OF ARTS AND SCIENCES. 



377 



TABLE III. — Rate of Change in Light. 



This table also affords a method of determining the point of mini- 

 mum light. At tliis point the rate of change should be zero, or should 

 change from positive to negative. This evidently occurs between the 

 times 50 and 25 minutes. Interpolating with the values given in 

 either the second or third column gives for the exact time 36 minutes. 

 This value agrees closely with 37 minutes, the value derived above from 

 the points of equal light. The best method of determining the time 

 of beginning and ending of the variation in light is from this same 

 table. It will necessarily be subject to considerable error, since the 

 observed curve must be extended according to the judgment of the 

 observer. The times — 300 and -j- 300 have been found in this 

 way. In other words, the star begins to diminish about 2G3 minutes 

 before the minimum, and does not recover its original brilliancy until 

 337 minutes after. The most rapid diminution would occur at — 140 

 or 100 minutes before the minimum. The variation would be then 

 0-.0064 per minute. 



The most rapid increase would occur at 100 minutes after the min- 

 imum, and would amount to 0.0055 magnitudes per minute. 



in Table IV. the values of R and a are arranged in groups. A 

 current number in the first column is followed in the second by the 

 condition determining the groups. The next columns give the number 

 of sets of twelve readings each, the arithmetical sum of the values of a, 

 the arithmetical sum of the residuals, and their algebraic sum, giving to 

 each the sign of i? divided by a. The seventh column gives the cor- 

 vection to the assumed minimum found by dividing the sixth column 

 by the fourth. The eighth column gives the probable error of the 



