134 CARNEGIE INSTITUTION 



tant stars are very faint. The future study of stellar ev^olution, 

 especially in its more detailed aspects, which have as yet received 

 little attention, x\nll, therefore, in large measure be confined to the 

 most powerful telescopes obtainable. In work of this kind every 

 inch of aperture counts, and instruments far larger than any now 

 available could be used to immense advantage. 



The other field of stellar spectroscopic research opened up by 

 Huggins relates to the motions of the stars in the line of sight. 

 Early recognizing the astronomical significance of Doppler's prin- 

 ciple, Huggins proceeded in 1868 to determine stellar motions with 

 its aid. Viewed from our present standpoint, the attempt was a 

 bold one, for with the instruments at his disposal and ia the absence 

 of the almost indispensable aid of photography, the detection of the 

 minute displacements of stellar lines which result from the star's 

 motion with respect to the earth was next to impossible. It is 

 therefore not surprising that Huggins' early results show but little 

 agreement with those obtained by modern methods. The important 

 fact remains that an immense field was opened up through these 

 pioneering efforts. 



The prosecution of this research at Greenwich, though persisted 

 in for many years with inadequate facilities by Maunder, yielded 

 few results of value. It was not until Vogel, whose first attempts 

 had been made in 1871, again attacked the problem at Potsdam in 

 1887 that the great possibilities of the method were in any degree 

 realized. Profiting by the adv-anlages offered by photography, and 

 employing a spectrograph especially designed for this research, Vogel 

 substituted for the fluctuating and unsteady image seen at the tele- 

 scope, a photographed image of the spectrum, accouipanied by a 

 comparison spectrum of hydrogen or iron. The plates, measured 

 under a microscope in the most favorable laboratory conditions, 

 yielded the first reliable determinations of radial velocity. With 

 the exception of a single classical research by Keeler, who in 1890 

 measured with remarkable precision the radial velocities of the plan- 

 etary nebulae and certain stars by visual observations with the Lick 

 telescope, all accurate determinations of this nature have been ob- 

 tained by the aid of photography. 



Once set on foot by Vogel, the study of radial velocities ad\'anced 

 rapidly. In 1S98 a great step forward was made by Campbell, whose 

 improvements of methods, embodied in the Mills spectrograph of 

 the Lick Observatory, enabled him to secure results far surpassing 

 in precision those previousl}' obtained. After measuring the radial 



