272 PROCEEDINGS OF THE AMERICAN ACADEMY 



line drawu uearly through these points serves to convert the scale 

 of grades into photometric magnitudes, or into actual light ratios. 

 From this we may conclude that the maximum and minimum liglit 

 of the variable expressed in photometric magnitudes is 3.7 and 4.5. 

 The range is accordingly less than that ordinarily given, but this is 

 partly due to the dilTerence in the scales. The observations of Iler- 

 scliel expressed iu grades equal 9.6, G.l, 8.5, and 1.0. The first 

 three of these correspond to periods of 1*^ 19^, 1"^ 4'\ and l** 14*", after 

 a minimum, if the light of the star was increasing, or to 3*^ 4**, o'^ G'', 

 and 3** 14'', if the light was diminishing. The other observation, 

 No. 4, which was made six days later, on July 25, serves to decide 

 between these two hypotheses. Tiie light was then sensibly that of a 

 minimum at 1.2 grades. As the period of the star is about 7'' 4'\ 

 a minimum about a day preceding the observations 1, 2, and 3 would 

 also be indicated by observation 4. This would agree with the hy- 

 pothesis that the light was increasing on July 19, but would contro- 

 vert the view that it was diminishing. 



The time of the observations is defined only by the limits of twi- 

 light, which in the latitude of Slough would extend to within about 

 two hours of midnight iu July. The star would culminate near mid- 

 night, and could be easily observed as long as darkness lasted. The 

 times of observation may therefore be written, 1795, July 19'^ 12** ± 2*^, 

 and 1795, July 2o'^ Ti"* ± 2'\ The mean of the three results on July 

 19 would give an interval from the minimum of 1*^ 12*^, or would 

 place the preceding minimum at 1795, July 18"^ 0\ with an uncertainty 

 of several hours, since a small error in the light corresponds to a large 

 deviation iu the time of minimum. The elements of Argelander indi- 

 cate a minimum 1795, July 18*^ 19'> 42™ Paris M. T. 



g HercuUs. — The variations in light of this star are irregular, or 

 the law governing them has not yet been discovered. Should this law 

 ever be determined, these observations may have great value, since they 

 anticipate by sixty-two years the first observations of equal accuracy 

 previously known. Probably the variations are so slow that the hour 

 at which the observations were made will not be needed. 



u Ilerculis. — The same remark applies to this star as to the preced- 

 ing. The observations are accordant, and anticipate other similar 

 observations b}' seventy-four years. 



o Ceti. — Some other observations by Sir William llerschel are 

 given by Argelander.* As this star had been observed for many 

 years previously, these observations are not of especial importance. 



* Bonn Beob.. vii. 329. 



