314 
MR. J. WELSH’S ACCOUNT OF METEOROLOGICAL 
travel, between its entrance into the tubes and its leaving the wet thermometer, was 
about 4 inches : the diameter of the tubes enclosing the bulbs was 0-4 inch, and that 
of the connecting tube 0-25 inch. The aspirator was a cylindrical bellows ; the valves 
being so arranged that, when the aspirator was forced open, the air could only enter 
it by passing over the thermometers : it was worked by attaching a weight to the 
lower end which pulled it open, the upper end being fixed; when it had opened to 
nearly its full extent, it was closed by means of a cord passing over pulleys and 
drawn up by the hand; a large valve allowing the air to escape rapidly from the 
aspirator as it was closed, and a second valve preventing the air from being driven 
backwards over the thermometers. Care was taken, in the construction of the different 
parts, that the aperture of the tubes should not be smaller between the thermometers 
and the external air, than between them and the aspirator ; otherwise the air might, 
by undergoing a certain degree of expansion, have come in contact with the bulbs in 
different conditions with respect to temperature and capacity for moisture from those 
of the external air. This was guarded against by applying a stopcock near the aspiiatoi, 
whose aperture was sufficiently small. A second flexible tube, with a stopcock, con- 
nected the aspirator with Regnault’s hygrometer ; so that the same aspirator might 
be used simultaneously for both instruments. Two different sizes of aspirator were 
used in the different ascents ; the one being 12 inches diameter, and extending to 
about 18 inches, occupying about 1^ minute in its descent; the other was 9 inches 
diameter, extending to 12 inches in 30 to 40 seconds. This was sufficient to pro- 
duce a current of air over the bulbs at the rate of 12 to 14 feet in a second; the 
vertical velocity of the balloon seldom exceeding 4 or 5 feet. The thermometers 
employed were of great sensibility ; the bulbs being cylindrical, the diameter not 
exceeding of an inch, and the length varying from a half to three-quarters of 
an inch. The length of one degree of the scale was from ^th to ^th of an inch, so 
that they could readily be read by estimation to 0°-] . The graduation extended to 30° 
or 40° below zero of Fahrenheit. The scales of those used in the first ascent were of 
brass, but afterwards of ivory, in order to render the column of mercury more visible. 
The errors of all the thermometers were determined throughout the scale, from about 
0° to 70°, by comparison with standards at the Kew Observatory; the comparisons 
below the freezing-point being made in mixtures of ice and salt. The corrections have 
been applied to all the readings. These thermometers were found to acquire the tempe- 
rature of the air very rapidly i when heated 20 above the tempeiatuie of the aii, and 
allowed to cool at rest in a confined room, they returned to within 0°-5 of the pre- 
vious reading in about 100 seconds; when gently fanned, by being carried through 
the room at the rate of 5 or 6 feet in a second, they returned to within 0°-5 in 40 
seconds ; when under the action of the aspirator they returned to within 0°-5 in 30 
seconds, and exactly to the original reading in 45 seconds. Any correction on 
account of sluggishness in the thermometers must thus be very small : this is shown 
by the observations of October 21, when the descent took place with about the same 
