SCIENCE 



[N. S. Vol. XXVIII. No. 728 



The mean of the residuals obtained by subtract- 

 ing — H'.SO from each of the above clock cor- 

 rections is 0'.04. 



Observations were discontinued for some weeks 

 beginning the latter part of June, so that, in order 

 to show how well the clock correction could be 

 predicted a new rate formula was deduced using 

 the observations from February 6 to June 1. 

 Using this rate formula 



-f 0'.00207 — 0».01230 (B — 677°"".5 ) 



— 0^000373 (T — April 14.5) 



and the clock correction for April 15.0 as 

 — 14^.39, the predicted clock corrections for June, 

 1905, were obtained as given in the accompanying 

 table : 



CLOCK COEEECTIONS FOE JUNE, 1905 



Ilhimination of the Reflex Zenith Tube: C. L. 



DOOLITTLE. 



As the instrument was originally constructed, 

 the field was illuminated by a small electric lamp 

 placed outside the tube, the light from which fell 

 on a diagonal reflector in the axis of the instru- 

 ment. The mirror was only eight inches from the 

 micrometer threads. The result was that when 

 the thread was not practically in the axis the 

 reflection back and forth from the under side of 

 the prism and the mercury surface produced a 

 feathery appearance in the threads, making it 

 impossible to obtain a sharp image of the same. 



After considerable discussion and some experi- 

 menting the following plan was adopted: 



The smallest electric lamp obtainable — some- 

 times called a dental lamp — is placed in the axis 

 of the instrument, six inches above the mercury 

 surface. This is held in place by a thin strip of 

 brass attached to the tube in such a way that it 

 may be readily removed in case repairs or re- 

 moval of the lamp are called for. The current is 

 furnished by a dry battery, and the brightness 



controlled by a rheostat. It is all of home con- 

 struction and gives perfect satisfaction. 



On the Constancy of the Period of the Variable 

 Star, M5 (Librs) No. 33: E. E. Babnabd. 

 This is one of the Harvard cluster variables 



and has been under observation with the 40-inch 



for nearly ten years. The period derived from 



these observations is 



12''2"7'.3040 ± 0^01781. 



The period was independently determined from 

 normal observations at intervals of one or two 

 years. The deviations among these values did not 

 exceed 0'.065. Though this is approximately 

 progressive with the time, it is not believed to 

 be real. 



An observation of the variable at its most rapid 

 light change is subject to a probable error of 

 about ±: 1.2 minutes, which, when three or four 

 normals are employed, is reduced to ± 0.5 minute. 



The light curve shows that the star, after re- 

 maining faint for a large part of its period, 

 rather suddenly begins to brighten and in one 

 hour has reached maximum. It remains but a 

 short time at maximum. The decrease for the 

 first half hour is almost equally as rapid as the 

 rise. It then fades slowly for the next six or 

 eight hours to minimum. The entire light change 

 is about 1.2 magnitudes from 15" ± to 14° ±. 



The Photoheliometer: Chas. Lane Poob. 



By an arrangement with Professor Frost a 

 25-foot photographic heliometer was mounted 

 upon the tube of the 40-inch Yerkes telescope. 

 The lenses were 2 inches in aperture and were 

 specially made for this work by Brashear. The 

 centers of the lenses were at a fixed distance 

 apart and this distance was so adjusted as to 

 make the photographic images of the sun overlap. 

 The common chord of these overlapping images 

 is a measure of the diameter of the sun, and 

 slight variations in the diameter produce rela- 

 tively large changes in the length of the chord. 



During 1907 a series of photographs were taken 

 by Mr. Fox, but owing to bad weather the number 

 was rather limited. They were sufficient, how- 

 ever, to test the value of the method and to show 

 the general lines upon which an instrument 

 should be built. 



A Possible Third Body in the System of Algol: 



E. H. CUETISS. 



Spectroscopic determinations of the center of 

 mass velocity of the eclipsing stars of Algol indi- 

 cate strongly that three bodies enter into this 



