MADE IN THE UNITED STATES. 317 



4. Turning again the face of the vertical circle to the east, (leaving the 

 needle as in the preceding position,) and reading off six times as before, it 

 makes twenty-four readings for the first magnetic state of the needle, which I 

 designate by the words poles direct. 



A second magnetic state of the needle is required for correcting the errors 

 resulting from the centre of gravity being out of the axis of the needle. To 

 that eifect, the poles are reversed by the action of two magnetic bars of 

 nearly seven inches in length, and the entire observation of the dip repeated 

 in the same order as to the four positions described above, which gives twenty- 

 four readings, the result of which I call poles reversed. The second needle 

 furnishes the same number of readings, making ninety-six readings in all. 

 This is the number actually taken at each of the places recorded in this paper, 

 with the exception of No. II. and III. where only two readings, one of each 

 pole, were made in each position of the instrument and needle. 



The instrument was placed upon a very convenient tripod, to which was 

 fixed a square table, with an azimuthal instrument to bring the vertical circle 

 in the magnetic meridian when on the zero of the vertical circle; and the 

 whole being well adjusted, stood firmly and steadily during the observation, as 

 proved by the spirit level, which very seldom required to be corrected. 



I was induced to adopt the preceding method by analogy with what is prac- 

 tised in taking altitudes with an astronomical circle, in which case, to com- 

 plete an observation, it is required to take one altitude with the face of the 

 instrument on one side, and then a conjugate observation on the opposite side. 

 I was not aware at that time that the mode of observing the dip I have fol- 

 lowed had undergone a slight modification on the part of many observers, who, 

 instead of turning the vertical circle 180°, to bring it to the second position, 

 prefer leaving it in its original position, and change only the marked side of 

 the needle in its brass Y's, with the view, I suppose of saving one turning of 

 the face of the vertical circle, in the observation of any one of the magnetic 

 states of the needle. But there can be no objection to the mode I have fol- 

 lowed, for, if for fear of disturbing the instrument it be thought proper to save 

 one turning of it, this is only avoiding one cause of disturbance to introduce 

 another, which is perhaps as great, since it compels them to open twice the 

 glass door, to change the face of the needle, with a manipulation, it would 

 seem, more likely to bring on a disturbance, than the simple and smooth 



