L. A, Bauer— Variation of Terrestrial Magnetism. 321 
These results show how useless it is to endeavor to determine 
the secular variation period from the supposed motion of the 
agonic lines around the earth, as some eminent investigators 
have done. Not only will the result depend upon the particu- 
lar agonic considered, but also upon the parallel of latitude 
along which the motion is supposed to take place. We are 
seemingly forced again to a conclusion which I have already 
formulated, viz., that we have no evidence as yet that the 
earth possesses a common secular variation period, it being 
questionable whether there really is a period at the close of 
which the same path will be traversed by a free magnetic 
needle. If the earth does possess a common period, then the 
curves at present being described at various stations are but 
branches or loops of the complete curve. If this period is as 
long as 2000 years, as above indicated, abundant opportunity 
is furnished for regarding the partial periods at the various 
stations as submultiples of the total period. 
It should be noted that the period of 2000 years deduced 
from the average motion of the agonics is identical with that 
which we obtain if we suppose that the secondary magnetic 
system revolves around the equator at the average rate of the 
two secondary poles, viz: 3£9, = 1860 years. That is, the 
average annual westwardly motion in longitude along the 
equator is about the same (nearly 0°-2) for the secondary mag- 
netic poles and the agonics. 
IT am inclined to think that the differences in the annual rate 
of motion as exhibited by the two secondary poles and, like- 
wise, by the two agonics is real—parts of the same magnetic 
system may move with different velocities, the rate of motion 
of the particular part being governed by the portion of the 
earth over which it is passing. We must then be careful not 
to suppose that because a certain phase has traversed a section 
of the earth in a certain time, its rate of motion will continue 
the same over another section. In this way there arises a 
warping, a distortion or overlapping of our magnetic system. 
A number of illustrations could be given as to how we can 
read the broad facts of the secular variation from the positions 
and the motions of the secondary magnetic poles. I shall give 
only two, however, in the present number. 
First. Where should we expect the largest and the smallest 
secular change in declination during the interval 1780-1885 ? 
The largest change in declination would be expected where the . 
change in the horizontal component of the secondary system 
is a maximum, i. e., in the region traversed by that branch of 
the secondary magnetic equator corresponding to the shorter 
distance in longitude of the secondary poles, hence over the 
eastern Atlantic and adjacent countries. 
