AT VARIOUS PLACES IN OHIO AND MICHIGAN. 3 



over the same pole. The other half of the needle I apply, in a similar manner, 

 to the opposite pole of the magnet, and repeat the entire operation three times. 

 In each set of observations the poles of the needles are reversed, and the same 

 number of observations made in the two magnetic states of the needle. My 

 mode of observing is as follows : — Having brought the plane of the vertical 

 circle into the magnetic meridian by the method already explained, with the 

 face of the instrument to the east, and the marked side of the needle also to the 

 east, I read off the graduation at both extremities of the needle. I do not, ordi- 

 narily, wait for the needle to come to a state of entire rest, but when the arc of 

 vibration is reduced to ten or fifteen minutes; take the mean of the extreme 

 oscillations. Without disturbing the position of the instrument, I now vibrate 

 the needle, centring it, and at the same time checking its vibrations by the cop- 

 per y's. When the arc of vibration is sufficiently reduced, I read off again, as 

 before. I repeat the same operation five times, thus obtaining ten readings in 

 the same position of the instrument and needle. These readings are commonly 

 nearly identical. In two or three instances, however, the extreme readings 

 have differed from each other to the amount of about forty minutes. This 

 occasional sluggishness of the needle may, perhaps, be ascribed to moisture, or 

 minute particles of dust settling upon the axis of the needle and upon the 

 agate supports, and acting, by friction, to retain the needle at rest, though out 

 of the position it would naturally assume. Leaving still the instrument in its 

 first position, I turn the east side of the needle to the west, and make ten read- 

 ings as before. Turning, then, the face of the instrument to the west, I repeat 

 the observations in the same order, making forty readings in one magnetic state 

 of the needle. Reversing the poles, I repeat the entire operation, which gives 

 me eighty readings with one needle. The other needle furnishes the same 

 number of readings, making a hundred and sixty in all ; and this is the number 

 actually taken at each of the places mentioned below, with the exception of 

 Hudson, where the observations were still more numerous. The preceding- 

 method was adopted in all of the observations, with the exception of those made 

 at Hudson in September, 1838, where only two readings (one of each pole) 

 were made in each position of the needle, but the system of reversal was pre- 

 cisely the same as that above described. 



