20 9 
considerable depths of the Caribbean sea. 
as the water appeared to have had great difficulty, and was 
probably some time in forcing its way into the interior of the 
closed cylinder ; and partly to the heat which so great a 
thickness of metal would retain for a considerable time ; the 
surface water was from 8 2°, 5 to 83°, 2 in the course of the 
afternoon ; the difference of temperature between the surface, 
and a depth exceeding 1000 fathoms was, therefore, 33 0 , 3 
by one thermometer, and 3 7°, 3 by the other, the indication 
of the latter being entitled to the most reliance. 
“ It may be reasonably inferred, that one or two hundred 
fathoms more line would have caused the thermometer to 
have descended into water at its maximum of density, as de- 
pends on heat, below which, consequently, no farther dimi- 
nution of temperature would take place ; this inference being 
cai the presumption that the greatest density of salt water 
occurs, as is the case in fresh water, at several degrees 
above its freezing point. 
“ It is desirable to repeat this experiment, should an op- 
portunity present itself, and to attach several register ther- 
mometers, at intervals of the line, to ascertain the progres- 
sion in which the temperature diminishes ; and it would be 
still more interesting to determine the depth beneath the sur- 
face at which water is found, at the term of the diminution 
of its temperature in different parallels.” 
The opportunity which was thus hoped for, did not occur 
in the remainder of the voyage ; the experiment requires fa- 
vourable circumstances of weather, and of sea, to both of 
which the season was adverse. 
I am not aware of any previous experiments on the tem- 
perature of the deep sea within the tropics, except those 
mdcccxxiii. E e 
