612 CYRUS F. TOLMAN 
cludes that the warmer portions of the ocean are constantly 
supplying CO, to the atmosphere, and the colder portions are 
always removing it.’ 
The falling temperature which causes the glacial period must 
affect the temperature of the surface of the equatorial water, 
especially after an ice advance is inaugurated, but how much we 
cannot estimate. Taking the average surface temperature of 
these waters as 21°, and postulating the same fall of tempera- 
ture as for the temperate waters, we see that the line a’ 0’ rep- 
senting the decreasing dissociation due to the falling tempera- 
ture, is much shorter than the line c’d’ or c’ e’, representing 
the increase in the dissociation caused by the decreasing partial 
pressure of CO,. However, the data upon which this part of 
the curve is based are much more conflicting and less satisfactory 
than those represented by the curve at 10°. But we may con- 
clude that in the equatorial region the diminishing partial pres- 
sure has a greater effect than the postulated fall of temperature. 
But, with the increasing cold, the heated areas which can 
give off carbonic acid to the air are greatly diminished, espe- 
cially after the glaciation has commenced, and this acts directly 
counter to the diminishing partial pressure. 
‘In the northern seas whose surface temperature is near zero, 
each molecule of carbonate has the power to take up carbonic 
acid until it becomes a fully saturated bicarbonate. Just how 
much increase this area will undergo on an approach of a glacial 
period we cannot well estimate, but perhaps we can get a fair 
idea of it by comparing the area of the ocean contained within 
the circle of latitude which bounds the northernmost point of 
Greenland glaciation with that contained within the circle bound- 
ing the southern limit of Pleistocene glaciation. The ratio 
- between these two areas is 2:5. These figures may not even 
approximately represent the real increase of the cold waters 
during glacial times, but they emphasize the fact. of an actual 
and great increase in the amount of these waters. 
The table on p. 609, as stated before, shows that at low tem- 
tChal. Rep., Vol. I, pp. 212, 213. 
