HYPOTHESIS OF CAUSE OF GLACIAL PERIODS 573 
Reade’s primary estimate, but they probably represent approxi- 
mately the relative ratios, and show the general nature of the 
eliminations that are requisite to change accumulating land 
waters into sea waters. 
While these general statements of the nature and limitations 
of chemical action, based on the more familiar doctrines of the 
older chemistry, are doubtless essentially true, the refinements 
of modern chemistry teach that there is an intricate series of 
dissociations and exchanges of acidic and basic factors, and of 
the various ions, in an effort to establish and maintain a new 
equilibrium between the salts, required by their new proportions 
and their new states of dilution. As the land waters contain a 
relatively large percentage of bicarbonates of calcium and 
magnesium, the readjustment affects these especially, with the 
result that probably a minor percentage of the second equivalent 
of carbon dioxide is set free. It seems necessary to state this 
with qualification on account of the extreme complexity of the 
reactions, and the incompleteness of existing data; but the 
Challenger, and similar investigations show that the quantity of 
second equivalent of carbon dioxide is less than sufficient to 
raise all of the carbonates into bicarbonates. This deficiency is 
apparently limited to 20 per cent. or less of the theoretical 
amount required. More rigorous experimental determination is, 
however, greatly needed. 
It is probable that the second equivalent of the land waters 
is, deficient in some like degree, but this has not been experi- 
mentally determined. If this be true, it must reduce the estimate 
of the carbon dioxide brought down to sea, and also the amount 
set free by dissociation. The total amount of carbon dioxide 
which may be supposed to be set free by inorganic reaction in 
the sea in its present state of concentration, is therefore probably 
much less than 20 per cent. It is obvious from the preceding 
considerations, and others that will follow, that to maintain the 
atmospheric status even approximately there must be a nearly 
or quite complete return of the second equivalent of carbon 
dioxide by some means. This is also implied by the fact that 
