OF ARTS AND SCIENCES. 397 



All the observations are fairly represented by formula (C), except 

 the second reduction of those of Dr. Schmidt. His original reduction 

 leaves a residual which might well be due to errors of observation. 

 Although the residuals of the Harvard College measures are small, 

 they are still much larger than their probable errors, and their values 

 evidently indicate systematic error. The last formula satisfies these 

 completely, giving average residuals of only 0.3 minutes, but does not 

 agree with the other observations. If we admit a vai-iatiou in tlie 

 period, the value 2'* 11*^ 49.9'" would seem to be that between E = 90 

 and E = 114, but not between E = and E = 90. 



It is scarcely worth while at present to discuss the relative probability 

 of these various formulas, since further observations which will doubt- 

 less soon be made will serve to decide between them with certainty. 

 If the original observations were published, so that all could be re- 

 duced according to the same method, doubtless much greater accord- 

 ance would be found in the results. There seems to be no reason 

 why the error in determining each minimum from observation during 

 the decrease and increase should not be reduced to two or three min- 

 utes or much less than those of /3 Persei. Should the discrepancy of 

 certain measures, as those in August of Dr. Schmidt, be confirmed, 

 they would indicate the existence of some disturbing body which 

 might also account for such a deviation as that noted in the minimum 

 of fi Persei on Nov. 22. No correction has been applied for the 

 a'berration. The star is so near the pole of the ecliptic that the 

 correction would never exceed two minutes, and would be masked by 

 the other errors. 



Harvard College Observatory, 

 Cambridge, Mass. 



