Reference Label	Details
Rossini 1931a	F. D. Rossini, J. Res. Natl. Bur. Stand. 6, 37-49 (1931) The Heats of Combustion of Methane and Carbon Monoxide
Rossini 1931b	F. D. Rossini, J. Res. Natl. Bur. Stand. 7, 329-330 (1931) The Heat of Formation of Water and the Heats of Combustion of Methane and Carbon Monoxide. A Correction
Prosen 1945	E. J. Prosen and F. D. Rossini, J. Res. Natl. Bur. Stand. 34, 263-269 (1945) Heats of Combustion and Formation of the Paraffin Hydrocarbons at 25° C
Rossini 1940	F. D. Rossini, Chem. Rev. 27, 1-16 (1940) Heats of Formation of Gaseous Hydrocarbons
note CH4	Rossini 1932 points out that Rossini 1931a used incorrect ly pressure corrections and gives 889720 +- 230 int. J/mol as the correct mean of Table 4 in Rossini 1931a for the negative of the enthalpy of combustion of CH4 at 30 deg C. We find that in addition Rossini 1931a has flipped the sign in the correction of experiments from temperatures 30.0x to 30.00 deg C (he subtracts, rather than adding, small corrections of the order of 1-3 J/mol, but the big correction, from 30 to 25 deg C has been used with the proper - positive - sign). Reworking the values in Rossini 1931a Table 4, and leaving out the first measurement ("A"), which differs the most from the others (as he did), results in a mean value of 889722.7 +- 126.0 int. J/mol, after introducing corrections to 30.00° C of the order of 1 J/mol for 0.01 deg C excess temperature, and the correction of -345.1 J/mol for the 0.122% CO impurity. This is slightly higher than Rossini 1932 claims. Also, the uncertainty given with our mean is simply one std. dev. For 5 measurements, it should be multiplied by 2.776451 to get the desired 95% confidence limit, resulting in +- 349.8 int. J/mol. (It appears that Rossini is assigning an uncertainty corresponding to the average abs. dev., which is 237 J/mol). For comparison, we correct the mean from the table to 25° C using Rossini 1931a value of -93 int. J/mol/K to get 890187.7 int. J/mol, or with a more curent value of 467 (int.) J/mol we get 890189.7 int J/mol. Rossini 1932 combined with Rossini 1931a implies a value of 890180 +- 300 int. J/mol. Note that these values do not seem to contain the correction to the ideal state of unit fugacity. However, Rossini 1940 gives a standard enthalpy of formation of CH4 as -74735 +- 310 int. J/mol, as well as the enthalpies of formation of H2O and CO2 of -285795 +- 40 int. J/mol and -393355 +- 46 int. J/mol. From these, one derives an enthalpy of -890210 J/mol, which is more negative than the implied value by 30 J/mol. On the other hand, Prosen 1945 gives the standard enthalpy of combustion of -212.798 +- 0.072 kcal/mol = -890198 +- 300 int. J/mol, and quotes a small change in the molecular weight of H2O to 18.016 (presumably from 18.0156) as well as a slight change in conversion to zero pressure. Undoing the change in molecular weight of Prosen 1945 leads to -890178 +- 300 int. J/mol. This is basically the same (only 2 J/mol less negative) as the value of -890180 +- 300 int. J/mol implied by combined Rossini 1932 and Rossini 1931a. One can calculate the corrections for non-ideality by using Hideal - Hreal of 8.06 J/mol for O2, 41.4 J/mol for CO2 and 8.06 + 15.9 x + 17.4 X^2 J/mol for (1-x) O2 + x CO2 mixtures, where x is the mole fraction of CO2 (note that the expression correctly results in 8.06 J/mol for x=0 and 41.4 J/mol for x=1) from Rossini 1939 and 15.5 J/mol for CH4 from Pittam 1972. Unfortunately, it is not explicitly known how much excess O2 was used in those experiments. Assuming a stoichiometric reaction, CH4 + 2 O2 the corrections are: -15.5 -16.1 and 41.4 J/mol for CH4, 2 O2 and CO2, resulting in 9.8 J/mol. For 100% excess oxygen, CH4 + 4 O2 -> CO2 + 2 O2 + 2 H2O, the corrections are -15.5, -32.2, and 45.9 J/mol for CH4, 4 O2, and 3 moles of x=1/3 CO2 + O2 mixture, for a total of -1.8 J/mol. For 50% excess oxygen, CH4 + 3 O2 -> CO2 + O2 + 2 H2O, the corrections are -15.5, -24.2, and 40.8 for CH4, 3 O3, and 2 moles of x=0.5 mixture of CO2+O2, for a total of 1.1 J/mol. In fact, the correction of 2 kJ/mol, implied by Prosen 1945, corresponds to 40% excess oxygen, CH4 + 2.8 O2 -> CO2 + 0.9 O2 + 2 H2O: -15,5, -22.6, and 40.1 for CH4, 2.8 O2 and 1.8 moles of x=0.55 CO2 + O2 mixture. Note that in the contemporary experiments on CO Rossini used 78 % excess O2, while for H2O the mixture was taken to be stoichiometric (or, maybe there was excess H2, which has a very small correction of -0.48 J/mol). At any rate, the correction is very small compared to the overall uncertainty. Assuming indeed that for CH4 the excess oxygen was 40%, the correction for non-ideality is +2 J/mol to DeltaH. Converting our reinterpretation of 889722.7 +- 349.8 int. J/mol to abs. J and to present molecular weight of water (which was used to determine the mass of CH4) of 18.01528 from 18.0156 results in the enthalpy of combustion at 30° C of -889849.3 +- 349.8 J/mol (corresponding to -890316 +- 350 J/mol at 25 deg C; for comparison Prosen 1945 value when correct for change in molec. weight is -890311 +- 300 J/mol). The current result should possibly be made less negative by 2 J/mol on account of the correction to ideal gas standard state, albeit in view of the overall uncertainty this correction is entirely negligible. Note that Cox 1970 quote the Rossini 1931 and Rossini 1931a result as 212.79 +- 0.06 kcal/mol = 890.31 +- 0.25 kJ/mol at 298.15 K, which would correspond to 889.84 +- 0.25 kJ/mol at 303.15 K.