9» GEOLOGY. 



highest velocities by the molecules of the atmosphere in its outer part, 

 where the opportunities for escape were greatest. The velocities of the 

 plane tesimals may have been several miles per second in many cases. 

 In a certain proportion of cases, the planetesimals would strike the 

 molecules in such a way that the rebound of the latter would combine 

 a large proportion of the planetesimal velocity with their own previous 

 velocity, and would therefore increase their liability to escape from 

 the control of the young earth. Now in the impact of the molecules 

 of watery vapor with the planetesimals, particularly when the latter 

 had acquired a white-hot temperature, the heat was sufficient to disso- 

 ciate the oxygen and hydrogen in considerable proportions, and these 

 dissociated elements would take on the velocities appropriate to their 

 respective weights, and would act individually until reunited. The 

 mean velocity of hydrogen is more than four times that of oxygen, 

 and it may be assumed that it would be at least four times as liable 

 to escape from the control of the earth. Such excessive escape would 

 leave a surplus of oxygen, even after the hydrogen that did not escape 

 had reunited with oxygen, as it would in time. As the planetesimal 

 impact was, by hypothesis, a very pronounced phenomenon during the 

 growing stages of the earth, it is assumed that a notable amount of 

 free oxygen might be accumulated by this dissociative and selective 

 process, and that this, with that from other sources, would suffice 

 for the support of the earliest plant life, which, though feeding chiefly 

 on carbon dioxide, required oxygen. 



(2) As soon as plants were introduced, they may be assumed to have 

 decomposed carbon dioxide, retaining the carbon and liberating the 

 oxygen as at present. While a portion of this oxygen doubtless .re- 

 united with the vegetable matter on its decomposition later, the de- 

 composing process was probably never complete, and a residue of free 

 oxygen resulted. The oxygen of all the later periods is probably mainly 

 derived from the decomposition of carbon dioxide by plant action. 

 For the primitive atmosphere there is theoretical need for only 

 enough oxygen to support the primitive plant life until it could 

 supply itself, after which it would produce a surplus. 



In the preceding discussion, the feeding of the atmosphere by the 

 extrusion of the internal gases has been assumed. It is necessary 

 now to consider how extrusive action might be initiated, and how 

 maintained. 



