486 ON IMPRESSIONS OF COLD 
agreement between theory and observation is most satisfac- 
tory. 
3. Pendant.—The zthrioscope might be reduced to a small- 
er and more compact form, by conjoining with it a pendant 
differential thermometer. Neither of the glass balls in this 
case requires to be gilt. But of the Pendant A‘thrioscopes 
there are two varieties. In the first, the balls of the pyroscope 
are enclosed by hemispherical brass cups. (See Fig. 6. Plate 
XI.) These, however, may be a little deeper, forming seg- 
ments about 10 or 15 degrees more than the hemisphere. 
The pyroscopic balls, with two-fifths. of the diameter of the 
cups, should be placed in the middle, between the centre and 
the bottom. In the second variety, the lower ball of the pyro- 
scope is encased by a hollow sphere of brass, composed of two 
pieces which screw together, and the upper ball occupies the 
focus of the cup, which needs scarcely be more than two inches 
wide. (See Fig. 4.) This variety of the instrument is more 
portable than any of the rest, and equally accurate; but owing 
to the brass casing, it is, under a change of temperature, rather 
slower in its action. The construction is rendered still more 
commodious, by having the stem of the differential thermome- 
ter inserted through a projecting circular piece, somewhat lar- 
ger than the upper ball, which screws into a perforation at the 
bottom of the cup, and then forms part of the same reflecting 
cavity. This arrangement allows not only the standard of the 
:nstrument to be occasionally detached, but also the cup itself, 
_-a circumstance at once conducive to the safety of carriage, 
and to the preservation of the metallic lustre and polish, 
On replacing the metallic lid, the effect is entirely extin- 
guished, and the fluid in the stem of the differential thermo- 
meter immediately sinks to zero. A cover of pasteboard has 
at first precisely the same influence ; but after it has itself be- 
come chilled by this exposure, it produces a small secondary ac- 
tion 
