68 McDIARMID— VARIABLE STARS. [April 24, 



that the fainter star was twice as bright on one side as the other. 

 The stars are nearly equal in size with a ratio of surface brightness 

 of 1:5.5. Ill this system the depth of the primary eclipse is 1.05 

 magnitudes and its duration a little over 6 hours. 



The last system, TX Cass., is the most interesting of the four 

 stars treated here. It was announced some years ago as being an 

 irregular variable; later its period was given by Zinner with the 

 note, that the period was probably changing. It was partly on ac- 

 count of these published notes that a thorough study of the light 

 curve of the star was carried out. Owing to the nature of the 

 variation the star has proved to be a difficult system for photometric 

 study. The eclipses, primary and secondary, which undergo a 

 loss of light of 0.54 and 0.33 magnitudes respectively and last over 

 18 hours, are difficult to observe, in fact it is impossible to obtain a 

 complete minim^um on any one night even during the long nights of 

 winter. The Harvard photogarphic measures have again proved 

 to be of extreme value and by combining them with the visual 

 observations I was able to establish a definitive peroid. The period 

 is 2 days, 22 hours, 14 min., and 41.7 sec. 



In the systems so far discussed the stellar disc was considered 

 of uniform surface brightness. Assuming this to be the case with 

 the system TX Cass, it was found that the observations could not 

 be represented at all satisfactorily ; the deviations were in many 

 cases three times the probable error. On the other hand assuming 

 the stellar discs to be similar to the sun bright at the center and 

 decreasing in brightness toward the edge, a very satisfactory repre- 

 sentation of the observations was found. The hypothesis of dark- 

 ened discs seems to be the correct one as it is confirmed by the 

 nature of the eclipses ; the secondary eclipse is total with a constant 

 phase of six hours while the primary eclipse has no constant phase 

 and the curve is distinctly round bottomed showing that the varia- 

 tion is continuous. This condition would exist with darkened discs 

 in case of an annular eclipse, and since the secondary is total our 

 natural and legitimate conclusion is that the primary is annular. 

 The system TX Cass, seems to offer very strong evidence in support 

 of darkening toward the limb in stellar systems. 



