AND THEIR CONNECTION WITH MAGNETIC CHANGES. 367 



A cursory glance at Tables VII., VIII., IX. is sufficient to show us that the 

 values of the departures for Class 2 are greater than those for Class 1 ; and that 

 in like manner those for Class 3 are greater than those for Class 2. But since it 

 may he objected that the means in Tables IV., V., VI., are not in all cases 

 thoroughly determined, I may likewise remark, that the range of the departures 

 follows the same rule as their absolute values. Indeed, the great characteristic 

 feature of the disturbing force is the large range which it causes between the 

 different values of earth-currents observed on the same line, and during the same 

 hour of the day, while at the same time the departures are perhaps as often to 

 the one side as to the other of the normal values, as far as this may be gleaned 

 from a somewhat limited number of observations. 



That this peculiar action of disturbances on earth-currents does not depend 

 upon the absolute amount of the disturbing force, magnetically measured, will, I 

 think, be rendered probable from the following considerations : — 



1st, In the two cases in which the declination needle was considerably de- 

 flected from its normal position, there is no evidence from the corresponding 

 earth-currents to show that these depend upon the absolute amount of the mag- 

 netic deflection. 



2d, For all the other disturbed observations of earth-currents, neither magnet 

 was far from its normal position. 



While, however, there is no evidence in favour of the theory which ascribes 

 magnetic disturbances to the direct action of earth-currents, there is much to 

 favour the induction hypothesis in the great range towards either side of their 

 normals apparent in the values of those earth-currents which correspond in 

 point of time with magnetic disturbances. For it is evident that such a range 

 from side to side would be produced if earth-currents were induced currents, due 

 to those small but rapid changes in the magnetism of the earth, which are repre- 

 sented by the peaks and hollows of the disturbance curves. 



Another remarkable feature of Tables VII., VIII., IX., is the tendency of 

 the undisturbed observations in juxta-position with one another, to exhibit 

 departures affected with the same sign ; and if we reflect that the order given 

 in these tables is that of time, we shall, I think, be disposed to conclude with Mr 

 Walker, that meteorological conditions in all probability influence the values of 

 the currents observed. 



If this be the case, there would thus appear to be three independent pheno- 

 mena which affect these values : — 



1st, The daily magnetic change. 



2d, Magnetic disturbances. 



3d, Meteorological conditions, probably those which alter the conducting power 

 of the superficial strata of the earth's surface. 



If we reflect on this, and bear in mind that we have as yet obtained only a 



VOL. XXIII. PART II. 6 H 



