89 
The times of minima appear to be more irregular than 
those of maxima ; but the magnitudes at maximum vary 
more than those at minimum. As the times of maxima are 
evidently better suited for the determination of the mean 
period than those of minima, I have treated them in the 
usual way, and have obtained the following elements : — 
Period =152*4 days. 
Epoch = 1865, January 24*1. 
Calculating the times of maxima from these elements, and 
comparing them with the observed times, we have the follow- 
ing differences : — 
0 .— 0 . 
days 
- 0*1 
+ 1*7 
-4*5 
+ 2*9 
The greatest difference is only 4*5 days, or about one 
thirty-fourth of the entire period, and the mean difference is 
2*3 days, or one sixty-sixth of the whole period. T Aquilse 
therefore belongs to the class of variables whose periods are 
moderately regular. 
The diagram which accompanies this communication 
shows the mean form of the light-curve of T Aquilse as 
derived from all the observations I have yet made. An 
examination of this curve has yielded the following results : — ■ 
Mean magnitude at maximum = 9*20 
Mean magnitude at minimum —11 *20 
Mean range of variation =•• 2*00 magnitudes 
Mean magnitude = 10*40 
Interval from minimum to maximum = 64*0 days 
Interval from maximum to minimum =88*4 days 
Interval from minimum to mean magnitude... = 43*0 days 
Interval from mean magnitude to maximum... = 21*0 days 
