of Terrestrial Magnetism. 115 



There is a second very notable geographical distinction between 

 the phenomena of the solar-diurnal and of the disturbance-varia- 

 tions, and it is one which bears strongly on the probability of a 

 difference in the mode of the sun's action in producing the two 

 classes of effects. It consists in the very different relative mag- 

 nitudes of the range of the two variations in different parts of 

 the same hemisphere. In the case of the disturbance-variation, 

 we have already noticed that its amount is much less in the 

 middle latitudes of Europe and Asia than in Canada, while in 

 certain parts of the north-western portions of the American 

 Continent it attains a development which renders the dispro- 

 portion excessive. We have nothing analogous to this in the 

 solar-diurnal variation, which exhibits in different meridians a 

 remarkable approach to constancy in its amount (varied only by 

 distance from the dividing line between the two magnetic hemi- 

 spheres, and by differences in the antagonistic horizontal force of 

 the earth), as well as in its form and turning-hours — a constancy 

 which contrasts strongly and systematically with the phenomena 

 of the disturbance-variation, and is apparently quite uninfluenced 

 by any peculiarities of land and sea, near or remote. 



I have yet another remarkable distinction to bring before you, 

 which seems to indicate a difference in the mode of operation 

 in the production by the sun of the mean solar-diurnal variation 

 of the entire year, and of the semiannual inequality previously 

 described to you. In the latter, if you will look again at Plate 

 II., in which it is represented in various latitudes and in both 

 hemispheres, you will see everywhere a striking uniformity; 

 the amount is nearly the same at all the stations, and the same 

 end of the needle is deflected in the same direction at the same 

 local hours in both hemispheres. In the mean annual solar- 

 diurnal variation, on the other hand, the same end of the needle 

 is deflected in opposite directions at the same local hour in the 

 two hemispheres ; the variation itself disappears altogether as 

 the line of separation between the two magnetic hemispheres is 



north end of the magnet being directed towards the north, the solar-dim'nal 

 variation is at 8 a.m. to the east, and at 2 p.m. to the west, both magneti- 

 cally and geographically. The arctic voyages and travels have shown that 

 between these extreme cases of dissimilarity every possible intermediate 

 variety may be found. Thus, whilst the deflection caused by the solar- 

 diurnal variation is always in the same magnetic direction at the same local 

 solar hour, its geographical direction over a very considerable area around 

 the dip of 90° has every conceivable variety. We must regard, therefore, 

 the causal action which produces the solar-diurnal variation as operating 

 through the intermediation of the earth's magnetism, and not simply as a 

 direct action upon the needle itself. In these remarks on the solar-diurnal 

 variation, it is always to be understood that the effects of the disturbances 

 have been either wholly or for the most part eliminated. 



12 



