200 Mr. W. Ellis. On the Simultaneity of 



graphical representation of the relation between the values of devia- 



tion and dip. 



Of iJiffj?/ /7/S-J7 



dtppi n a 



Magncti 

 Eauato 



dippin r/. 



There would seem to be here a tendency to negative values with 

 northerly dip, and to positive values with southerly dip. Remarking 

 how five stations (two with extreme northerly dip, two with extreme 

 southerly dip, and one intermediate) hang together, this apparent 

 tendency may be accidental, and yet there is an undoubted drift to 

 the left in the upper part of the diagram, and to the right in the 

 lower part A variation with latitude certainly appears to be one 

 better admitting of physical explanation than does a variation with 

 longitude. Can anything depend on the distribution of land and 

 water ? Whether the observed deviations represent a real inequality, 



