ON COMPARING AND REDUCING MAGNETIC OBSERVATIONS. 79 



south to north in the northern, and from north to south in the southern 

 hemisphere, and producing about- 2 p.m. a maximum westerly deflection 

 of the north-seeking pole of the magnet in the northern hemisphere, 

 while that in the southern hemisphere for the same pole is of an opposite 

 character, being easterly. 



4. Again we know that the air in the upper atmospheric regions 

 travels from the equator to the poles, forming (in those regions) a south- 

 west current in the northern hemisphere, and a north-west current in 

 the southern hemisphere, these being, in fact, the well-known anti-trades. 

 If, then, owing to their passage across the earth's lines of force, these 

 moving conductors are animated by electric currents, these currents 

 mast, according to the well-known law, be in such a direction as to stop 

 the atmospheric motions.' 



5. For the purpose of the following argument we may without sensible 

 error imagine the magnetic eartli to be really similar to the model that 

 is sometimes used to represent it ; that is to say, we may regard it as a 

 globe wrapped round continuously with insulated wires all in the same 

 direction, and conveying a current, the circles of these wires being small 

 near the jsoles, and, of course, large at the equator. If we should take a 

 bird's-eye view of this system from above the point which I'epresents 

 the north pole (which corresponds approximately to the south pole of a 

 magnet), we should find that the positive current in the wires would 

 circulate in the direction of the hands of a watch, ascendino- on the east 

 and descending on the west side. Such hypothetical currents may, there- 

 foi-e, be imagined to move along the earth's surface from east to west. 



6. Let us now take the upper systems of atmospheric currents and 

 consider that element of their motion which is from west to east. This 

 motion is common to both systems. If in the northern hemisphere these 

 upper winds be animated by a positive electric curi-ent goino- north, 

 this current will be attracted by the hypothetical magnetic current on 

 the west side and repelled on the east ; that is to say, there will be a 

 tendency to stop the easterly motion of the atmospheric current. In the 

 same way it may be shown that if in the southern hemisphere the upper 

 winds be animated by a positive current going south, this will tend to 

 stop the easterly motion of the atmospheric cui-rent. 



7. It thus appears that the electric currents with which, accordino- to 

 this argument, the upper trade-winds in the two hemispheres ouo-ht 

 to be animated are precisely such as will account for the solar-diurnal 

 variation of declination, this being alike the most prominent and the most 

 simple feature of the solar-diurnal variations. 



8. While it is advocated that the provisional working hypothesis thus 

 accounts, as far as direction is concerned, for the positive currents going 

 north and south — which are presumed to be the main causes of the 

 diurnal variation of declination — it is also necessary to remark that such 

 currents will naturally present a decided diurnal fluctuation. Indeed, if 

 this were not the case, they could not properly account for a variation 

 one marked feature of which is its prominence during the day as dis- 

 tinguished from the night hours. Now we may conceive that the upper 

 atmospheric currents maij be stronger, and will at any rate be better 



' Sir J. Henry Lefroj' lias long been of opinion that the key to the mao-netic 

 movements in both liemispheres is to be found by studying the simultaneous effects 

 in both produced bv the action of the sun on the equator (see p 18'> of his 

 'Survey,' 1851). 



