230 CARNEGIE INSTITUTION OF WASHINGTON. 



is made obvious. Thus the order in which the possibilities of col- 

 lision are brought into effect favors direct rotation. 



From the previous discussion it will be seen that a planetary 

 nucleus gathers planetesimals that have orbits both smaller and 

 larger than itself, and hence in effect it sweeps a space both outside 

 and inside its own zone. The breadth of this space is dependent 

 on the eccentricity of its own orbit and on the eccentricities' of the 

 orbits of the planetesimals it gathers in on either hand. 



It is obvious that there may have been two or more nuclei orig- 

 inally within the same zone. If one of these was notably smaller 

 than the other, it might be picked up by the latter the same as if it 

 were a planetesimal. Two of equal size might perhaps unite, though 

 this would not necessarily take place. Two nuclei in nearly the 

 same zone must feed upon the same belt of planetesimals and must 

 mutually interfere with one another's growth. If there were little 

 difference in their masses at the outset, that one which was best 

 spaced out from the nuclei in neighboring zones would be likely to 

 become dominant by superior growth, for it would have a better 

 feeding-ground, so to speak. Even a nucleus that was smaller at 

 the outset, if well separated from large competitors might become 

 the dominant one by a better growth. 



If there were originallj^ man}' nuclei of minor mass and if these 

 were much scattered, especially if the planes of their orbits were 

 diverse, the dominance of any one might be avoided and a scanty 

 growth of all result, as in the case of the asteroids. 



It seems to be a sure inference that in the process of growth the 

 nucleus must have ivorked toward the center of the zone from which 

 it gathered, as a consequence of the superior feeding on the richer 

 side. For example, if more planetesimals were picked up on orbits 

 smaller than its own, its orbit must have grown smaller as a me- 

 chanical result of the accretion, for a new orbit, arising from the 

 union of two bodies, is intermediate between the two previous orbits, 

 and hence smaller than the larger one. If more planetesimals were 

 picked up on the outer side, the orbit of the nucleus must have 

 grown larger. The nucleus, therefore, must have worked toward 

 the center of the richer feeding-ground, or in average cases of equable 

 original distribution, toward the ground not preyed upon by other 

 nuclei. 



The foregoing processes tended toward a selection of nuclei for 

 dominance and to an automatic spacing out of the successful nuclei. 

 This process, if our hypothesis be true, should find verification in the 



