The following artificial sea-water was found excellent for the growth of marine Protista and experiments in places where sea-water can not be obtained (the volumes being correct for : If is used, it should be dissolved in the before adding the other salts or it will be impossible to get into solution. We used a thick sirupy solution of water-glass (said to be and made an analysis of the Si concentration because we wished to ascertain the buffer value, but for the growth of organisms it is sufficient to consider the commercial solution as equivalent to 6 m SiO, or 1.5 m and the quantity desired should be obtained by successive dilutions, as a strong solution will precipitate when added to the other salts.
For plants at least it is permissible to add a drop of the strong solution per liter and allow it to precipitate. If cultures are grown in soft ease, no further additions of Si will be necessary to supply that used up by diatoms.
When sea-water is evaporated and redissolved oxides of manganese and iron, CaSO4, Si02, CaC0s, CaF2, Ca(B002, Al20s, BaSO4, and, under certain conditions, are diffi cult to get into solution, but by adding HC1 and evaporating, a paste containing gypsum will be obtained which will go into solution in about 3 days and may be neu tralized and 0.21 gram per liter added.
Another method of determin ing the tension on board ship or in laboratories without general chemical equipment was developed. The portable micro apparatus for determining the total described in section 4, was used and conversion tables were prepared for deter mining the tension from the total the temperature, and the excess base. Figures 7, 8, and 9 are conversion tables for finding the tension of sea-water of 23, 24, and 25 excess base, respectively. Each curve is for the indicated temperature only, and the values for 0° were extrapolated. Tables for sea-water of any other excess base titration may be plotted with the aid of figure 10, which shows the change of with change in excess base when other factors are constant.
Although it is theoretically a simple matter to determine the tension directly, in practice it has been most difficult and attended with large experimental errors. Perhaps the most direct method is to shake a large quantity of the sea-water with a small quantity of air in a bottle until equilibrium is established, and analyze the air in the bottle. Errors arise in failure to reach equilibrium or to correctly determine the pressure and temperature or to correctly analyze the air. The analysis of one part of in 10,000 parts of air may be done by the titration method, in which there are solubility and other errors, or by measuring the gas-volume contraction after absorption of in which there are temperature, pressure, and meniscus errors. A conversion table was made by Fox. For those who may be confused at first by logarithmic paper the relation of P. and CO, tension is shown on ordinary coordinate paper in figure 11.
The determination of of the air is made by placing about 10 c.c. of natural or artificial sea-water containing the thymolsulfophthalein in a nonsol test-tube of exactly the same bore as the sealed tubes and bubbling air through it to equilibrium and comparing with the tubes. The tension of the air is read from figure 6. In order to hasten equilibrium the test-tube should be tall and have a series of constric tions above the portion necessary for colorimetry. A rubber stopper with 3 holes is inserted and a glass tube leading to the bottom, its closed end being pierced by fine holes; a tube is connected to a suction pump and a mercury manometer inserted in the third hole. The of the air is the tension X 760 4- (the barometric read ing—manometer reading). Thymolsulfophthalein is decolorised by bubbling air through its solution for a long time, but this seems not to:be an oxidation or a reduction.