48 ASTRONOMICAL OBSERVATIONS 



latitude deduced from them is too small. In the later observations, the mer- 

 cury was removed about thirty inches from the object-glass of the telescope, 

 and the reflected observations were then found to accord quite as well as the 

 direct observations. The direct and reflected observations were made alter- 

 nately, from ten to sixteen at a culmination, and with the following result: 



Upper culmination, August 8, .... . 41° 14' 39"'8 



10, 36-7 



13, 36-8 



14, 37-8 



15, 40-8 



17, 36 -6 



Mean of these six culminations, 41 14 38 -1 



The latitude deduced from the upper culmination of 6 Ursae Minoris, Au- 

 gust 13th, is 41° 14' 35". 1; August 17th, 41° 14' 36".2. Mean latitude by S 

 Ursae Minoris, 41° 14' 35".7. Mean of observations on a ard S Ursae Minoris, 

 41° 14' 37".5. 



In order to determine the error of the readings of the microscopes, the fol- 

 lowing observations were made August 12th. The object was to ascertain if 

 five revolutions of the micrometer exactly measure the interval between two 

 divisions upon the limb. The numbers below give the excess of each micro- 

 scope, for a reading of five minutes, for the north polar distances contained in 

 the first column, being the points employed for the observations of Polaris. 



rth Polar Distance. 



A. 



B. 



c. 



Mean. 



358° 25'— 30' 



— l'-3 



+ l"-0 



-1- 2"-9 



-f- 0"-87 



1 30 —35 



— 1 -7 



-1- 1 -2 



+ 4-7 



+ 1 -40 



275 55 —60 



— 1 -9 



— 0-2 



+ 5-8 



-1- 1 -23 



279 0—5 



— 1 .3 



4- .4 



-1- 3 -8 



-f -97 



The numbers m the last column furnish the correction to be subtracted from 

 the micrometer reading when this amounts to five minutes. A proportional 

 part is to be taken for any other reading. This correction, although affecting 

 the latitude by only a fraction of a second, has, nevertheless, been applied to 

 all the observations. I assume, then, for the latitude of Hudson Observa- 

 tory, 41° 14' 37", and think that future observations cannot vary much from 

 this result. 



