LL ne 
AND INSTRUMENTS AT MUNICH. 515 
are required; for this purpose their relative position may be 
most conveniently arranged as follows :— 
3 
— 
= 
_—- 
laa 
i 
i 
SPE SSS) 
—s 
~ 
— 
ns 
The telescope and scale are at F. The telescope has a vertical 
movement, so that it may be directed to the two differential 
instruments which are placed perpendicularly one above an- 
other; that for the declination at D, that for the intensity at I, 
m and m! are the marks of reference for the telescope. The 
influence of the instruments D and I upon each other, though 
not considerable, must be taken into the calculation. The two 
instruments can thus be observed with only two or three seconds’ 
interval. In ordinary cases such readings may be regarded as 
simultaneous ; where greater precision is required, D, I, and D, 
may be read successively at equal intervals, and the mean of the 
two readings of D will be simultaneous with the reading of I. 
I propose to take a future opportunity of describing a con- 
struction and arrangement better adapted for dwelling-houses 
in particular. 
A careful consideration of the above construction will show 
that, in consequence of the symmetrical position of the deflecting 
magnets, and their perpendicular direction relatively to the sus- 
pended magnet, no change which can occur in the adjustment 
of the instrument (supposing it not to be very considerable) will 
produce error in the reading: a check therefore is unnecessary. 
The formula shows that the readings ‘of this instrument re- 
quire a correction on account of temperature, partly because the 
magnetism of the deflectors is affected by temperature, and 
partly because their distance from the suspended magnet is 
altered by the expansion of the brass bar. 
It appears to have been latterly supposed that the influence 
of temperature on magnetized steel bars could not be brought 
under calculation. However, in constructing my new magnetic 
apparatus, I have made a series of experiments which agree with 
previous experiments of M. Kupffer in showing, that in all tem- 
peratures which are met with in the open air, and even in some 
considerably higher, the change in the magnetism of the bar is 
