4 Introductory Chapter [CH. i 



After allowing statistically for the effects of projection on the celestial 

 sphere, Russell* finds that triple systems consist normally of a close double 

 with a third star revolving at a considerable distance about their centre 

 of gravity, the ratio of the actual separations being about 10 to 1. The 

 bearing of this on questions of cosmogony will be considered later ; for the 

 present it is sufficient to notice that the multiple systems observed in the sky 

 shew no resemblance to our own solar system. 



Thus we have found a very definite uniformity of arrangement inside our 

 system, and a very definite uniformity of arrangement outside, but the two 

 arrangements are different, and the question of whether there are other 

 systems arranged like our own has to remain an open one. It may perhaps 

 be mentioned that some astronomers believe that there are irregularities in 

 the motion of binary systems which are too definite to be ascribed merely to 

 errors of observation. These may ultimately be found to point to the exist- 

 ence of planetary bodies revolving at a great distance round the central binary 

 system, but the evidence is certainly too vague at present for definite con- 

 clusions to be drawn. 



Our search outside our own system has, however, disclosed the existence 

 of a second uniformity of structure, namely that of binary stars having masses 

 not far from equal. 



Spiral and other Nebulae 



3. These two uniformities, namely the planetary formation and the 

 double-star formation, although perhaps the most striking, are by no means 

 the only uniformities which have been discovered by astronomy. Principal 

 among the remaining ones is the spiral nebula formation which appears to be 

 very distinctive and uniform. The characteristic spiral nebula consists in- 

 variably of a nucleus with two arms emerging from opposite points; the 

 convolutions of the two arms are similar, the curve of each being approxi- 

 mately an equiangular spiral f. This formation is very freely scattered in 

 space : Keeler and Perrine estimated the number of nebulae easily discoverable 

 with the Crossley reflector to be of the order of half a million, while Keeler 

 found more than half of the nebulae recorded on his plates to be spirals +. 

 Although the spiral nebulae are only special instances of the more general 

 nebular formations found in the sky, they are nevertheless the most frequent 

 and the most distinctive of these formations known ; for cosmogony they are 

 the most interesting because the definiteness of their formation must contain 

 a valuable clue to their origin and condition. Besides spiral nebulae there 

 are other types of nebulae, which are commonly described in the following 

 terms. 



" Astrophys. Journ. 31 (1910), p. 200. f v. Pahlen, Ast. Nach. No. 4503. 



Campbell, Stellar Motions, p. 36. 



