MOSTmethods currently used to determine the concentration of unsaturated compounds in gaseous hydrocarbon mixtures were introduced years ago. The procedures have been modified occasionally to increase accuracy, but some methods give results which are inherently inaccurate and are retained either because of convenience or because errors have not been recognized or investigated. Recent improvements in older refining processes and the introduction of new processes, such as polymerization, alkylation, and the production of synthetic rubber raw materials, which involve olefinic compounds, have demanded higher accuracy in the analysis for these substances.A method for determining unsaturates in gaseous hydrocarbons by catalytic hydrogenation has been described by McMillan, Cole, and Ritchie (5). It is claimed to give results accurate within a few tenths of 1 per cent•and to be as rapid as others in general use at present. The method is apparently satisfactory from the chemical standpoint: the hydrogenation is complete, free of side reactions, and sufficiently rapid for routine use, and the catalyst is easily saturated with gas and produces no interfering effects. However, the method, as originally described by McMillan, Cole and Ritchie, is based on the assumption that the lower hydrocarbons are ideal gases. Since these compounds are known to deviate considerably from the ideal gas laws, the nature of the errors introduced by the assumption was studied.In order to formulate a convenient procedure for correcting the analytical results for such deviations, it is necessary to consider the nature and sequence of the gas volume measurements made during an analysis. These are:

THE accelerated development of the carbon monoxide hydrogenation process for the production of gasoline during the years immediately following World War II, will have, perhaps in

Notes 627 excess of water is zero. The Gibbsian surface excess will be proportional to the difference between the area under the concentration curve and the area of the rectangle --0-s. For this case it is easy to see that, even though there is a surface layer of solute (positive adsorption at the interface), the Gibbsian surface excess may be positive, zero or negative depending on the concentrations in the double layer region below the interface. It is also plausible that changes in the bulk concentration will change the concentrations in this region, and thus change the surface excess, even though only slight changes occur in the outer ionic layer. Thus the apparent paradox of a surface deficiency and yet a surface tension less than that of water is readily understandable.

Discussion of ResultsThe results show that uranium can be separated from iron and titanium by electrolysis of a solution containing these elements followed by the precipitation of titanium with cupferron. The uranium then can be determined volumetrically.The average value of the uranium content found for kolm is slightly higher than that of Wells,2 0.462% compared to his average, 0.432%. This value, 0.462, also increases the discrepancy between the chemical analysis and the radioactive analysis of the kolm.It is of interest to note that the uranium content of the Canadian uraninite found by the above described volumetric method (53.47%) is in close agreement with the value (53.52%) found by Wells, who made a gravimetric analysis on material taken from the same twenty-pound sample. Summary 1. Large amounts of iron can be separated completely from small amounts of elements such as titanium, vanadium and uranium by electrolysis using a mercury cathode.2. Uranium cannot be determined accurately in presence of titanium by titration with permanganate after reduction by zinc. 3. The color change of permanganate in cold colored solutions is remarkably intensified by the presence of a little diphenylamine sulfonic acid.4. Uranium in kolm ash and in a Canadian uraninite was determined volumetrically. Results on both substances agree well with the gravimetric results reported by R. C. Wells.

No abstract