204 CARNEGIE INSTITUTION OF WASHINGTON. 



slight variations in the radiation of bright stars will depend upon the precision 

 with which the atmospheric absorption can be eliminated (p. 237). 



Mr. Abbot has applied an improved type of vacuum bolometer and gal- 

 vanometer to the measurement of the energy spectra of Aldebaran, Capella, 

 and Betelgeuse at the coude focus of the 100-inch reflector. The first 

 successful bolometric determination of the distribution of energy in the 

 spectra of stars is an event of great importance in the progress of astron- 

 omy. Using a prism of U. V. crown glass with an equivalent angle of 36°, 

 a galvanometer with a period of single swing of 1.56 seconds and a mirror- 

 scale distance of 4.7 meters, Abbot observed visually maximum deflections 

 of 14 mm. in the infra-red of Betelgeuse and about one-half that amount in 

 Aldebaran and Capella. The energy distribution in the spectrum of Capella 

 is found to be similar to that of the sun, while that of Betelgeuse is very dif- 

 ferent and shows a sharp maximum at 0.83 ix. The spectrum of Aldebaran 

 shows results intermediate between those for the other two stars. 



The extraordinary degree of sensitiveness required for these observations 

 is shown by the fact that with the apparatus used the current passing through 

 the galvanometer corresponding to a deflection of 1 mm. on the scale is 2.5 X 10 _u 

 amperes, and the corresponding rise of temperature of the bolometer strip 

 8X10 -8 degrees centigrade. In Abbot's opinion an increase of at least ten- 

 fold in sensitiveness can be attained, and with the aid of photographic regis- 

 tration deflections of 0.5 mm. can be observed without difficulty. This will 

 make it possible to secure spectrobolometric observations of stars to the third 

 and possibly the fourth magnitude (p. 239). 



Professor Michelson's investigations of the possibility of measuring the 

 velocity of light between two stations separated by a distance of over 20 

 miles show that while the intensity of the return beam in the direct Foucault 

 arrangement with the rotating mirror is insufficient, ample light may be 

 secured by a combination of the Foucault and Fizeau forms of apparatus. 

 In the form adopted the observer sees the light returned from the distant 

 station with an intensity which depends upon the speed of the rotating mirror. 

 The light gradually becomes less intense as the speed of the mirror increases, 

 falls to zero, and then suddenly rises to its full amount when the speed is such 

 that a stcond face of the rotating octagonal mirror replaces the first. The 

 observation accordingly consists in adjusting the speed of the mirror to the 

 rate at which the light suddenly reappears, and with good seeing Michelson 

 considers that this can be done with an error not exceeding 1 part in 55,000. 

 Apparently the measurement of the speed of the mirror with the requisite 

 accuracy will present no serious difficulty, to judge from previous experience. 

 The essential question of the intensity of the return light being answered 

 satisfactorily, we expect to be able during the coming winter to construct the 

 apparatus necessary for a complete test of the method when Michelson returns 

 next summer (p. 242). 



Reference was made last year to the experiment devised by Michelson for 

 examining the effect of the earth's rotation on the velocity of light. If a 

 beam of light is divided and sent around a circuit of 1 mile in opposite direc- 

 tions, the resulting interference fringes, according to the generalized theory of 

 relativity on the hypothesis of a stationary ether, should show a displacement 

 of about 0.15 fringe. If no displacement or one smaller than this amount were 



