BY
FOLKE YOROLEFF AND ER\ST-GUSTAF FINNILA
Salinity. The i1 o h r-I n u cl s e n method was used. In the laboratory on board the salinity was titrated with a whole-automatic burette, model S t ii t z e r b a c h.
The scale on the burette, ranging from 0 to 10 ml, is subdivided in 0. o 4 ml.
Dissolved Oxygen. The samples were drawn off into glass-stoppered bottles with a capacity of about 150 ml. For storage, the filled bottles were placed in glass cylinders which were filled to overflowing with sea water shortly after the samples were drawn.
Oxygen was determined in 100 ml according to «' i n k I e
is
method, as modified by KALLE (1).Alkalinity. The determinations were made according to the indirect titration method by GRIPENBERG (2). The carbon dioxide set free by the acid was expelled by letting a. strong current of carbon dioxide free air pass through the solution for half an hour.
Hydrogen Ion Concentration (p1-I). For the measurements a B e c k m a n pH-meter model GS with electrodes nos. 1190-80 (glass) and 1170 (calomel) was used.
We used the lower sensitivity, which is the same as for model G or = 0. o a pH.
As is well known, there are two essential conditions which must be fulfilled for a successful determination of pH in situ. Firstly, the vessel containing the water sample during the measurement should be of such construction and should be filled in such a way that CO2 cannot penetrate from the atmosphere into the sample, nor be given off therefrom. Secondly, the difference in temperature in situ and at the moment of determination should be taken into account. The first of these requirements necessitates a special electrode vessel, and we have used a 75 ml weighing bottle (diam.
10 2. ANALYTICAL -METHODS
4.7 cm.) with a. rubber stopper.. The vessel is filled through a glass tube I.< • 9 extending to the bottom of the vessel, r
¢ II and water escapes through another
glass tube inserted level with the lower surface of the stopper. Several vessel volumes are allowed to pass through.
1 As to the temperature compensa.- tion, the B e c k man pH meter is fitted with the temperature
compa-rator, allowing measurements at tempe-ratures between 10° and 40°C only.
Bucx and NZNAs (3) have tested the accuracy of the compensation ,.' ., and found that the temperature was C r: somewhat overcompensated; they also found that the electrodes require a rat- 6. Electrode vessel her long time for adjusting themselves to temperature changes. For these reasons we did not use the compen-sator at all; instead, we standarized the pH meter at 20°C and calculated the temperature correction according to NYN,is (4).
If the temperature of the sample was below 15°C, the vessel was placed in a thermostat oven, adjusted at 50°C, for one minute. The electrodes and a thermo-meter were fitted into a rubber stopper similar to that of the vessel. Besides, the stopper had a small vent to prevent any pressure difference when the vessel was fitted to the stopper. The pH was measured after waiting one minute and the tem-perature read also.
pH in situ = pH, + (t1-20) (-0.0034 pH1 +0.0041) + a (t,—t).
t1 = temperature of the sample during the measurement.
t = temperature in situ.
a = function of Cl, pH-I and t according to Buda and NZ,`*äs (loc. cit.) The factor 0.äö11 in the formula above is determined according to NYNAs (loc. cit.), and differs from one instrument to another.
Dissolved Phosphate. The determinations of dissolved inorganic phosphate were made by the molybdenium-blue method, as modified by -WOOSTER and RAXESTRAW (5) for use with a photoelectric colorimeter: 50 ml of the sample was treated with reagents. In the interval of 10 to 30 minutes after addition of the stannous chloride, the extinction of the sample at 700 my was measured, using a B e c k m a n model B spectrophotometer and 7.2-cm absorption cells. The standard phosphate solution was made of reagent grade KH,PO1. The analyses were made on board, within two hours after the samples were drawn from the water bottle. Samples showing a visible turbidity were filtered through a Zh h a t m a n filter paper no 42. No salt correction has been applied. The accuracy of the analyses is estimated to be at about +5 per cent.
Silicate, The determinations of silicate were made according to the method by D i e n e r t and tid' a n d e n b u 1 c k e, slightly modified by RoBINsorr and Taorirsou (6). Again 50 ml of the sample was treated with reagents. The ana-lyses were made on board, and sulfuric acid was added to the samples immediately
2. ANALYTICAL METHODS 11
after they were drawn from the water bottle. A sample treated in this way is stable for 12 hours, but the analysis was in most cases carried out within two hours. In the interval of 10 to 60 minutes after the addition of ammonium molybdate, the extinction of the sample at 435
was measured in the spectrophotometer, using7.2-cm absorption cells. A buffered solution of K 2CrO,1 was used as standard. Samples showing a visible turbidity were filtered in the same way as the phosphate samples.
The relative errors for the analyses do not exced + 5 %.
Calcium and Magnesium. Wide use of ethylendiamintetraacetic acid or its disodium salt (Complexone III, Versenate or EDTA) as a volumetric agent has been possible, on account of the broad chelating power of these substances and the stability of the chelates. Complexone III forms a 1 to 1 chelate with a large number of cations. As a base for the determination of Ca and Mg in sea water, the methods described by SCHWARTZFNBACH (7) were used.
Firstly, the sum of the metals was determined as follows: 25 nil of the sample was buffered to pH 10, with ammonia-ammonium chloride buffer. The mixture was warmed to 40° C and titrated at this temperature with a 0. o sM standard solution of Complexone III, using Eriochrome Black T as indicator.
Secondly, for the determination of calcium, an indirect titration method was used as the magnesium content of sea water in equivalents is about five times that of calcium. The procedure was a follows: 10 to 20 ml of the standard Complexone III solution were added to 50 ml of sea water. The amount of standard solution depends on the salinity of the sample. The pH of the solution was raised to 12 by adding 10 ml of 1N sodium hydroxide, and the excess of Complexone III titrated with a 0. o aM calcium standard, using Murexide as indicator.
This method gives the sum of calcium and strontium. All calcium values given include the strontium present.
References KALLE, K. Ann. d. Hydr. 67, 267 (1939).
GRIPENBERG,