MAGNETIC DECLINATION NEAR THE MAGNETIC EQUATOR. 689 



declination in high latitudes, having a greater range in summer than in winter. 

 A similar operation then performed on the curves of temperature would produce 

 a similar result ; but, as we know, the temperature curve is not inverted in the 

 southern hemisphere, the maximum temperature occurring somewhat after noon 

 everywhere. 



Though this process of subtracting the mean curve from its two parts cannot 

 have any marked value in the consideration of the phenomenon of the diurnal 

 variation of magnetic declination, yet I desire to point out that the subtraction 

 of curves, or of their equivalent equations, from each other, indicated by me in my 

 original note already cited, may lead in other ways to definite ideas of the incre- 

 ments of movement (the curval increments, if I may so term them), which are 

 due in one place to a change of position of the sun, or which are due for the same 

 position of the sun to a change of latitude. These operations, which have been 

 already performed by me for different places, may be of importance with reference 

 to the comprehension, not only of the rapid changes of law near the equator, but 

 also of the changes due to an approach to the magnetic poles. 



Ingenious attempts have been made to connect the diurnal variations of mag- 

 netic declination with the electrical currents observed by means of telegraph 

 wires. These currents are, I believe, quite local ; it is extremely probable that they 

 produce variations of a second order : others of the first order are due to the 

 polar currents, connected with the polar lights ; but these are superposed on the 

 regular solar variations. We have, I think, only to look at the group of curves 

 for the months of May to August, to see the fallacy or the insufficiency of all the 

 hypotheses proposed as yet to explain these variations. We have, it appears to 

 me, to deal with a general (in opposition to local), if not a cosmical cause. The 

 position of the station as regards coasts, centre of continents, or as to height, 

 does not seem to produce any marked variation in the regularity of the law and 

 its change from north to south, in the months referred to. 



I have already noticed the great effect of the lunar action, sometimes exceed- 

 ing that of the sun, and have offered a suggestion, which might lead to a theory. I 

 would only add, at present, that since writing the note, page 680, it has occurred to 

 me that I endeavoured to show in 1850, in the General Results of the Makerstoun 

 Observations (I am not able to cite the page at present), that the aurora borealis 

 was most frequent near the epoch of full moon. Though the number of 

 aurorae, from which this conclusion was derived, was too limited to give it great 

 weight, yet, as the aurorse were sought for on the appearance of the slightest 

 irregularity in the movement of the magnets (a sure premonition) it merited a 

 greater weight than one deduced from masses of observations by different 

 observers, made without system, and noted chiefly when most easily seen — that 

 is, when the moon does not shine. I believe I also pointed out the occurrence of 



VOL. XXIV. PART III. 9 A 



