I9I2-] OF THE GLOBULAR CLUSTERS. 1(33 



of the theory of repulsive forces, however, it is now possible to 

 explain these projectile motions, which Herschel likewise pointed to 

 as the chief agency for the preservation of sidereal systems. The 

 only assumption necessary is an unsymmetrical figure of the primor- 

 dial nebula, giving a whirling motion about the center as the system 

 develops ; and since the dust gathers from all directions it is certain 

 that this lack of perfect symmetry will nearly always develop, as we 

 see also by the spiral nebulae. 



10. It is this unsymmetrical form of the spiral nebulae produced 

 by the gathering of the dust from the stars, or the slight relative 

 tangential motion of stars formed separately but finally made to 

 revolve together as a binary system, that gives the binary stars the 

 projectile forces, with which they are set revolving in their orbits. 

 In no case have they resulted from the rupture of a rotating mass of 

 fluid under conditions of hydrostatic pressure as formerly believed 

 by Darwin, Poincare and See. 



11. Even if the rotation could become rapid enough to produce 

 a separation, under conditions of hydrostatic pressure, by rupture 

 of a figure of equilibrium, there would still be the equal or greater 

 difficulty of explaining the origin of the primitive rapid rotation. 

 This last difficulty escaped notice till we came to assign the cause 

 of rotations, and found that mechanical throwing ofif was impossible 

 under actual conditions in nature. It is therefore recognized, from 

 the definite proof furnished by Babinet's criterion in the solar sys- 

 tem, that such a thing as a throwing off never takes place; but that 

 all planetary and stellar bodies are formed in the distance, and after- 

 wards near the centers about which they subsequently revolve. 



12. TJiis gives us a fundamental laiv of the firmament — tht^ 

 planets being added on to the sun, the satellites added on to their 

 planets, the moon added on to the earth, and the companions added 

 on to the double and multiple stars — which is nozv found to be beau- 

 tifully confirmed by the dynamical theory of the globular clusters. 

 It is not often that such a great lazv of nature can be brought to light, 

 ami it is zvorthy of the more consideration from the circumstance 

 that it explains all classes of stellar systems by a single general 

 principle. 



