72 REPORT—1 885, 
as great in the early morning hours as it is in the afternoon. Conse- 
quently we must either disregard a great many observations by day, which 
are really disturbed, or include a great many by night, which are not, 
unless we say that instability is the same thing as disturbance. 
7. What, then, is to be done with the photographic registers? How 
can they be compared unless by ordinates, measured at points agreed 
upon, such as the Géttingen hours ? 
Ireply (1) that I think that each observer should minutely scan his own 
records, and note the time, direction, and amount of movements. (2) That 
the efforts of magneticians should be addressed to the cheap publication 
and prompt interchange of the registers of each week, reproduced and 
reduced by photography to a uniform scale, say 15mm. to 1 hour, with a 
view to the discovery of periodically recurring movements of whatever 
nature; of movements apparently local, or not generally traceable ; and 
of movements which were general, in one or more elements, over a large 
part of the earth’s surface. 
It hardly meets this suggestion to say that we have hundreds of 
projections of disturbances already, and that nothing has come of it. It 
is true; but these projections are scattered through many volumes, are 
upon all sorts of scales, and are rarely comparative. 
8. The student having by his eye-comparison grasped the general 
features of the movements constituting disturbance at some particular 
epoch, or presenting an exceptional character, the need of measurements 
would arise, and if a reference to the mean of the day or the mean of n 
days or of the calendar month is necessary, such mean can be ascertained. 
I am not sure that it often will be, and I doubt whether our adherence to 
the calendar month is rational. Why should movements on May 31 be 
referred to the mean for May rather than the mean for June? The more 
accurate, though more laborious, plan would be to refer them to the mean 
for May 31 + 10 days. 
9. The end of the needle which points to the equatorial region has in 
every locality a mean position in relation to the meridian from which it 
is continually deviating, and to which it always returns. It appears to 
me open to question whether the relation of the direction of the move- 
ment to the absolute declination, as increasing it or diminishing it, has — 
much to do with the question. At least it seems to assume that the 
normal position is due to the same physical causes as produce the devia- 
tions, and therefore I think that the deviations, whether of the polar or 
equatorial end, should be simply noted as east or west without regard to 
sign. In the southern hemisphere it is the equatorial end that we observe. 
Regions where the north end is actually directed to the south, as at Port 
Kennedy and the Alert’s winter quarters (1875-6), will require negative 
signs, 
10. It seems probable (1) that the mean position of the needle above 
referred to is always perpendicular to the direction of electric currents in 
the crust of the earth, or the atmosphere, or both, originating in a thermo- 
electric action of the sun on the meridian, and propagated north and south 
from the ecliptic; (2) that the position of the meridian of the place, in 
reference to the sun, determines the direction of the mean deviation of 
the needle from its normal position or the mean solar-diurnal movement, 
and that the amount is determined by a balance of forces still to be clearly 
defined. The amount is known at a sufficient number of stations to test 
any law laid down. 
