TIDAL ZONES 219 



tion, it is 9,000 miles from the center of the earth. This feature adds its 

 own embarrassment to that of the Eoche effect in all attempts to Avork 

 out into specific terms a theory of the origin of the moon by the fission 

 of an earth-moon body. A still more formidable difficulty arises from the 

 determination, by Moulton, that the fission of an earth-moon body, if 

 assigned to progressive rotational changes (such as the passage from an 

 oblate to a prolate spheroid, thence to a pear-shaped body which by pro- 

 gressive constriction divides into two bodies), could only take place by 

 shrinkage to a density far beyond that now attained — indeed, far beyond 

 that rationally assignable under any known conditions. The fission 

 theory appears thus to be entirely untenable in the earth-moon case and 

 in similar cases.* 



DlVISIOJiT BETWEEN THE COLLISIONAL AND ULTRA ATMOSPHERES OP 



THE Earth 



The line C. A. marks roughly the zone of transition between the highly 

 collisional state of the earth's atmosphere near the surface and the vault- 

 ing state above, described long ago by Johnstone Stoney.^ This vaulting- 

 state, or krenal ultra-atmosphere, in turn passes outwardly into an orbital 

 state, or orbital ultra-atmosphere, as I have elsewhere shown. *^ These 

 transitions are, however, very gradual ; there is no definite line of separa- 

 tion. Collisions prevail in lessening frequency throughout the whole 

 sphere of control, but they are only dominant near the earth. The line 

 C. A. is placed at about 300 miles from the surface of the earth merely to 

 represent a horizon at which a large proportion of the free paths of the 

 molecules are great enough to encircle the earth, and so give rise to orbital 

 movements, in contrast to the short, straight, free paths of the molecules 

 in the atmosphere at the earth's surface. The diagram serves also to show 

 how enormously greater is the space occupied by the krenal and orbital 

 ultra-atmospheres than by the collisional atmosphere; but this greater 

 volume is more than offset by the extreme attenuation of the outer atmos- 

 pheres. In their extremely sparse way, the ultra-atmospheres occupy all 

 the outer space of the earth's sphere of control. Beyond that, and envel- 

 oping it, lies the sphere of control of the sun, occupied in like manner by 

 the sun's ultra-atmospheres, as previously noted. 



* The tidal and other problems. Pub. No. 107, Carnegie Institution of Washington, 

 D. C, 1909, p. 159. 



5 On atmospheres upon planets and satellites. Trans. Roy. Dublin Soc, 1898, p. 305. 

 «The origin of the earth, 1916, pp. 19-27. 



