OP ARTS AND SCIENCES. 393 



second and fourth comparisons cannot be reduced, since the light is 

 less than that adopted for the minimum. All of those preceding 

 eleven hours have also been bracketed, since the variation in light is 

 so small that an exceedingly small weight should be assigned to them. 

 Retaining them would not sensibly affect the result. The mean of the 

 remaining twelve gives for the time of minimum 9"^ 47"" INIoscow 

 mean time. The proximity of the minimum does not affect the resid- 

 uals of the last column. The last five are alone rejected, since, ow- 

 ing to the cause stated above, they indicate errors too large to be 

 accidental. The mean of the eighteen residuals retained is 0.06 

 magnitudes. 



The most complete series of naked-eye observations of this variable 

 are those of Dr. Schmidt of Athens. Five minima were observed 

 by him in August.* As all the comparisons were made after the 

 period of least light, he was obliged to wait for their reduction until 

 October 8, when he observed the star both before and after the min- 

 imum. The first reduction of these observations was made from % 

 curve derived from the measures of October 8. Later he has given a 

 discussion of thirteen minima,t from which he infers a rapid increase 

 in the period. In this paper he omits the observations of August 22, 

 although in his former paper he had assigned to it and to the minimum 

 of August 17 weights double those of any of the others. No reason 

 is given for this omission. There also seems to be a misprint iu line 12, 

 p. 89, of this same article. December 7 should apparently be Decem- 

 ber 2, as this date is employed below. Otherwise, an error of nineteen 

 minutes would be indicated in the observed minimum. A second re- 

 duction is given of the August observations, by which the time of 

 minimum is increased more than half an hour. As the original com- 

 parisons have not been published, it is impossible to rediscuss them. 

 As the star varies only a few hundredths of a magnitude during nearly 

 two hours, it is obvious that large differences may arise in the as- 

 sumed time when the light is least. Dr. Schmidt has also deter- 

 mined the period by a method free from this criticism. He has com- 

 pared the intervals between the times at which the variable equals 

 one of the comparison stars in brightness. Unfortunately, he has not 

 stated the times at which this occurs, so that a comj^arison with other 

 observers is not practicable. 



Mr. George Knott has also observed seven of the minima by the 

 method of Argelander.$ On September 23 and 28 the variable was 



* Astron. Nach., xcviii. 283. t Astron. Nach., xcix. 87. 



J Astron. Nach., xcix. 109. Nature, xxiii. 542. 



