340 



cerned, the epoch of maximum velocity of the diurnal motion is also 

 an epoch of minimum disturbance. 



When we compare the results for the irregular disturbance, with 

 reference to the separate elements of magnetic declination and mag- 

 netic dip (see Ed. Trans., Vol. xix., Part 2), with the velocities of 

 motion as deduced from these figures, we find, that when the diurnal 

 motion is most rapid the departures from the direction of that mo- 

 tion are least, and when the diurnal motion is slowest the irregular 

 departures from, the hourly mean position are greatest. 



Thus, if we examine the mean disturbance of magnetic declina- 

 tion for each hour, as deduced from two years' observations, we find 

 it a maximum during the hours from 8 p.m. till 2 a.m. ; this is the 

 period for which the motion of the needle is at once slowest and least 

 as regards the declination ; about 21*^ (referring to the figure for the 

 year, see Plate VIII., Edin. Trans., Vol. xix.. Part 2), the motion 

 is most rapid and nearly altogether in declination, the minimum 

 disturbance in declination occurs immediately before this hour ; 

 another and nearly equal minimum occurs under the analogous cir- 

 cumstances about o^ P.M. ; a secondary maximum occurring about 



Jh oj. 2h P.M. 



If we approximate to the hourly mean disturbance of the magnetic 

 dip by means of those deduced for the two components of force, we 

 find the minimum to occur about G*"— 7*^ a.m., when the velocity of 

 motion is considerable, and when almost wholly in the direction of 

 dip ; the disturbance increases from that time till about 2*^ a.m., shew- 

 ing a secondary minimum about 1'^ p.m. and about 8'^ p.m., at both 

 of which times the direction of motion is chiefly that of dip : the 

 maximum disturbance occurs from about 10'^ p.m. till 3'' a.m., during 

 which period the velocity of motion is least. 



On the whole, then, the magnetic disturbance appears to be chiefly 

 at right angles to the direction of the motion of the needle, and to 

 be inversely as the velocity of motion. 



It is scarcely possible to connect the previous facts of area, 

 perimeter, or velocity of motion with the laws of variation of tem- 

 perature. In the mean for the whole year, the temperature changes 

 most rapidly between 8^^ and 9** a.m. ; but it changes with nearly 

 equal rapidity between S^ and 6"^ p.m. There is no corresponding 

 fact in the previous numbers. When we compare the variations of 

 temperature with the variations of position for the suspended mag- 



