After a discussion of the requirements to be met by the method of stirring used in liquid calorimeters for accurate heat measurements, a description is given of a calorimeter constructed on a new principle. The development of heat occurs in a cylinder which is somewhat smaller than the calorimeter vessel. Water is circulated between the two walls, where the thermometer is placed by means of a centrally fixed centrifugal pump. In this way, the actual wall temperature of the calorimeter‐vessel is measured; this is essential for the accurate calculation of the “radiation” correction. A simple method is described for determining the heat due to stirring. Finally, the required minimum duration of the main period is discussed.

The heats of combustion of some samples of pure benzoic acid, partly supplied by the National Bureau of Standards (Washington) and partly prepared by the authors, have been carefully measured by reference to electrical standards. The mean value of the absolute heat of combustion of benzoic acid under standard bomb conditions in some series of determinations carried out in the years 1946‐1948 is 26438 ± 4 j/g. New determinations in 1954 showed a mean value of 26435 ± 4 j/g.

A series of five 2‐n‐alkylnaphthalenes has been synthesized. Several physical properties of these substances have been determined. Their purity has been determined from melting curve data.

In order to be able to calculate readily the so-called "radiation" correction in calorimetric measurements. the jacket-temperature must be kept constant and the contents of the calorimeter vessel must be stirred at a constant speed. This article describes a thermo-regulator for this purpose, in which the fluctuations are not greater than 0.002".For every calorimeter used for accurate work, it is desirable to investigate beforehand whether the heat exchange between the vessel and the jacket satisfies Ncwfon's equation. W i t h the help of recent data it has been investigated how wide the air layer may be before convection occurs. This is necessary in order to obtain a system which satisfies Newton's equation.Finally. various methods of approximation for calculating the "radiation" integral are discussed. $ 1 . Introduction.In general, the following equation holds good for the "radiation" correction in calorimeters: Giving T, a constant value, which is reached by a constant temperature of the water-jacket, T,,, (see § 2 ) and by ensuring a constant rate of stirring (compare Communications I and 11). the carrying out of observations in such a way that equation (,1) can be computed, and also the method of calculating, are considerably simplified.In deriving equation ( 1 ) the validity of Newfon's equation is the