Decembeb 16, 1910] 



SCIENCE 



881 



an exact duplicate of the primary except as to 

 brightness. The orbit was computed from seventy 

 plates obtained at the Allegheny Observatory. 

 Ten lines are the maximum number measured for 

 the primary, though twenty-eight others are dis- 

 tinct enough for an approximate determination of 

 their positions. Only seven lines of the secondary 

 can be measured even on the best plates. The 

 ratio of the mass of the brighter star to that of 

 the fainter is 1.43. If the surfaces of the two 

 were of the same brightness per unit area, and the 

 densities the same, the difference in mass would 

 imply a difference of but 0.27 magnitude, which is 

 undoubtedly too small; therefore the secondary is 

 either denser, or its surface brightness is less. 

 The point of special interest in this star is the 

 fact that the calcium lines H and K show con- 

 stant velocity. K. is measurable on sixty plates, 

 from which is derived a mean velocity of + 12.4 

 km. with a probable error of ± 4.27 km. for an 

 average plate, while the probable error of the 

 mean is ± 0.55 km. If instead of Rowland's 

 wave-length of 3933.825 we use the mean derived 

 from seven first type stars, 3933.768, the velocity 

 becomes + 15.4 km. If we use the value derived 

 by St. John at Mt. Wilson, 3933.667, the velocity 

 is increased to -|- 25 km. As the velocity of the 

 center of mass of the system is + 18.46 km. it 

 is impossible to say whether the velocity of the 

 calcium vapor is the same or not, but it can not 

 be much different. The H line of calcium gives a 

 velocity of + lS-4 km. or + 29.4 km. depending 

 upon whether we adopt Rowland's or St. John's 

 wave-length. This line is a difficult one to meas- 

 ure because of the proximity of the diffuse He 

 line, and this may account for the difference be- 

 tween its velocity and that of the K line. This 

 paper is to be published in volume II. of the 

 Publications of the Allegheny Observatory. 



Th« Spectrum of the Chromosphere and the Ap- 

 plication to it of some Recent Laboratory In- 

 vestigations: Waltee S. Adams and Henky G. 

 Gale. (Read by Professor S. A. Mitchell.) 

 The first part of this paper is a continuation of 

 the work of Hale and Adams, Astrophysical Jour- 

 nal, 30, 222, on the photography of the flash spec- 

 trum without an eclipse. The number of lines 

 obtained on these plates compares favorably with 

 that afforded by eclipse plates, being much richer 

 in the red but not so rich in the blue. About 

 97 per cent, of the bright lines can be identified 

 with dark lines in Rowland's table; of the re- 

 maining 3 per cent, a few are due to helium. A 



marked feature of probably all the bright chromo- 

 spheric lines is the tendency to double reversal as 

 the sun's limb is approached; out farther they 

 assume the cliaracter of simple bright lines. The 

 enhanced lines, as compared with the arc lines, 

 appear with much greater intensity as bright 

 lines in the chromosphere than as dark lines in 

 the ordinary solar spectrum; this agrees with the 

 previous results of Lockyer, Evershed, Dyson and 

 others. An attempt was made to identify the 

 coronal line at X 5303.26 it 0.15; the photographs 

 show a well-marked line at X 5303.36, but it is 

 probable that this is coincident with the dark 

 line at X 5303.401 in Rowland's tables. 



The second part of the paper is concerned with 

 an investigation of the spectrum of the electric 

 spark under pressure, made by Mr. Gale at the 

 Pasadena laboratory. In the case of titanium it 

 was found that in the region around X 4300 the 

 arc lines become completely reversed under pres- 

 sures of five or six atmospheres while the en- 

 hanced lines remain bright. On plates taken at 

 longer wave-length, however, some of the arc lines 

 do not reverse, and in general the proportion of 

 unreversed lines increases with the wave-length; 

 this accords with Hale's result in Astrophysical 

 Journal, 15, 227. Similar results have been ob- 

 tained for iron and chromium, except that higher 

 pressures seem to be necessary to produce the 

 maximum number of reversals. The authors then 

 apply these results to the spectrum of the chromo- 

 sphere as follows: Fabry and Buisson have shown 

 that the pressure in the sun's reversing layer is 

 between five and six atmospheres, which is the 

 pressure used for most of the above laboratory 

 experiments. We should accordingly expect that 

 the enhanced lines will appear bright in the chro- 

 mosphere while most of the arc lines remain dark, 

 and this accords with the observed facts. More- 

 over, as we pass to the longer wave-lengths both 

 the chromosphere and the plates of the spark 

 spectrum taken in the laboratory show an in- 

 creasing number of arc lines that do not reverse. 



Some Results of the Study of the Spectra of 

 Sirius, Procyon and Arcturus vnth Bigh Dis- 

 persion: Waltee S. Adams. (Read by Mr. 

 Plaskett.) 



The material discussed by Mr. Adams in this 

 paper consists of six plates of Sirius extending 

 from X 4200 to X 6G00, four of Procyon from 

 X 4200 to X 4900, and nine of Arcturus from 

 X 4300 to X 6600. These were secured with an 

 auto-collimating prism spectrograph of 18 feet 



