108 



EIGHTH REPORT — 1838. 



the error may be regarded as nearly constant*. Defective, 

 therefore, as the apparatus is in this respect, there is reason to 

 conclude that the differences of dip obtained with it in Ireland 

 may be relied on within the usual limits of probable error, and 

 that to obtain the true dip from the observed results, we have 

 only to apply a positive correction, which may be regarded as 

 constant throughout the series. 



The instrument referred to in the preceding pages having 

 been much employed in Dublin, and with very consistent re- 

 sults, we shall take, as the basis of its correction, the dip in 

 Dublin as deduced from the observations with Gambey's nee- 

 dles, Table XXXII. In these observations, made according 

 to the method of arbitrary azimuths, the bearing points of the 

 axle^ and the position of the needle with respect to the limb, 

 are different in each azimuth ; so that the results may be re- 

 garded as, virtually, the results of different instruments. 

 Their accordance is sufficient to show that the errors of axle 

 and of limb are inconsiderable. For the convenience of refe- 

 rence, the observations are put together in the following Table ; 

 the dips being reduced to the 1 st of January, 1 838. 



Table XXXIV. 



The mean of these results is 70° 57''9. If we combine with 

 this the mean result obtained by Captain Ross at the same 

 place, as deduced from six observations with four needles, and 

 reduced to the same epoch, (namely, 71° 1''7,) we have, for 

 the mean dip in Dublin, on the 1st of January, 1838, 



70° 58'-8. 



* A comparison of the results with those of other instruments seems to point 

 to the conclusion that this error diminishes with the dip, and is somewhat less 

 in England than in Ireland. 



