions; but since it is so weak an acid, this change will be slight. Part of the boric acid will displace from bicarbonates and some of the will dissociate and increase the number of H ions still more. Since boric acid is weaker than it will not decrease the P. as much as the same number of molecules of hence the error in estimating the will not be as great as the same molecular concentration of boric acid. would have even less effect than boric acid, but phosphoric acid would have a relatively greater effect. We know of no determinations of in sea-water, but Bang found 4 mg. per kilogram of fish.
The conversion table in figure 5 has a very limited range, and the further it is extended graphically the greater the error. We did this for the Princeton aquarium water because it had the very abnormal excess base of about 44.5. We determined the total and P„ on a portion of it and Dr. L. R. Cary determined the P, of another portion and the change in on further additions of The results are: The difference of 6.2 c.c. at P. 8.1 is only 0.0004 m, but the difference at 6.6 is probably partly due to an error in the extended table or the experimental data. It may be remarked that the history of this abnormal water does not completely solve the question of the origin of the difference in excess base. It is possible that some con crete may have been dissolved, although it was said to have been par affined. Concrete may possibly have been dissolved by the rain-water used to compensate for evaporation, or limestone dust may have gotten into the rain-water by being blown onto the roofs. These character istics of the water were not investigated when it was first received.
In using the conversion table in figure 5 for the study of the respira tion of marine organisms, only relative values are necessary, and it is thought wholly adequate for the purpose. It would be of little advan tage to extend the table to a greater P. range unless it is first absolutely established that the abnormal P„ does not make the organisms physio logically abnormal and that oxygen is still present in the water (see oxygen scale at left of figure 5 for temperate and tropical seas). The scale at the extreme left shows that the temperature is raised only 0.04° by animal respiration that uses up all of the oxygen in the warmer seas. This shows the impossibility of the existence of a warm-blooded animal with gill respiration.