SOME RESULTS FROM THE OBSERVATION OF 

 ECLIPSING VARIABLES. 



By RAYMOND S. DUGAN, Ph.D. 



{Read April 24, 1015.) 



At Princeton we have been using for over ten years a stellar 

 photometer devised by Pickering and similar to the one used for so 

 many years to such good purpose by Wendell. This photometer has 

 many virtues and but few vices. Its construction is such that the 

 observer has the very comfortable conviction that nearly all the 

 sources of systematic error he can think of are being rendered 

 innocuous by the program of observation. In accuracy it is ap- 

 parently excelled by the electrical photometers alone. 



A large part of our researches at Princeton has for some time 

 been the observational and theoretical study of eclipsing variables. 

 From the beginning we were of the opinion that the patient dis- 

 tribution of observations repeatedly throughout the entire period of 

 light-variation might very possibly bring us several facts to repay 

 us for the great labor involved. 



The first star subjected to this process was RT Persei.^ 14,464 

 measures made during the years 1905-8, combined into a mean 

 curve, showed at once the existence of a secondary minimum, 0.13 

 of a magnitude in depth, and with satisfactory distinctness a slight 

 change in brightness between eclipses. This latter change was in- 

 terpreted to mean that the two stars are ellipsoidal and brighter on 

 the sides facing each other. An asymmetry, established with con- 

 siderable certainty in the curve of primary eclipse, was found to 

 prevail throughout the entire, curve. Combining this with the 

 knowledge of one component of the eccentricity derived from the 

 observed displacement of secondary minimum toward the preceding 



1 See Contributions from the Princeton Univ. Obs., No. i. 



52 



