THE DISTURBANCE OF ISOMAGNETICS. 



169 



As we suspected that the change in the vahie of n might be due 

 to the change in the tempering of the magnet, caused probably by the 

 annual variations of temperature during the years intervenino-, we 

 made experiments on a similar magnet newly made, lîefore applying 

 this correction to the previous results. The bar was tempered as hard 

 as we could make it, and the experiment was performed both in the 

 unmagnetized and in the magnetized state. In l^oth cases the value 

 came out in slight excess of the values of the old magnet, contrary to the 

 above supposition ; we therefore apply the correction to our previous 

 results. The values of li for 1887 recalculated by applying all the 

 necessary corrections are found in the appendix. 



Zy'if. — This was also the instrument that was used in 1887, i.e., 

 iSTo. 88 Dover. Needle No. 2 was always used. 



Results of Ihe Survey. 



The individual determinations of the mai]fnetic elements are 

 given in full in the table appended to this paper. We here compare 

 the mean values with tliose of the previous survey. 



DIP 



