108 Proceedings of the Royal Society of Edinburgh. [Sess. 
necessarily in effectual contact with the neck-tube at the point, the 
temperatures obtained were not strictly those of the metal ; the difference, 
however, was probably not more than a few degrees, and was greatest 
where it mattered least — namely, near the mouths of the flasks. 
The results are set forth graphically in fig. 3. 
Three of the bottles were 50-lb. containers of dimensions substantially 
those indicated by fig. 1. The fourth was a German-built vessel — the 
largest vacuum flask in the country — capable of holding over 300 lbs. of 
liquid air. 
The first container examined (see curve labelled “ German 50-lbs.,” fig. 3) 
TEMPERATURE IN CONTAINER NECKS 
was a brass vessel built and evacuated 
in Germany before the war. After many 
years of continuous service at the New- 
castle mine rescue station its vacuum 
was showing signs of breakdown ; its 
evaporation rate at the time of the test 
was 1*85 litres of gas per minute at 
18° C., which is equivalent to a daily 
loss of over 7 lbs. The top of the 
outer neck was thickly coated with ice. 
The inner neck was of German silver, 
f-in. bore, the thickness of the metal 
being 0 02 in. As fig. 3 indicates, the 
neck enters the vacuum space 3 T 9 -g- ins. 
from the mouth of the flask. The 
temperature T 7 -g- in. below that point was 
as low as — 172° C., and at all parts 
below 4^ ins. from that point the temperature was the same as that 
of the liquid in the flask. Under the conditions then obtaining, therefore, 
the loss due to neck-conduction was nil, and if the neck had been shorter 
by 5 ins. it would still have been nil. 
The second and third containers examined were made in this country, 
and were in satisfactory condition. They were of copper, with German- 
silver necks 13^- ins. long and J-in. bore. Their rates of gaseous discharge 
(measured at 16° C.) were respectively 1T5 and T03 litres per minute at the 
time of the tests. The curve labelled “ L. 12 and L. 13,” fig. 3, expressed the 
fall of temperature down the neck for both these flasks. The temperatures 
at the bottom of the necks were, in these cases, appreciably above the 
temperature of the liquid, a fact which seems to indicate that the evaporated 
gas gained a little heat from the upper part of the inner globe before it 
