Dip and Variation of the Magnetic Needle. 



103 



southerly — 1'.6, at the ten most easterly — 0'.2, at the ten most 

 westerly +0''.8, at the nineteen intermediate — O'.S. 



The preceding comparisons furnish us with the following re- 

 sults, where column third shows the dip, column fourth the direc- 

 tion of the isoclinal line, and column fifth the increase of dip in 

 the normal direction for sixty geographical miles at the points 

 whose latitudes and longitudes are shown in the first and second 

 cokuTins. 



With these data, we may 

 isoclinal lines of 72° and 73°. 

 lowing points. 



trace with some confidence the 

 They will pass through the fol- 



The isoclinal lines are curved lines passing through the above 

 named points and in the directions here given. In the eastern 

 part of the United States this curvature is very gradual and uni- 

 form, but near the longitude of 83° 30' there is a point of inflex- 

 ion. The isoclinal lines become concave towards the south. 

 This interruption of the general regularity soon ceases, and in 

 the neighborhood of the Mississippi the isoclinal lines are tolera- 

 bly well represented by portions of the same circular arcs as near 

 the Atlantic. The position here described differs but slightly 

 from that represented on my chart, Vol. 39, p. 41. Other isoclinal 

 lines may be drawn on each side of these from the same data, but 

 with somewhat greater uncertainty. How now are we to regard 

 the differences contained in the last column of the preceding ta- 

 bles ? Do they indicate errors of observation,* errors of the hy- 



* I much regret that Prof Locke should have interpreted a former paper of mine 

 (Vol. XL, p. 85) as a ' special effort to discredit his observations.' Such was far 

 from my intention. I had always regarded Prof Locke's observations as exceed- 

 ingly valuable, especially as they were made in a region hitherto wholly unex- 



