COSMOGONIC HYPOTHESES. 23 1 



In building up the above hypothesis, the following facts of 

 observation have been borne in mind : — 



1. The mutation of heavy elements, such as Uranium 



(atomic weight 238.5) into Helium (atomic weight 

 4) with an enormous liberation of energy spread over 

 thousands of millions of years. 



2. The changing of stars into nebulae of which some four 



or five cases are known, whereas the reverse process 

 is unknown. 



3. Gaseous stars (spectra showing helium and hydrogen, 



with or without^ bright lines) are very light, their 

 density not exceeding i-qth that of the Sun, whilst 

 their gravitative ix)wer seems to be " nil." Thus the 

 brilliant close pair of a Crux shows no orbital motion, 

 whilst the essentially wider solar-type star a Centaurus 

 is a rapid binary pair.* 



4. Nebulous matter is found near most stars of the gaseous 



types — thus the nebulous regions of Orion are in the 

 midst of helium-type stars, nebulous matter i> un- 

 known near solar-type stars. 



As under the explosion hypothesis the Sun is liquid, it cannot 

 maintain its temperature by contraction, because liquids are virtu- 

 ally incompressible, hence Helmholz's theory of the maintenance 

 of solar heat is not applicable. It is further improbable that 

 gaseous or stellar-type masses always contract as they radiate 

 heat; on the contrary, Kelvin's investigations indicate very 

 strongly that such masses of gas may expand. The argument 

 that spiral nebulae are systems in formation overlooks the palpa- 

 ble fact that these objects are exceedingly faint. Long exposure 

 photographs give very misleading pictures of spiral nebulae. In 

 nearly every case the total brightness is less than any one of its 

 neighbouring small stars. It would be of the same order of 

 reasoning to assert that islands are formed out of wisps of cirrus 

 cloud. 



The comn.ionly received view that gaseous stars are hotter 

 than liquid or sun-type stars has perhaps been engendered by the 

 classification really based on the nebular hypothesis, vis., that 

 white stars are the hottest, and that sun-type stars already show 

 signs of cooling, but Huggins clearly shows that solar-type stars 

 are the hotter — thus in his " Atlas of Spectra," 1S99, p. 85, he 

 says : — 



In strong contrast with this falling off in Vega at about X 3.700, 

 the continuous spectrum of the solar stars. Procyon and notably Capella — ■ 

 that is to say, the narrow bright intervals between the numerous strong 

 dark lines ... is obviously far more intense. 



And it may fairly be asked, if the gaseous or stellar 

 type of star is the hotter, why it should not show metallic lines in 



*" Wider" is used in a general sense and includes the effect of 

 surface luminosity, because as far as distance alone goes, a Centaurus 

 is not so wide as a Crux. See Newcomb's remark on a star in Orion, 

 which is quoted in the references given later. 



