An analysis has been made of the heats of formation and combustion of organic gases and liquids, and of the heats of vaporization of liquids. The worlr has been done as far as possible with homologous series, in order to discover systematic effects. The data are converted into heats of atomization, i.e., the heats required to convert the gases or liquids into their constituent atoms. It is shown that the heats of atomization of the gaseous aliphatic hydrocarbons (paraffins, olefins, and acetylenes) can be accurately represented by a scheme in which a distinction is made between primary, secondary, and tertiary carbon-hydrogen bonds, and between bonds that are next to, and next but one to, multiple bonds. For aromatic molecules an appropriate correction for resonance is proposed. For other types of compound it is found that suitable values for the various bonds (e.g., C-CHO, C-01-1) will give rise to good agreement with experiment.I t is shown that to a reliable approsinlation heats of vaporization are also amenable to the same treatment. Since this is so it is possible to assign bond values on the basis of which it is possible to m a l e predictions about heats of formation of liquids.A system of coefficients is worked out by means of which the numbers of atoms of various kinds in a n~olecule can be expressed in terms of the numbers of the different lrinds of bonds. On the basis of these it is shown how bond contributions t o heats of formation and heats of combustion can be calculated.A table (Table X) gives the contributions proposed for the heats of atomization, ' heats of formation, and heats of combustion for both gases and liquids.
INTRODUCTIONVarious attempts have been made from time to time to devise systems that will enable reliable values of heats of formation and of heats of combustion to be deduced on the basis of chemical structure. Very approximate values can be calculated assuming strict bond additivity, using bond strengths such as those listed by Pauling (26) and by Syrkin (37); however such a procedure can lead to considerable errors in organic molecules of even only moderate complexity. Various schemes for the inclusion of terms for i~lteractio~ls (27) between bonds or between atoms have beell proposed, notably those of Taylor, Pignocco, and Rossini (42), Tatevskii (38, 39,40), and Bernstein (4). Of these workers Taylor, Pignocco, and Rossini did not specifically consider heats of formation, but Tatevskii showed that reliable predictions could be made for paraffins by recognizing three different types of carbon-hydrogen bonds and 10 different types of carbon-carbon bonds. Bernstein later showed that equally good agreement could be obtained with fewer types. In the present paper it is shown that satisfactory agreement for the lower paraffins is obtained by treating all carbon-carbon bonds as the same and recognizing three different types of carbon-hydrogen bonds. For olefins and acetylenes six additional types of bonds are required, and in most cases it is shown that the scheme is equiv...