r 



Magnetism in the United States. 5 



Yorkj eastward to Black Rock, and westward to Princeton, are 

 larger than tliose deduced, while '•^'^ar Cape May, in the lower 

 part of New Jersey and Delaware, they are less ; in the western 

 part of Massachusetts they are greater, in the eastern part they 

 are less. For the Avant of more ample material, especially of 

 observations in the interior, no attempt has at present been made 

 to represent these irregularities in the system of lines. They 

 are greatest in amount in the eastern and middle states, where 

 the average of the residuals, irrespective of sign, is 16', exclud- 

 ing the class first noticed. Along the shores of the Gulf of 

 Mexico it is 6' ; and on the western coast of the United States it 

 amounts to 10'. 



The residuals of the dip observations are less than those of 

 the declination, amounting in the average to 9', and there are a 

 less number of large disturbances. When we consider that the 

 disturbing polarities probably act nearly in the plane of the 

 horizon, we may conclude that the irregularities of the dip 

 should bear a still smaller proportion to those of the declination ; 

 and we see that the constant instrumental errors, which are 

 larger for the dip than for the declination, are not without sensi- 

 ble effect on the magnitude of the residuals. 



Upon the same consideration it will not be surprising that the 

 residuals of the horizontal force observations are larger in pro- 

 portion than those of the dip, being, in the mean, about the six- 

 tieth part of tlie actual values of the horizontal intensities. A 

 variation of 10^ in the dip would correspond to one of the one 

 hundred and twentieth part of the horizontal force. 



Comparison of Maps. — A comparison of the maps herewith 

 presented, with other similar ones that have heretofore been con- 

 structed, cannot fail to be interesting and instructive. 



Declination, — Allowing for the change in ten 3-ears, the lines 

 on Prof Loomis's map for 1840 {SilL Journ. vol. xl), agree w^ell 

 with the present map; considering the comparatively small 

 number and often unreliable character of observations they were 

 based upon, the agreement is remarkable, and leads to the hope 

 that valuable results may be derived from the recent observa- 

 tions made in connection with surveys of public lands in the 

 country west of the Mississippi, in which Burt's solar compass 

 has been used to a great extent. 



Colonel Sabine's chart of the declination in the Atlantic ocean 

 {Phil Trans. 1849), covers only the northeastern portion of our 

 map, which in that portion is based in part upon the same ob- 

 servations of 1811: and '45 to 1840, the epoch of the map, the 

 secular change used had not been considerably in error, as 

 already noticed in Mr. Schott's paper on the secular change of 

 the declination (d S. Report for 1855). As it is, the lines agree 

 well in direction, but occupy too nearly the same position for 

 the differences of epoch. 



% 



