100 PROCEEDINGS OF THE AMERICAN ACADEMY. 



Specific Heat at Constant Pressure. — The specific heat at 

 constant pressure is shown on Folder XI; the cur^'es for the twelve 

 liquids at foiu* temperatures in Figures 87 to 98, and the collection 

 of the average results in Figure 99. The quantity listed as change 

 of Cp is the specific heat at atmospheric pressure minus the specific 

 heat at the pressure in question. A positive change means, there- 

 fore, that the specific heat is less at the pressure in question than 

 at atmospheric pressure. In order not to confuse the curves, the 

 origin for each temperature has been displaced with respect to the 

 neighboring cin-ves. The scale of the drawing is shown at the right 

 hand side. 



The twelve liquids show a bewildering variety, so bewildering that 

 speculation as to the cause of all the variations is hopeless. It is to 

 be pointed out nevertheless, that such great variety is to be expected 

 if the molecules take up different positions more or less symmetrical 

 in arrangement with increasing pressure. The process is similar in 

 many ways to a process of association, which is accompanied by much 

 greater changes in the specific heats than in the volume, or compres- 

 sibility, or dilatation. The curves show some points of similarity, 

 however. It is an almost universal rule that the initial change of 

 Cp at any temperature is a decrease. For the majority of liquids 

 the specific heat on the whole decreases at the high temperature and 

 increases or does not increase so much at the low temperatures. We 

 have seen that as a rule the specific heat at atmospheric pressure is 

 higher at the higher temperatures. For some liquids the tempera- 

 ture efi'ect may be very marked. The change under pressure is in 

 such a direction as to bring the specific heats at high pressures more 

 nearly to equalit;y' for all the temperatures. The three halogen com- 

 pounds are an exception to the rule, however. The very large in- 

 crease of Cp for ethyl chloride at 80° is very much like that already 

 found in the case of water. It may mean an abnormally high rate 

 of dissociation at the higher pressures. The four normal alcohols 

 show similarities in the abnormally large decrease of Cp at 80°. The 

 decrease evidently cannot go on indefinitely. For methyl and ethyl 

 alcohol, the maximum pressure is sufficient to change the decrease 

 into an increase, as it must eventually, but for propyl and amyl 

 alcohols, the reversal in direction must be at higher pressures than 

 reached here. It seems to be at hand for propyl alcohol. 



There is one very rough check which may be applied to the values 

 given here for the specific heats; namely, in no case must the specific 

 heat decrease by an amount more than its original value, for a nega- 



