ON COMPARING AND REDUCING MAGNETIC OBSERVATIONS. ye | 
to the solar diurnal variation for undisturbed observations, it should have 
been +3, the number +6 would be inserted in the table, and so on. 
20. It will be seen at once that we shall be able to ascertain by the 
method now described, if disturbance (as Broun supposed) alters the 
daily average values of the horizontal force. For in the horizontal force 
instrument any comparatively short period change of average daily value 
is hardly likely to be caused by instrumental alteration, but is most pro- 
bably due to magnetic causes, more especially if the same change takes 
place simultaneously at various stations. 
There are, however, more serious difficulties connected with the 
vertical force instrument, but into these I cannot now enter. 
II. By Sir J. Henry Lerroy, K.C.M.G., F.RB.S. 
1. The statement of the question appears to assume that the first, or 
chief, object of continuous automatic registers of magnetic changes is to 
extend the large number we already possess of mean determinations of 
solar-diurnal variations, and to add fresh numerical or quantitative values 
of the deviations from these means, produced by the causes we class as 
irregular. 
2. This appears to me to be persevering in a path we have been travel- 
ling for forty years without reaching, or even seeing the way, to any 
physical explanation of the phenomena. 
3. There are about seventy-five points on the globe at which the 
diurnal variation, including disturbances, has been determined by eye- 
observations, hourly or bi-hourly, with more or less completeness and 
precision. The irregular, or non-solar-diurnal, effects have as yet been 
eliminated for a few only (ten or twelve) of these points, but this nnmber 
has proved sufficient to bring out pretty clearly certain general laws to 
which no key has yet been found. 
4. Unless it can be shown that a multiplication of numerical data 
promises to bring us to a conclusion, I am inclined to think that the 
laborious compilation of more data of the same kind by measurements 
from photographic registers, which are less precise than the old eye- 
observations, is rather a misdirection of energy, unless indeed at stations 
widely remote from any others, and where new facts may be expected 
(see, for example, the very anomalous diurnal curve at Reikiavik, Iceland, 
‘ Athabasca volume,’ p. 297). The recent circumpolar stations would 
have come into this category if they had used self-recording instruments. 
5. Airy and Sabine have both taken +3'3 of declination as the 
measure of a disturbed observation at Greenwich and Kew respectively.! 
If it is true, as remarked by Professor Balfour Stewart (par. 5), that the 
precise measure is of no great consequence, is it worth while to spend 
much time over making out a new value independently for any part of 
Great Britain ? 
_ 6. The arbitrary nature of Sabine’s mode of treatment of observations 
is to me a strong objection to the continuance of it. 
For example, he threw out as disturbed all the observations at Point 
Barrow which deviated 22/87 from the normal,? and at Fort Carlton ? all 
which deviated 60. But I think I have sufficiently shown that in high 
latitudes in America the mean value of disturbance is about three times 
1 Phil. Trans. 1860-1863. 2 Phil. Trans. 1857. 3 St, Helena, vol. ii. 
