262 
POPULAR SCIENCE REVIEW. 
derful. My arguments are founded on the clear and simple 
obviousness of the fact that a given quantity of dead organic 
matter diffused through a large quantity of water sullies it 
less than if it were small, and on the necessity of main- 
taining an evenly moderate temperature for the reasons already 
given, avoiding the high and low ranges of the atmosphere ; 
and I show that the easiest manner of attaining this is by 
having a large reservoir sunk in the earth at a distance giving 
a known temperature. Thus, referring to the sunk ther- 
mometers at the Greenwich Observatory, with a thermometer 
having its bulb on a level with the scales of the sunk instru- 
ments, the lowest (January) mean monthly reading in a named 
year was 36*4° F., with a mean daily range of 6*9° F. ; and under 
the same circumstances the highest (July) mean monthly read- 
ing was 66*9° F., with a mean daily range of 1 9*9° F. But 
from the showing of other thermometers whose bulbs are 
sunk in the ground to the respective depths of one inch, three 
feet, twelve feet, and twenty-five feet, the temperatures become 
strikingly even for the whole year through — so much so, that 
at twenty-five feet deep the mean monthly reading of January 
was 52° F., with a mean daily range of only 0*025° F. ; and the 
mean monthly reading of July was 49*0° F., with a mean daily 
range of but 0*06° F., the highest mean daily range at that depth 
in any month of the year being 0*07° F. in August. To apply 
this to aquaria, I have made the accompanying diagram (p. 261), 
it being an ideal vertical section of an aquarium, consisting 
of a show-tank a, with its reservoir b in the earth, with an inlet 
supply-pipe c, and an outlet pipe d, the six arrows showing the 
direction of flow of water, e is a pipe supplying water to com- 
pensate for evaporation, which, both for marine and fresh-water 
aquaria, should be distilled water. For simplicity, the engines, 
pumps, &c., which circulate the water are omitted, the showing 
of results being alone aimed at. 
Supposing that in any part of an English year the temperature 
of B would be 60° F., and that in summer a would rise to 75° F., 
that would be much too warm for an aquarium containing 
British animals. Or it might in winter sink to 30° F. or less, 
that would be much too cold. But on a sufficient circulation 
being established between a and b, then their mean tempera- 
tures would be expressed by the seven following formulas, vary- 
ing according to the size of B : — 
Formula No. 1. A 2 . B 1 . Mean result 70° F. 
„ „ 2. A1 . B1 67*5° F. .. 
