720 



SCIENCE 



[N. S. Vol. XXXVIII. No. 986 



"We thus have strong eonfirmation here of 

 what, on totally different grounds, is 

 pretty generally considered as the order of 

 the different ages in a star's life. 



But to proceed : As the younger the stars, 

 the smaller their internal motion, it follows 

 at once that from whatever matter our 

 youngest stars — the helium stars— may 

 have been evolved, that matter must, in all 

 probability, have still smaller internal mo- 

 tion. Let us call this matter primordial 

 matter. As the internal velocity of the 

 helium stars is already so very small we 

 come to the conclusion that primordial 

 matter must practically have hardly any 

 other motion than the motion of the cloud 

 to which it belongs. 



There is more. According to the second 

 of the observed facts, the second stream, 

 which is rich for the older stars, is much 

 poorer for the younger ones ; it almost dies 

 out in the helium type stars. "We must ex- 

 pect, therefore, that for primordial matter 

 there will practically be no second stream 

 or second star cloud. 



Therefore, finally, we must expect that 

 the particles of primordial matter will all 

 move in practically parallel lines, and that 

 in the direction in which all but a very few 

 of the helium stars move, and with the 

 same velocity. 



Now it is a very general notion that it is 

 from the nebulae that the stars are formed. 

 Therefore that what we called primordial 

 matter would be nothing else than the 

 matter of the nebulas. "What precedes gives 

 us the means of testing the notion by ob- 

 servation. What then does observation 

 show? 



The number of available data is as yet 

 extremely small. The determination of 

 what we call astronomical proper motion 

 of these very ill-defined objects is extremely 

 difficult, and has been up to the present 

 time invariably unsuccessful. For the de- 



termination of the radial velocity by the 

 spectroscope, the faintness of the nebulaB 

 is a serious obstacle. The consequence is 

 that, as yet, we know the radial velocity of 

 only fourteen of these objects in all. Still, 

 even this limited number is decisive in 

 showing that there can be no question that 

 the real motions of these objects are ap- 

 proximately parallel to the motion of the 

 helium stars, or even parallel to any fixed 

 direction whatever. Their velocity, more- 

 over, is exceedingly unequal. Must we con- 

 clude that the nebulee are not the birth- 

 place of the stars ? It may seem so. 



Meanwhile let us not go too fast. There 

 are nebulse and nebulae. It so happens — and 

 there is ample practical reason for it — ^that 

 with one exception observation of radial ve- 

 locity has, up to the present time, been con- 

 fined to what we call the planetary nebulas — 

 elliptical or round nebulee — which show an 

 appearance remotely like that of a plane- 

 tary disc. Herschel saw in them a likeness 

 to what, according to Laplace's cosmog- 

 ony, must have been the primitive stage of 

 our own planetary system and so imagined 

 that these planetary nebulse must be the 

 birthplace of the stars. 



According to what precedes, this view 

 seems now untenable. The planetary neb- 

 ulas can not be the birthplace of the stars. 

 If they were, they would show the parallel 

 and equal motion of practically all the 

 helium stars. Their motions, on the con- 

 trary, are extremely unparallel and un- 

 equal, and we must rather assign these ob- 

 jects a place at the end of the order of evo- 

 lution than at the beginning. 



"We may, perhaps, see an independent 

 confirmation of this view in the stars called 

 temporary stars, but time will not permit 

 me to pursue the argument further. 



As I said just now, there is one nebula 

 for which the radial velocity has been de- 

 termined which is not a planetary. This 



