267 
of Edinburgh, Session 1884 - 85 . 
is caused to move very rapidly up and down, and having power 
enough to prick holes in this legal cap paper. On attaching a pro- 
peller or stirrer to it, it was found that there was sufficient power to 
keep the propeller in motion, and obtain the results that are men- 
tioned in Professor Barker’s paper. 
To prevent the dangerous escape of the bisulphide of carbon vapour 
from the prism, which you justly fear, a mercurial stuffing box was 
employed In the neck of the bisulphide prism bottle, a 
little below the upper edge, is a cork, that has through its centre 
a small glass tube ; through this passes a steel wire or shaft having 
attached to its lower end the propeller, half an inch in diameter. 
On the upper end of the shaft is fastened a small glass cap with its 
opening downwards, so as to dip below the surface of the mercury, 
in the upper portion of the neck of the bottle above the cork. 
On the top of this cap is fastened a pulley to receive motion from 
the Edison electric motor (just described) by means of a thread or 
small belt. It was found that, even when the propeller was caused 
to revolve very fast, there was not the slightest leakage of the vapour 
of the bisulphide of carbon 
The controlling of the temperature inside of the prism box is ac- 
complished by the use of two thermostat bars that were made under 
my direction for brother Henry. They are on the same principle as 
those that I invented many years ago for the taking of observations 
on temperature at the New York Observatory. Each bar consists 
of a strip of vulcanite a foot long, one inch wideband of an inch 
thick, riveted to a similar strip of brass only yjy of an inch thick. 
The rivets are in two rows set one inch apart. One end of the bar 
is fastened firmly to a block, while the other end is free to move 
with the changes of temperature. The vulcanite expanding more 
than the brass, causes the thermostat bar to bend towards the brass 
side, on an increase of temperature ; but if it falls, the bar bends the 
reverse way. On the loose end of the bar is an adjusting screw 
(with a platinum point for electric contacts), so that it can be set 
for any temperature. We used two of these bars so as to have very 
quick action. If there is the slightest fall of temperature in the 
prism box, the platinum points of the' bars come in contact, causing 
the current of the battery to pass through the electro-magnet, and 
open the valve of the gas lamp to give more heat in the metal pipe 
VOL. XIII. 
T 
