CARBON DIOXIDE OF THE OCEAN 603 
magnesium bicarbonates, partially dissociate at ordinary temper- 
atures into the normal carbonate, giving out CO,. Treadwell 
and Reuter have shown in the latest and most complete research 
on the subject (see reference above) that when calcium carbon- 
ate is dissolved under different partial pressures of CO,, that 
the proportion of calcium to carbon dioxide, indicates that prac- 
tically a pure bicarbonate, and not a mixture of bicarbonate and 
carbonate, is present in the solution. 
The difference between the sodium, magnesium, etc., bicar- 
bonates on the one hand, and the calcium bicarbonate on the 
other is seen from the following: (1) For the dissociation of 
magnesium bicarbonate we have in a saturated solution, 
Mg (HCO,;), == Mg CO, + H,CO,. 
We have seen that the H,CO, is directly in equilibrium with the 
CO, in the air; therefore at a given temperature, the degree of 
dissociation depends upon the partial pressure of the CO, in air. 
With variations of temperature, the colder the water, the larger 
is the portion of bicarbonate present, and therefore, the larger 
is the proportion of carbon dioxide in the ocean to that in the 
air. (2) For calcium bicarbonate we have : 
€ai(HiCO,), === CaCO, + H,CO,. 
But Ca CO, is very slightly soluble, and therefore practically a 
constant for a saturated solution. Therefore, we do not have a 
mixture of bicarbonate and appreciable quantities of the carbon- 
ate, but the salt in solution is nearly all bicarbonate, and by 
lessening the partial pressure of the CO,, Ca CO, is precipitated 
from the saturated solution and the bicarbonate decreases. 
However, this difference between Ca (THEO) , and Mg 
(HCO,), which Treadwell has emphasized, only holds for 
saturated solutions. It is evident that for an under-saturated 
solution of Ca (HCO,), we may have a larger proportion of 
normal carbonate than in the saturated solution and still no 
precipitation. Also, in ocean water (see p. 591), the numerous 
different salts tend to increase the solubility of the normal 
