Febeuaey 27, 1903.] 



SCIENCE. 



325 



Bruce spectrograph of the Yerkes Observa- 

 tory. Stars with spectra of this interest- 

 ing type, which seems certainly to char- 

 acterize an early stage of stellar develop- 

 ment, have not hitherto received much at- 

 tention in respect to motion in the line of 

 sight. These spectra are not adapted to 

 measurements of any such degree of accu- 

 racy as is possible for the solar stars, be- 

 cause of the comparative fewness and the 

 generally hazy and ill-defined character of 

 their lines. The results have, however, 

 proved more accordant than was antici- 

 pated. The general good adjustment and 

 trustworthiness of the spectrograph are at- 

 tested by the measures of the moon 's radial 

 velocity. The twelve lunar spectra pho- 

 tographed during the year gave a mean 

 difference of 0.2 km. per sec. between the 

 observed and the computed radial velocity, 

 the largest difference being 0.7 km. per 

 sec. The titanium spark was chiefly used 

 for furnishing the comparison spectrum, 

 but the iron and the chromium spark and 

 a helium tube were also employed at times. 

 A perfectly definite amount of self-induc- 

 tion and capacity was always maintained 

 in the secondary circuit. 



The lines commonly present and meas- 

 ured in the stellar spectra were those due 

 to one or more of the following elements : 

 helium, oxygen, silicon, nitrogen, hydro- 

 gen, magnesium. The observing list in- 

 cluded about 150 stars of this type brighter 

 than the sixth magnitude. The present 

 paper included only those of which three 

 or more plates have been obtained and 

 measured. Those foimd to vary in their 

 radial velocity, six in number, were not 

 included in the discussion. The stars of 

 the Orion Unpe are peculiarly distributed 

 in the sW, being for the most part grouped 

 in or near the Milky "Way. As many of 

 the twenty are near the apex or anti-apex 

 of the sun's way, the ob-served velocities 

 clearly show the effect of the solar motion. 



If a correction were applied for this mo- 

 tion, the resulting absolute radial velocities 

 would be small. The angular pi-oper mo- 

 tions of these stars are also small, and sug- 

 gest a relatively great distance from our 

 sun, as well as a 'community of interest' 

 of these stars. The radial velocities ob- 

 served, expressed in kilometers per second, 

 are as follows: 



(Paper will appear in full in the 'De- 

 cennial Publications of the University of 

 Chicago.') 



Neiv Spectroscopic Binaries: Edwin B. 



Frost and Walter S. Adams. 



During the observations described above, 

 six stars of the Orion type were found 

 whose radial velocity varied. Preliminary 

 statements have already been published as 

 to three of these ( vj Orionis, <> Persei, /? 

 Cephei). The others are <^ Ceti, Z Tauri 

 and ■^Eridani. Of ''^Ceti we have obtained 

 eleven plates since November 1, 1901, which 

 give a range from -\- 6 to +16 km. per 

 sec. The period is short, but observations 

 on consecutive nights will be necessary for 

 its establishment. 



The plates of ? Tauri available are ten 

 in number (from November 8, 1901, to 

 December 18, 1902), and give a range from 

 + 7 to + 34 km. The period can not yet 

 be given, but may, perhaps, be about four- 

 teen days. The spectrum is rather unique 

 in respect to its very sharp and strong y 

 and (5 lines of hydrogen, with the other 

 lines (some of them metallic) very faint. 



