264 SCIENCE AND METHOD. 



But this does not matter, since it is not about the 

 rectilinear motion of our system that we ask them for 

 information, but about its rotation. The fixed stars, 

 by their apparent motion, disclose the diurnal rotation 

 of the Earth, although their distance is immense. 

 Unfortunately, the possible rotation of the Milky 

 Way, rapid as it is, relatively speaking, is very slow 

 from the absolute point of view, and, moreover, bear- 

 ings upon nebulae cannot be very exact. It would 

 accordingly require thousands of years of observation 

 to learn anything. 



However it be, in this second hypothesis, the figure 

 of the Milky Way would be a figure of ultimate 

 equilibrium. 



I will not discuss the relative value of these two 

 hypotheses at any greater length, because there is a 

 third which is perhaps more probable. We know that 

 among the irresoluble nebulae several families can be 

 distinguished, the irregular nebulae such as that in 

 Orion, the planetary and annular nebulae, and the 

 spiral nebulae. The spectra of the first two families 

 have been determined, and prove to be discontinuous. 

 These nebulae are accordingly not composed of stars. 

 Moreover, their distribution in the sky appears to 

 depend upon the Milky Way, whether they show a 

 tendency to be removed from it, or on the contrary 

 to approach it, and therefore they form part of the 

 system. On the contrary, the spiral nebulae are 

 generally considered as independent of the Milky 

 Way : it is assumed that they are, like it, composed 

 of a multitude of stars ; that they are, in a word, 

 other Milky Ways very remote from ours. The work 

 recently done by Stratonoff tends to make us look 



