It is worthy of notice that the indicator-diagram method of graphically representing the changes in a cycle of operations is adopted in Thomson's "Account," but does not occur in Carnot's essay. The cycles consist of two isothermal changes and two adiabatic changes; that is, two changes at the temperatures of the source and refrigerator respectively, and two changes--from the higher to the lower temperature, and from the lower to the higher. These changes are made subject to the condition in each case that the substance neither gains nor loses energy in the form of heat, but is cooled in the one case by expansion and heated in the other by compression. The indicator diagram was due not to Thomson but to Clapeyron (see p. 99 above), who used it to illustrate an account of Carnot's theory. There appeared in the issue of the Edinburgh _Philosophical Transactions_ for January 2, 1849, along with the "Account of Carnot's Theory," a paper by James Thomson, entitled, "Theoretical Considerations on the Effect of Pressure in Lowering the Freezing Point of Water." The author predicted that, unless the principle of conservation of energy was at fault, the effect of increase of pressure on water in the act of freezing would be to lower the freezing point; and he calculated from Carnot's theory the amount of lowering which would be produced by a given increment of pressure. The prediction thus made was tested by experiments carried out in the Physical Laboratory by Thomson, and the results obtained completely confirmed the conclusions arrived at by theory. This prediction and its verification have been justly regarded as of great importance in the history of the dynamical theory of heat; and they afford an excellent example of the predictive character of a true scientific theory. The theory of the matter will be referred to in the next chapter. CHAPTER VIII THERMODYNAMICS AND ABSOLUTE THERMOMETRY