DR. J. T. BOTTOMLEY ON THERMAL 
fiOO 
apparatus which L found it necessary to use for temperatures higher than the boiling 
point of water. 
When I had finished a considerable number of determinations, carried on precisely 
as described above, with boiling water for the superior temperature, I attempted to 
get to a temperature of 200° C. or 2.50° C. I found, however, that the soldered joint 
would not bear a temperature of anything like 200° when the Sprengel vacuum 
was applied. The solder became crystalline and brittle and leaky below 150°. The 
solder used was good ordinary solder ; and I think that slight leakiness (very per¬ 
ceptible to the vSprengel) manifested itself at a temperature at which the solder was 
for ordinary purposes quite good and sound. I am not sure, however ; and it is quite 
possible that by trials I might have got a better solder by using less tin in the com¬ 
position. I did not delay with such trials, but preferred to alter the enclosure to the 
form shown in fig. 2. cm is a copper shell nearly spherical, but wuth a wide mouth 
and neck. The shell is beaten into shape. The neck is carefully brazed on, and the 
parts round the brazing finished by beating. The neck has a flange with the upper 
edge ground flat. The whole is very thick and solid, and an outer shell of copper 
was subsequently added, squeezed into shape and fitting loosely round the other for 
tlie purpose of equalizing the heating (and cooling) of the enclosure. The brass tube 
tt and outer water-jacket are also shown in the diagram; and it will be seen that, 
to close the copper shell, the whole of this piece is brought down on the flange at the 
neck of the shell ; and the soldering is made at this joint. This construction was 
found to be in every way an improvement on that used at first. The only little 
trouble being the lowering in of the copper globe without touching the sides, and the 
making the measurements necessary for bringing the copper globe to the centre of 
the enclosure. These troubles are scarcely worth mentioning, and the new arrange¬ 
ment gave much greater freedom for working at the globe while it was hanging on 
the wire before being lowered into its place. 
In the arrangement just described it will be seen that the soldering is kept cool by 
the proximity of the cold water in the water-jacket; and this can be done even 
although the temperature of the shell enclosure is very high. In practice a stream 
of cold water was kept running through the water-jacket during the heating process 
described above, and for heating the enclosure a Bunsen burner or, in the latest 
experiments, one of Fletcher’s high-power burners was applied direct to the 
bottom of the copper shell.^' When the temperature of the globe inside was found. 
* A curious phenomenon, which requires farther examination, was noticed during the heating of the 
globe described in the text. I found it impossible to exhaust the enclosure with the Sprengel pump 
while the flame was applied; but as soon as the flame was removed I could exhaust the enclosure per¬ 
fectly. Whether the copper shell became porous at the high temperature (just visible redness) and thus 
admitted gas enough to prevent my obtaining a vacuum’; or whether some minute crack or opening was 
made at some of the joints owing to the great variation of temperature from point to point of the shell, 
I am unable to say at present. The leaking ceased, however, as soon as the heating ceased. This leak- 
