614 



SCIENCE. 



[N. S. Vol. XVII. No. 433. 



variable star regions referred to above, 

 and twenty-five exposures for miscellaneous 

 purposes. 



A small slitless speetograph was designed 

 by Professor Keeler for use on faint ob- 

 jects with the Crossley reflector. It was 

 completed on the day of his departure from 

 the mountain. It was tested promptly by 

 Messrs. Campbell and Palmer, who found 

 it necessary to use convex and concave 

 quartz lenses in connection with the quartz 

 prism, in order that the rays should be 

 parallel when passing through the quartz 

 prism. These changes were designed by 

 Mr. Palmer, and the instrument was used 

 extensively by him. He secured seventy 

 spectrograms of the smaller planetary 

 nebulffi and of other small objects. Many 

 interesting facts resulted from these ob- 

 servations. I shall refer only to his suc- 

 cess in photographing extremely faint 

 spectra. A strong image of the spectrum 

 of Nova Cygni, visual magnitude about 

 15.5, was obtained with ease. Successful 

 exposures could probably be made on stars 

 at least a magnitude fainter. His photo- 

 graph of Nova Cygni demonstrates that the 

 spectrum, which was nebular in 1877, has 

 noAV become continuous, like that of the 

 ordinary stars. 



In addition to the observations of Eros, 

 positions of asteroids 1900 GA, Ohio, and 

 Palatia, were determined by Mr. Palmer, 

 from photographs taken with the Crossley 

 reflector. Mr. Hussey secured eight ob- 

 servations of the asteroids Minerva, Edna, 

 440, and Chicago. Messrs. Palmer and 

 Curtiss have recently secured photographs 

 of several asteroids whose positions were 

 requested. 



Three nights per week with the thirty- 

 six-inch equatorial have been devoted to 

 the determination of the motions of the 

 brighter stars in the line of sight, with the 

 Mills spectrograph, during the past six 



years. The accuracy of the Lick Observa- 

 tory determinations has steadily progressed 

 until, for the stars containing fine lines, 

 the probable error of a single determina- 

 tion of velocity is only about 0.25 kilo- 

 meter. 



To the list of fifteen spectroscopic bi- 

 naries discovered prior to Director Keeler 's 

 report of July 1, 1900, I desire to make 

 twenty-three additions, as follows: 



Beta Hereulis, 12 Persei, 



Xi Ursae Majoris 93 Leonis, 



Delta Bootis, Beta Scuti, 



113 Hereulis, 2 Scuti, 



Eta Andromedae, Kappa Pegasi, 



Pi Cephei, 31 Cygni, 



Xi Piscium, Tau Persei, 



Xi Prime Ceti, Epsilon Hydrse, 



Delta Equulei, Alpha Equulei, 



Zeta Hereulis, Phi Persei, 



Omicron Andromedae, Eta Geminorum. 

 Gamma Canis Minoris, 



These thirty-eight systems have been dis- 

 covered since 1898. 



There is room for reference to only two 

 of the stars on the above list : Zeta Hereu- 

 lis is a short-period visual binary star, 

 completing a revolution in about thirty- 

 three years. The velocity of the principal 

 star in the line of sight is slowly varying. 

 Kappa Pegasi is one of the most interest- 

 ing visual binaries known, period eleven 

 and one third years. Until the discovery 

 of the true period of Delta Equulei, this 

 was supposed to be the shortest period 

 known. One of the components of Kappa 

 Pegasi is a spectroscopic binary, having a 

 period of only six days. 



These binary systems have been discov- 

 ered in the process of determining the 

 velocities of about 350 stars ; in this list of 

 350 previous observers had discovered three 

 binaries. Without taking into accouiit a 

 list of several suspected binaries, it is ap- 

 parent that of the brighter stars at least 

 one in every seven or eight is attended by 

 an invisible companion. When we con- 



