314 CARNEGIE INSTITUTION OF WASHINGTON. 



monly styled "gaseous," only a portion exhibit such a systematic dis- 

 tribution of luminosity as to imply a true gaseous organization in the 

 recognized physical sense. The pecuUar distribution of matter in 

 these small nebulae, so far as such distribution can be judged from 

 luminosity, awaits dynamic study as fast as the requisite definition 

 becomes available. Meanwhile, it is unwise to rest much on their 

 "gaseous" nature. Their spectra imply free molecules, indeed, so 

 charged that they are luminescent, but the dynamic control of these 

 free luminous molecules is, in most cases, as much a problem yet to be 

 solved as if they were small aggregates of matter in other than molecu- 

 lar states. Probably nothing can be safely regarded as really gaseous 

 that does not conform to the laws of distribution of true gases. 



The foregoing analysis of the assigned conditions of genesis of nebulae 

 bears definitely on their distribution, and only a few words are needed 

 to make clear their bearings on the alternative views of the nature of 

 spiral nebulae offered respectively in behalf of the revived Herschellian 

 hypothesis and of the earlier hypothesis of dynamic encounter. The 

 former places the star-clusters that masquerade as nebulae far out in 

 space in positions in no assignable way dependent on our galactic 

 system. Their mean distances are presumably about equal in all direc- 

 tions. With equally good seeing in all directions, no specific localiza- 

 tion is assignable. 



The hypothesis that the nebulae are dispersed stellar matter places 

 the large spiral nebulae in the open outer zone of the galactic system, con- 

 centric to its denser part. They thus vary greatly in distance from our 

 point of observation; those in the outer zone beyond the rim of the 

 denser galactic disk may be ten times as far from us as those at the 

 sides of the disk. 



Under the Herschellian hypothesis the intrinsic light of the star- 

 clusters is great; their dimness depends on distance. At such great 

 distances the gradient of diminution of light is very low and great 

 differences in remoteness are required to produce a given degree of 

 dimness, if clear seeing is assumed. Under the alternative view, the 

 intrinsic light of the nebulae is relatively low, their dimness being due 

 as much to this as to remoteness. At the shorter distances to which 

 they are assigned, the gradient of diminution in luminosity is higher 

 and lesser differences of distance yield a given diminution. It is thus 

 not unreasonable to beUeve that such differences of distance as the place- 

 ments under this hypothesis postulate may mark the change from 

 visibility to invisibility. 



In so far as intervening matter is a cause of invisibility, both views 

 stand on essentially the same basis, except that a comhination of greater 

 distance with increasing intervention of matter is more available for 

 the latter than for the former. 



