812 
Proceedings of the Poycd Society 
that salt water has no maximum density point, led us to suspect a 
maximum density point below zero centigrade. Thus, for 1 ton 
per square inch pressure, our experiments show that a line drawn 
as in the fresh water diagram above, would cross the axis at about 
- 5°, that the corresponding 2 ton line would cross at - 8° - 5, the 
3 ton line at -11°, and the 4 ton line at ~ 13°. Whereas, the 
corresponding points in the fresh water diagram are + 1°*5, - 1° *5, 
-3°, and - 4° *2 respectively. We found also that for equal 
initial temperatures the heating effect produced by a sudden in- 
crease of pressure was greater in salt than in fresh water. As the 
experiments with salt water were performed at temperatures nearly 
the same as those with the fresh water, they gave us greater confi- 
dence in the correctness of the conclusions we arrived at in the 
case of the latter. 
Note, added July 21, 1882. 
A careful consideration of the diagram on page 811 will show 
that if the rate of change of the maximum density point with pres- 
sure be uniform, as is assumed by Professor Tait in his note which 
♦ 
follows, then when a line crosses the axis, that is, when there is 
neither heating nor cooling shown on a sudden release of a certain 
pressure, the water is at the temperature of the maximum density 
point for half the pressure that the line represents. The following 
example will make this clear : — Water at a temperature of 1 0, 5 cent., 
when suddenly released from 1 ton on the square inch (150 atmo- 
spheres) pressure, is left at the same temperature after the release, 
as is shown by Line I. in the above diagram. This would at first 
sight seem to indicate that' the maximum density point for water 
under 1 ton pressure is 1 0, 5 cent. But if the temperature of the 
maximum density point is lowered by pressure, it must rise as the 
pressure falls off, and if the lowering is proportional to the pressure, 
it will rise uniformly. So, if in the example given above there is 
no final heating or cooling of the water, it shows that there must be 
first a cooling and then an equal heating ; that is to say, that the 
maximum density point must be as much below 1 0, 5 cent, at the 
beginning of the release of pressure, as it is above it at the end. 
Now we know that at the end it is 2°*5 above l°-5 (’viz., 4° cent.), 
therefore it must be 2° '5 below 1 0, 5 when the pressure begins to 
