APPARATUS EMPLOYED IN DEEP-SEA EXPLORATIONS. 283 
errors, to indicate too high a temperature, from the pressure of 
the sea-water at great depths (which amounts to one ton per 
square inch for every 800 fathoms) slightly compressing the 
bulb and forcing the mercury, or spirit, too high up in the 
column. Many expedients were tried to protect the thermo- 
meters from this source of error, and ultimately the plan 
suggested by Professor W. A. Miller, Treasurer E. S., was 
adopted. A figure of the instrument, and a detailed descrip- 
tion of it, will be found in the Proceedings of the Royal 
Society , June 17, 1869. It consists of the ordinary form of 
Six’s self-registering thermometer, the bulb of which is sur- 
rounded by an outer bulb, hermetically sealed, and the inter- 
vening space partly filled with spirits of wine. A careful 
series of experiments made on shore with the assistance of a 
hydraulic press, showed that the pressure only acted upon the 
outer bulb, the inner indicating the true temperature, and 
that the difference in indications between this and an ordinary 
thermometer increased regularly with the pressure. These 
differences were carefully noted at intervals up to three tons 
to the square inch, and it was exceedingly interesting to 
find that a set of independent experiments made at sea, when 
the protected and ordinary thermometers were simultaneously 
employed, gave precisely the same differences between the 
two instruments when the pressure was calculated at the rate 
of one ton per square inch for every 800 fathoms of depth. 
In addition to the precautions against pressure, these instru- 
ments, to the construction of which Mr. Casella paid the 
utmost attention, were furnished with registering indices of a 
peculiar kind, which fitted so tightly in the tubes that they 
could not be displaced by the shaking to which they were oc- 
casionally subjected, notwithstanding the most careful handling, 
and which could only be set by powerful magnets, with grooves 
cut in the end of their poles, so as to partially surround the 
thermometer tube. The whole instrument was protected by a 
copper cylinder open at each end. The expedition was provided 
with six of these instruments, but the two which were originally 
employed were found to work so entirely satisfactorily that they 
were used for every observation, and it was calculated that they 
travelled vertically up and down in the sea more than a hundred 
miles during the three months that they were in use. 
The apparatus for collecting samples of sea-water from 
various depths was of very simple construction. It consisted 
of a cylindrical brass tube (the interior being coated with 
varnish, to avoid the action of the sea-water upon it), provided 
at each end with an accurately ground valve opening upwards. 
As long as this was descending vertically, the water passed 
freely through it, but when the motion was reversed the 
