32 



SCIENCE 



[N. S. Vol. XXXVII. No. 940 



as prismatic spectrum. Region X4100-X3850, 

 diffraction spectrum of nearly uniform intensity 

 and nearly as strong as at H^. Prismatic spec- 

 trum, none. It is evident that with this disper- 

 sion the diffraction spectrograph is superior, espe- 

 cially towards the violet and ultraviolet. When 

 it is compared with a single prism spectrograph 

 of lower dispersion the advantage will likely be 

 the other way, but if a spectrum of uniform in- 

 tensity over a wide range is required, and in the 

 red end, where the prismatic spectrum is unduly 

 compressed, the grating spectrograph will un- 

 doubtedly possess many advantages. As soon as 

 possible a more complete test of its performance 

 will be made. 



The Cincinnati Astronomical Society: DeLisle 

 Stewart. 



This paper describes the organization and activi- 

 ties and plans of the society. The suggestion is 

 made that some definite connection between the 

 national and local societies might be worth con- 

 sideration. 



Eadium and the Chromosphere: S. A. Mitchell. 

 This paper has been published in A. N., 4600, 

 BO only the conclusions need be given here. 



At the 1905 eclipse, while a member of the 

 V. S. Naval Observatory expedition in Spain, the 

 writer photographed the ' ' Flash Spectrum ' ' with 

 such dispersion and definition that it permitted 

 the determination of wave-lengths which are in 

 error but a few hundredths of an Angstrom. A 

 comparison with the radium spectrum showed that 

 a majority of the lines due to radium nearly coin- 

 cide in the chromosphere with lines already satis- 

 factorily identified from other sources. The third 

 strongest line of radium at X4S26.10, if present 

 in the chromosphere, must be very weak. The 

 comparison between the spectra of radium and of 

 radium emanation and the chromosphere has led 

 me to the conclusion that there are no radium lines 

 in the sun. A similar lack of uranium lines like- 

 wise appears. 



It wUl be necessary to obtain photographs of 

 novse with a much greater dispersion than that 

 used at Bonn before deciding that radium or 

 uranium lines are found in their spectra. 

 Absorbing Medium in Space: Edward S. King. 



The results given in Harvard Annals, 59, indi- 

 cated the presence of an absorbing medium in 

 space. The amount found was 0.0377 mag. for 

 the photographic rays, and 0.0184 mag. for 

 the visual rays, while traversing the unit of dis- 

 tance, which is measured by 32.6 light-years. 



These figures having been derived from 26 stars, 

 it was thought best to improve the data by apply- 

 ing the method to other stars. Accordingly, a 

 supplementary list of 22 stars was prepared from 

 the Publications of the Groningen Laboratory, 

 No. 24, and observed. The procedure was in all 

 respects precisely that followed previously. The 

 resulting values had the same sign, indicating 

 absorption by greater redness, but were somewhat 

 larger than before. By selecting from both lists 

 only those stars having the parallax most accu- 

 rately determined, the reduction gave values iden- 

 tical with the above results from Harvard Annals, 

 59. Thus, the indications from this work also 

 point to the presence of an absorbing medium in 

 space, or some factor which produces effects sim- 

 ilar to absorption by making the more distant 

 stars appear redder. 

 Proper Motions of Faint Stars: George C. Com- 



STOCK. 



The writer, who has been engaged for some 

 years upon a study of the motions of the faintest 

 stars for whose determination adequate material 

 can be found, presents in this paper results re- 

 lating to more than 250 stars fainter than the 

 tenth magnitude of the Harvard scale. These 

 motions have been derived from micrometer ob- 

 servations by which the positions of the faint stars 

 are referred to neighboring brighter ones of known 

 proper motion, the time interval covered by the 

 measures being in a few cases as small as a quarter 

 century but more frequently extending to fifty or 

 seventy-five years. The probable error of the 

 motion assigned each star in each coordinate has 

 been determined, and from a comparison of these 

 numbers with similar data given in Boss's Pre- 

 liminary General Catalogue it appears that the 

 proper motions here considered, tenth to thirteenth 

 stellar magnitude, are at least of equal precision 

 with those given for the fainter stars of the Boss 

 Catalogue, 7.5 to 9.0 magnitude. The proper 

 motions here considered are all referred to the 

 Boss system. 



In a certain number of cases the motions here 

 investigated are shown to be orbital, the faint star 

 being physically connected with its brighter com- 

 panion; but when the angular distance separating 

 the stars exceeds the Struve double-star limit of 

 32" this connection is found to be very infrequent 

 and in about 80 j)er cent, of the cases here in- 

 vestigated, i. e., more than 200 stars, the motion 

 of the faint star appears to be uninfluenced by its 



