62 THE CLIMATES OF THE GEOLOGICAL PAST 



Enormous ev^aporation in equatorial regions led to 

 salinity obtaining a greater influence than temperature 

 over the specific gravity of sea water, hence the salinity 

 controlled the manner of circulation. Heavy salt waters 

 sank and drifted polewards from the equator, and lighter, 

 fresher and cold waters formed surface currents from the 

 poles to the equator. 



Submarine eruptions, by pouring hot lavas into the 

 sea, would cause increased evaporation and increased 

 salinity, and would saturate deep-sea waters with carbonic 

 acid, liberated partly from the lava itself, and partly from 

 the decomposition of corals and coral limestone by the hot 

 lava. The carbonic acid carried poleward in solution 

 under pressure would be liberated as the waters rose to 

 the surface. Carbonic acid thus liberated would be of 

 great value to plant life, both because it is a plant food, 

 and because it renders the climate warm and equable. 



It is therefore clear that both the facts of an ice age 

 and the phenomenon of a rich polar floi^a can be explained 

 on very simple assumptions, and that there is no need to 

 suppose those vast astronomical and physical revolutions 

 which many theories have hypothesled. 



M. Eugene Dubois, in a celebrated essay, tried to 

 explain climatic variations on the earth, and also ice ages, 

 on the assumption that the sun is a variable star. The 

 objection to this view is that variability of the sun implies 

 a harmonious periodicity, and indeed, we possess no 

 evidence to show that intervals between warm and cold 

 periods of geological history were of equal duration. In 

 fact, most of our data tend to disprove any regular periodi- 

 city. On Dubois's assumption, glaciation should also 

 have been universal in the cold cycles, and simultaneous 

 in the two hc-mispheres. This was not the case. Besides, 

 M. Dubois's theory is unnecessary, far-fetched, and un- 

 proved. 



There may be some variability in the amount of heat 

 obtained from the sun in dirEerent epochs, but as the 

 evidence which we possess, proves it to be of an irregular 

 nature, it must be due to extra-solar or accidental causes. 

 Effective causes of such fluctuation in solar heat might be 

 either (1) the passing of our sun and planetary system 

 through an excessively cold cr hot region in space ; or (2) 



