May 25, 1917] 



SCIENCE 



531 



other great universes of stars verj^ far dis- 

 tant from our own universe of stars is re- 

 ceiving favorable consideration, as far as 

 the spiral nebulffi are concerned. In fact, 

 the related hjqDothesis that our own stellar 

 system, if viewed from a stupendous dis- 

 tance, would be seen as a spiral nebula, has 

 been seriously proposed and is receiving 

 favorable consideration. There is much 

 merit in the hypothesis that if an observer 

 went very far away from our sj^stem, in 

 the direction of one of the poles of the 

 Galaxy, and looked back to our stellar sys- 

 tem, he would see it fairly circular in out- 

 line, and that the cloud forms which we see 

 in the galactic structure would to the dis- 

 tant observer resemble the ill-defined con- 

 densations so characteristic of the outer 

 structure of spiral nebulae (see Fig. 5). If 

 the observer were at a very great distance 

 outside of our system in the plane of Milky 

 Way extended, we certainly think he would 

 see our system as a greatlj^ elongated ellipse 

 resembling the many well-known spindle 

 nebulfe which are interpreted as spirals 

 seen edgewise (see Fig. 6). 



Whether we regard the spirals as very 

 large bodies at enormous distances, far out- 

 side and independent of our system, or as 

 within our system and therefore compara- 

 tively small bodies close to us, we encounter 

 apparent difficulties of interpretation. 

 These proceed chiefly from their avoidance 

 of the direction of the Milky Way. We are 

 not content to think of this relationship of 

 avoidance as a coincidence or accident of 

 nature. If the spirals are far outside our 

 system, we should expect to see a great 

 many of them in the direction of the Milky 

 Way, but beyond its structure. 



We have seen reasons for believing, or at 

 least strongly suspecting, that there is an 

 immense amount of obstructing material in 

 our system, and that this would be the most 

 extensive and the most efiiective in the long 

 dimensions of our sj^stem, and the least ef- 



fective in the direction of the short axis of 

 the system. If such obstructions are 

 operating effectively upon the light of ex- 

 tremely faint and extremely distant nebula, 

 they should jDroduce something like the 

 distribution which we actually observe to 

 exist amongst the spirals. 



Fig. 20. Hypothetical Cross section of our 

 Stellar System, showing spaces N and 8 occupied 

 by the spiral nebulae if they are within the stellar 

 system. 



Let us suppose, on the other hand, that 

 the spirals are in our system. Why do we 

 not find them in the Milky Way structure ? 

 Why do they absolutely avoid that struc- 

 ture ? Why do they congregate around the 

 shorter axis of our stellar system? If they 

 are within our system they must be distrib- 

 uted as to distance and direction somewhat 

 as sketched in the conical volumes N and S 

 in Fig. 20, which I'eiDresents a cross-section 



