876 



SCIENCE 



[N. S. Vol. XXXII. No. 833 



The systematic study of spectroscopic binaries 

 led to interesting conclusions which have since 

 been published in Lick Observatory Bulletins, 

 Volume 6, p. 17. 



Items as to New Spectroscopic Binaries: Edwin 

 B. Feost and Olivee J. Lee. (Read by Mr. 

 Parkhurst. ) 



Particulars were communicated regarding the 

 following twenty spectroscopic binaries, recently 

 detected at the Yerkes Observatory: <r Andro- 

 medai, B. D. 59°. 146, 86 p Tauri, /j. Eridani, <a 

 Orionis, v Geminorum, 42 Camelopardalis, </> Ger- 

 minorum, 7 Cancri, d Hydrae, o- Leonis, 23 Comae, 

 T] Coronae, i Serpentis, 7 Coronse, w Serpentis, 68 

 Ophiuchi, 13 Vulpeculse, 33 Cygni, 16 Lacertae. 



Probable Errors of Radial Velocity Determina- 

 tions: J. S. Plaskett. 



This paper presents the results of the measures 

 of a number of plates of the same star with three 

 different dispersions, a three-prism spectrograph 

 with a long focus and a short-focus camera and a 

 single-prism spectrograph, whose linear disper- 

 sions are approximately as 3, I4, 1. It is shown 

 that the probable error of a plate by no means 

 increases proportionately with decrease of disper- 

 sion, but that, so far as these results go, probable 

 error is only increased about 40 per cent, for a 

 decrease of dispersion from 3 to 1. A discussion 

 of these results is followed by a consideration of 

 the effect of change of spectral type, with the con- 

 sequent change in the number and quality of the 

 lines measured, upon the errors of radial velocity 

 determinations. A tabulation of the probable er- 

 rors of single plates obtained in the determination 

 of numerous spectroscopic binary orbits shows 

 how closely these errors depend upon the quality 

 of the lines for measurement. 



Visual and Photographic Magnitudes, Colors and 

 Spectral Types of the Stars to Magnitude 7.5 

 in the Zone between 13 and 77 Degrees "North 

 Declination: J. A. Paekhuest. 

 The photographic determination of star-colors 

 and their relation to spectral types was begun in 

 1906 at the Yerkes Observatory by F. C. Jordan 

 and the writer. In March, 1908, work was begun 

 on a zone centered at + 75°, with the idea of 

 extending it to the pole. The present paper de- 

 scribes the results for the stars between + 73° 

 and + 77°, 290 in number. All the work was 

 photographic, using a Zeiss doublet of 145 mm. 

 aperture and 814 mm. focal length. The photo- 

 graphic magnitudes were taken from Seed plates 

 exposed 6 mm. from the focus, giving extra-focal 



images 1.2 mm. in diameter. The opacity of these 

 images was measured with a Hartmann surface 

 photometer so calibrated as to give magnitudes 

 on an absolute scale. The " visual " magnitudes 

 were obtained from micrometer measures of the 

 diameters of focal images taken on Cramer tri- 

 chromatic plates, using a " visual luminosity " 

 filter. The magnitudes are based on the system 

 of the Potsdam Photometric Durchmusterung. 

 The spectral types were estimated from plates 

 taken with a 15° objective prism over the same 

 doublet. The Harvard classification was used. 

 The probable error of a catalogue magnitude is 

 ± 0.03 for the photographic, and it 0.04 for the 

 visual values. The color-perception of the plates 

 is a little less than the Potsdam, and about equal 

 to the Harvard catalogues. The color-spectrum 

 curve is steeper than that given in Vol. 59 of the 

 Harvard Annals. 



A71 Independent Method of Determining the Extra- 

 telluric Solar Radiation: Feank W. Veey. 

 Static methods in actinometry are at present in 

 disrepute on account of the difficulty of obtaining 

 accurate values of the instrumental corrections. 

 The proposed method overcomes this objection by 

 completely eliminating the most troublesome of 

 these corrections. There remain only theoretical 

 difficulties, and some of these will continue to 

 exist by whatever method we may approach the 

 problem. A recent new determination of the melt- 

 ing point of platinum by Day and Susman fixes 

 its value at 2048° C. Abs. within 5°. If we expose 

 a strip of thinnest platinum, coated with plat- 

 inum black, to the solar rays at the focus of a 

 condensing mirror, and gradually increase the 

 aperture of the mirror by opening an iris dia- 

 phragm, having first heated the foil until the con- 

 duction into the supports has assumed a steady 

 state, which will require about two minutes, a 

 point is reached where a further increase of the 

 aperture will almost instantly melt the platinum. 

 The experiment avoids the necessity of applying 

 an uncertain correction for loss of heat by convec- 

 tion, because, owing to the viscosity of air at high 

 temperatures and the slow motion of the air, the 

 loss by convection in the time required to melt 

 the platinum is negligible. The temperature of 

 the environment need not be considered, since 

 radiation at ordinary temperatures is insignificant 

 compared with that of melting platinum. The 

 platinum black passes to bright platinum in melt- 

 ing, but lasts long enough for the purpose of the 

 comparison, which gives the means of confronting 

 for an instant surfaces which are more nearly 



