1. The 25°C. isotherm for the system lithium chloride-dioxane-water has been presented. The system shows immiscibility for mixtures of high dioxane content.2. A new solvate, LiCl • H20 • (CMD/L, has been prepared and identified.3. A method of preparing anhydrous lithium chloride from the double solvate has been described.

1. The method proposed by Eversole Doughty for the calculation of diffusion coefficients is in error except at the lowest concentrations.2. A new equation is derived which expresses D as a function of concentration for all concentrations.

in which the gas flow was slowed down to rates such as those encountered in practice, the recoveries are above 97.5%. This is considerably better than could be expected from the gravimetric method.Acetyl chloride is very easily hydrolyzed and in the presence of water only a small amount reacts with aniline to form acetanilide. This compound shows an absorption maximum at 240 µ in methanol (Figure 4). Chloroacetyl chloride quantitatively forms -chloroacetanilide in aqueous aniline. It also shows an absorption maximum at 240 m/j in methanol. APPLICATION TO STACK GAS ANALYSISThe precipitation of materials other than 1,3-diphenylurea may cause high results by the gravimetric procedure. This was emphasized by the analysis of two samples from a titanium chlorinator stack. The phosgene was precipitated as 1,3diphenylurea according to the procedure of Olsen and coworkers (7) and the precipitate was weighed. The precipitate was then dissolved in methanol and the 1,3-diphenylurea was determined by the ultraviolet method. The results are shown in Table II. As diphenylurea in the precipitate must inevitably be revealed by its absorbance, it is apparent that the gravimetric method was giving high results on these samples, particularly in the first case.

No abstract

An investigation was undertaken to determine whether or not the Preston hydrostatic pressure machine, modified so that an initial pressure of 25 lb. per sq. in. may be obtained, is capable of predicting bottling‐line production breakage by pressure tests on samples of the bottles in use. The quart and fifth‐gallon bottles investigated were round, whereas the pint and half‐pint bottles were of the flask type. Correlating data between production and the percentage of test breakage for quart and pint bottles are significant, and data for fifth gallons, although less significant, are definitely positive. Half‐pint data yielded no significant relationship between the percentage of production breakage to test breakage, but this is probably the result of abnormal bottling‐line conditions. A method of determining the range of production breakage percentage to that of test breakage by the use of the standard error of estimate is described.