94 



Dip and Variation of the Magnetic Needle. 



amount we have no means of estimating. In order to obtain 

 more light upon this subject, I addressed a letter to Prof. Farrar, 

 of Cambridge. In his answer he remarks : '•' I have every reason 

 to believe that the only dipping needle belonging to the apparatus 

 of the college, at the time of Prof Williams' observations, was 

 one made by Nairne, and is not, I think, to be depended upon. 

 When I have attempted to use it, I have found the needle to rest 

 in positions differing several degrees from each other." This last 

 remark corresponds very well with the observations of Prof. Wil- 

 liams. If the original manuscript containing the observations 

 can be found, it is possible that the suspicion which now attaches 

 to them may be removed. At present, however, I place no con- 

 fidence in them. 



The next dip observations in point of age, are those made du- 

 ring Long's expedition to the Rocky Mountains in 1819. The 

 dip at St. Louis, June 16, 1819, is given 70° 30'. I found the 

 dip, Sept. 29, 1841, to be 69° 25'.5, shewing a decrease of about 

 three minutes per year. The dip^at Shippingsport is given 70° 15'. 

 Prof. Locke found the dip at Louisville in 1840, 70° 4', shewing 

 a diminution of half a minute per year. The discordance of 

 these results excites a suspicion of inaccuracy somewhere. A 

 more satisfactory result may be expected by combining all the 

 observations made in 1819, so as to obtain the mean dip at some 

 one station. By this method we may be expected to eliminate 

 in a measure the influence of local attraction, and also accidental 

 errors of observation. The following table presents a summary 

 of these observations, treated in the manner described in a former 

 number of this Journal, Vol. xl, p. 85. As the observations are 

 all comprised within a moderate period, I have not applied any 

 correction to reduce them all to a common instant. Adopting for 

 the central position latitude 39° 3' N., longitude 91° 37' W., we 

 obtain the annexed equations of condition. 



