ON COMPARING AND REDUCING MAGNETIC OBSERVATIONS. 83 
X. Letter from George M. Whipple, Esq., to Professor Stewart. 
Kew Observatory : 
July 29, 1885. 
Dear Prof. Stewart,—I have carefully read the paper you were so 
good as to forward to me, ‘ Suggestions for the Committee on Magnetical 
Reductions,’ and must confess that I am in most points fully in accord- 
ance with Sir H. Lefroy. 
I would much rather trust to the solution of the various problems of 
Terrestrial Magnetism by a farther and more extended series of com- 
parison of curves than by an extension of numerical processes. 
The reduction of the Fort Rae observation shows how enormously 
large and frequent the variations may be in some parts of the earth; 
and such being the case, I fail to see how any useful purpose could be 
served by the repetition of the calculations of Gauss. 
I think that magneticians should endeavour, if possible, to enter into 
communication with geologists and seismologists, and endeavour to trace 
out clearly the causes of (what I would term) superficial variations, pro- 
bably due, Prof. Schuster says, to electric currents, for localities well 
furnished with magnetic observatories, such as Europe, rather than to 
attempt at once to solve the whole problem of distribution throughout the 
earth of magnetic matter. Tam, yours faithfully, 
G. M. Wuiprtz, Superintendent. 
P.S.—I enclose also copy of some remarks addressed by Capt. Dawson 
and myself to the Vienna Congress on the subject. 
Further and additional remarks on the questions to be submitted to the 
Vienna International Polar Conference. 
We are of opinion that careful inspection of the observations them- 
selves will suffice to show the days and hours when the diurnal curve 
follows its normal course. From days and hours selected by this inspec- 
tion, mean curves may be obtained, and ultimately by interpolation a 
Series of hourly values may be arrived at for every day in the year. 
Readings differing from these values by more than a certain separat- 
ing value should be set aside and discussed as disturbances. It appears 
to us probable that the principle of determining the mean monthly diurnal 
curves for each station from observations selected only on such days as 
are shown by evidence of magnetographs elsewhere to have been mag- 
netically calm, assumes beforehand a uniformity of magnetic conditions 
_ over the globe, and might, therefore, fail at certain stations. A rough | 
comparison of Fort Rae and Kew Observatory results indicates to us that 
itis rather more advisable to deal with hours and not with days as a 
whole, and that therefore some rule, either Sabine’s or Lloyd’s, must of 
: 
: 
: 
. 
necessity be adopted. 
There seems no objection to the application, first, of Lloyd’s rule to 
throw out disturbances, and then to the subsequent classification of these 
disturbances after the method suggested by Wild. 
We fail to see as yet any method of introducing possible corrections 
for sun-spot periodicity into observations made during so short an inter- 
val of time at stations where no previous observations have been taken ; 
and therefore recommend that this disturbing element be omitted entirely 
G2 
