238 ANNUAL, REPORT SMITHSONIAN INSTITUTION, 1931 



same amount of matter in the universe as now, but it would be packed 

 into a smaller space; the density of the primeval gas would be 

 greater than we have calculated for it. Consequently, the distance 

 apart of the condensations which ultimately formed nebulae would 

 be less than we have calculated. After they had formed, their 

 rushing apart would increase their distances, with the result that 

 by now these distances would be nearly, but not quite, as far apart 

 as those given by a calculation which ignores the expansion of the 

 universe entirely. 



The upshot of the whole matter is that, whether the universe is 

 expanding or not, the actual condensations of a primeval gas ought 

 to represent the present nebulae fairly well. 



If this account of the origin of the nebulae is accepted, it becomes 

 possible to trace out the mechanical evolution of the universe from its 

 origin as a uniform gas spread throughout primeval space. We 

 have in succession : 



1. A uniform tenous gas of density of the order of 10"^° and 

 of diameter at least thousands of millions of light-years. 



2. Condensations developing in this gas at points hundreds of 

 thousands, or perhaps millions, of light-years apart, and forming 

 separate nebulae with masses of the order of thousands of millions 

 of suns. 



3. Condensations developing in turn in the arms of these nebulae, 

 and forming stars with masses about equal to that of our sun. 



Further, according to the " Tidal theory " of the origin of the solar 

 system, we may add to this: 



4. Condensations developing in the arms of gas pulled out from 

 the stars by the tidal action of other passing stars, and forming 

 bodies of planetary mass. 



5. Condensations similarly developing in the arms of gas pulled out 

 tidally from the planets, and forming bodies of a mass comparable 

 with the satellites of the planets. 



This scheme covers five complete generations of astronomical 

 bodies, having masses of the order of 10=^ 10", 10^*, lO^^, 10-^ gm., 

 respectively, the birth of each generation from the preceding gener- 

 ation being through the agency of what I have described as " gravi- 

 tational instability." 



Owing to the repeated action of this agency, sometimes by itself, 

 but more often in conjunction with other agencies, we see the universe 

 gradually evolving from a single chaotically-spread primeval gas of 

 extreme tenuity, down to comparatively small dense bodies such as 

 our earth which form possible abodes for life. 



