August 6, 1885 | 
NATURE 
325 
going from the one level to the other. It thus appears 
that we shall get the greatest effect by selecting two metals 
near the opposite extremities of the list. Bismuth and 
antimony are generallly the metals chosen, and 36 or 
49 junctions of these are frequently used, the metals being 
packed close together, but insulated from each other, and 
thus forming a sort of cube, each end of which contains, 
say, 36 junctions. These junctions are generally covered 
with lamp-black. If the one end of this be heated we 
shall have a current in the one direction ; if the other end, 
we shall have one in the other direction ; while, if both 
ends be heated simultaneously and to the same extent, 
Th) ANTIMONY 
CALVANO 
METER 
Fic. 5. 
we shall have no current whatever. This arrangement 
forms what is termed a ¢hermopzle, and the cube of ele- 
ments is generally encased in a brass covering presenting 
two terminals, in which the wires of the galvanometer are 
to be inserted and screwed tight. Inasmuch as this 
arrangement is generally used for viewing and measuring 
heat rays, a brass cone polished in the inside is often 
attached to the thermopile (Fig. 8) with the view of 
catching a large area of heat rays and reflecting them 
into the pile. 
The galvanometer consists essentially of a magnet, 
which is delicately suspended by a very fine thread. 
Around it we have numerous coils of wire (but not in this 
case very numerous coils of very fine wire), which con- 
vey the current, each single coil counting separately in its 
action upon the magnet. 
The various coils must, of course, be insulated from 
each other. A comparatively weak current will thus pro- 
ores a visible effect if there be only a sufficient number of 
coils. 
But yet the result so obtained is not the best, because 
we are having, after all, a strife between the influence of | 
the current and that of the earth upon the small magnet. 
Assuming that the galvanometer was so placed to begin 
with that the magnet was in the magnetic meridian, 
then the current will tend to move the magnet to a 
position at right angles to this plane, while the earth’s 
magnetic force will tend to keep it where it is. There is 
thus a strife between the two, and this will greatly inter- 
fere with the delicacy of the instrument. What we have 
Fic. 6. 
to do is so to counteract the earth’s directive force that the 
little magnet may behave as if it was not under any 
external magnetic influence whatever. 
A needle for which the directive effect of the earth’s 
magnetic force is thus neutralised is said to be rendered 
astatic. 
There are two ways in which this may be accomplished. 
We may use two needles of as nearly as possible the same 
strength, joined rigidly together with their poles in oppo- 
site directions, as in Fig. 6. Numerous coils of wire are 
wound around the lower needle, one of which we have 
exhibited. Here the upper current will tend to twist 0’ 
above the plane of the paper, while the lower current will 
act on 2’ in an opposite direction, this lower current, how- 
ever, being further removed from the upper needle than 
the upper current, the latter will predominate, and the 
needle will, on the whole, be twisted round so as to place 
& above the plane of the paper. Furthermore, the lower 
needle will be twisted round by both the upper and the 
under currents so as to place a above the plane of the 
paper, and hence the two needles will be twisted by the 
current in the same way, whilst the directive force of the 
earth’s magnetism which opposes any motion of the needle 
will, by the arrangement above alluded to, be either 
altogether cancelled or rendered very small. 
A galvanometer of this kind was employed by Melloni 
along with a thermopile as already described, and it was 
with these that he obtained the valuable results which we 
shall presently mention. But before dismissing this subject 
let us allude to some still further refinements made since 
the time of Melloni, which have contributed very greatly 
to increase the delicacy of this combination. 
We have spoken about one way in which the effect of 
the earth’s force may be neutralised, but we may likewise 
adopt the method of Sir W. Thomson, indicated in Fig. 8, 
where an external magnet, M, is so placed as to cancel the 
earth’s action on the suspended galvanometer magnet 
which is supposed to be placed in the centre of G. 
A still greater refinement consists in the joint use of 
both the methods now described. A system of two magnets 
M 
Si AU 0 
Fic. 8. 
placed oppositely and united rigidly together is employed, 
a separate coil being made to surround each magnet. An 
external magnet is then so placed as to neutralize any 
directive force that may yet linger in the system. By this 
means very great delicacy almost amounting to instability 
may be obtained. 
We shall conclude by mentioning an optical arrange- 
ment introduced by Sir W. Thomson, which greatly adds 
to the delicacy of the galvanometer. In this arrangement 
(Fig. 7) a small galvanometer magnet is attached to the 
back of a mirror, which mirror is suspended by a very 
fine thread. 
Again (Fig. 8) there is a lamp behind the scale Ss’, and 
a slit, or, better still, a round aperture below the scale, with 
a wire in its middle, is lighted up by the lamp, and a 
