2 26 CARNEGIE INSTITUTION OP WASHINGTON, 



he^t generated in the course of time. It is to be noted that the 

 heat generated after a soHd nucleus was developed must have been 

 superficial and hence readily radiated awa^-. While the nuclei 

 remained assemblages of small bodies, perhaps gaseous in part in the 

 larger ones, planetesimals from without may have penetrated to the 

 interior and there developed heat not so readily lost. But this state 

 is only assignable to the earl}- stages. 



A further consideration bearing upon the critical subject of tem- 

 perature is the manner of collision. Since all the planetesimals and 

 planetar}^ nuclei were revolving in the same directio7i about the solar 

 mass, the collisions were all overtakes, and could have been violent 

 only to the extent of their differences of orbital velocity, modified 

 by their mutual attractions. These velocities are of a much lower 

 order than the average velocities of meteoritic collisions. Many of 

 the overtakes would obviously be due to differences of velocity barely 

 sufficient to bring about an overtake. When the relative mildness of 

 impact is considered in connection with the intervals between impacts 

 at a given spot, the conviction can scarcely be avoided that the surface 

 temperature would not necessarily have been high. It seems probable 

 that it would have been moderate throughout most of the period of 

 aggregation, and certainly so in the declining stages of infall. 



The development of the hypothesis has now reached a point where 

 it can be tested. It happens to be a point where all hj-potheses of 

 this class have been supposed to be fatally at fault. The crucial 

 feature lies in the direction of rotation which would result from the 

 gathering in of matter in this way. At the same time, the bearing 

 of the discussion broadens, for this vital question of direction of 

 rotation attaches to all forms of aggregation of independent bodies 

 moving in orbits about the common center of a system. For example, 

 if the evolution of the solar system be supposed to start with a gaseous 

 spheroid of the Laplacian type, and to proceed in the manner postu- 

 lated by Laplace until the planetar}' rings were formed, and if then 

 the velocities of the molecules resulting from mutual impact carried 

 them beyond the gravitative control of the rings, so that they were 

 scattered and revolved independently around the central mass, the 

 hypothesis of their aggregation would be as much subject to the test 

 of rotation as the special hypothesis now under consideration. So, 

 too, if iiistead of forming definite rings, the molecules were separated 

 from the supposed gaseous spheroid, one by one, as seems more 

 probable than separation by rings, their aggregation would be 



