288 



PROCEEDINGS OF THE AMERICAN ACADEMY 

 TABLE VIII.— Summary op Series. 



be applied to measures made in the corresponding position of the in- 

 strument. The position of the instrument actually employed was Op, 

 except in the observations to which Nos. 20 to 31 relate, when it was 

 0„. The corresponding excess is given in the next column. The 

 last two columns give the values of the collimation and level error, 

 in seconds of arc. 



During SejDtember the level was used in some geodetic observations 

 among the mountains of New Hampshire and Vermont. Nos. 20 to 

 25 were taken from the top of Mt. Moosilauk, and Nos. 26 to 31 

 from Mt. Mansfield. Although the instrument was carried in a wagon 

 over rough mountain roads, the effect on its constants seemed to be 

 inappreciable. The level error, as shown in the last column, did 

 not appear to change perceptibly during the entire series. The colli- 

 mation at first underwent a singular change which may have been due 

 to a looseness of the screws holding the reticule. No change appears 

 to have taken place during a single series of observations, since the 

 collimation was always substantially the same before and after it, that 

 is, in the pairs of measures made upon the same date. The effect is 

 therefore eliminated in the final results. No change appears to have 

 taken place after September 3. 



Nos. 11, 15, 16, 19, 32, 33, 34, 35, 36, and 37 relate to the needle 

 used to observe a Bootis ; in the other cases the intersection of the 

 cross-wires was observed. The interval between them was about 85 

 divisions, which affects the collimation, but not the level, to this extent. 



In No. 10 the reading of E„ has been assumed to be in error by one 

 turn of the screw ; otherwise, the observed value would be —46.7. 



