516 DR. LAMONT ON THE MAGNETIC OBSERVATORY 
perfectly proportioned to the change of temperature, and is in- 
dependent of the strength of the bar’s magnetism. We also 
found, that when a magnetic bar has lost a certain portion of its 
force it attains a steady condition, fitting it for observations on 
terrestrial magnetism. Magnets differ widely in their sensibility 
to heat, and whereas a magnet will lose under ordinary circum- 
stances 00012, and in more favourable cases 0°0006 of its mag- 
netism for an increase of temperature of 1° Reaumur, I have 
succeeded in producing magnets in which the loss is only 
0:00014. However, I do not regard these experiments as con- 
cluded, the more so as I find the mention, in the magnetic works 
of Col. Sabine, of a small bar used in experiments of vibration 
which underwent no sensible alteration either in increasing or 
decreasing temperatures. Without expecting to be able to con- 
struct magnets which shall be perfectly free from any effect of 
temperature, I may at least be justified in considering that the 
extreme attainable limit has not yet been reached. 
3. The apparatus for absolute declination consists in essentials 
of a small magnet furnished with a mirror, which may be in- 
verted in the usual manner for eliminating the error of collima- 
tion: I will particularize one circumstance only. 
In determining the declination it is necessary to make the 
optical axis of the telescope of the theodolite perpendicular to 
the plane of the mirror of the magnet. For this purpose I give 
to the theodolite telescope the following means of adjustment :— 
iH) s 
GA || 
b 0 s 
o is the eye-piece; its tube is pierced above at a; a mirror } 
at an inclination of 45° to the optical axis throws the light re- 
ceived from above towards the object-glass, and illuminates the 
wire at c. From c the rays, forming a cone, proceed to the 
object-glass OO, are rendered parallel by the object-glass, and 
arrive at the mirror of the magnet ss. Being reflected back 
upon the object-glass (still parallel), they produce at the focus of 
the telescope an image of the wire. The telescope is turned 
until the wire ¢ coincides with its image, when the optical axis 
of the telescope will be perpendicular to the mirror. 
The following ,observations, made in the magnetic observa- 
tory, show the degree of accuracy of which this instrument is 
susceptible. The first column in the table gives the approxi- 
