increased, and that more and more of the entering nitro group took up a position meta to the hydroxyl. However, a maximum value of about 0.8 was reached for the meta-ortho ratio. This change in directive power is ascribed to the salt-forming property of the oxygen atom of ^-cresol. The predominating tendency toward ortho substitution, even when a large excess of acid is present, proba.bly arises from the fact that the oxonium salt is in part un-ionized.In the case of £-cresyl carbonate, increase in the amount of sulfuric acid likewise increased the meta-ortho ratio. It is assumed that here also oxonium salt formation is responsible for the change.These changes in orientation are entirely analogous to the modification in the orientation of amino and substituted amino groups by the addition of sulfuric acid.

9‐Nitroanthracene reactant: Twenty grams (0.112 mole) of finely powdered anthracene product: 9‐nitroanthracene

The methods of synthesis for p-phenylenediguanidine, p-bromophenylguanidine, p-iodophenylguanidine, 4,4'-diguanidobiphenyl, 4,4'diguanidodiphenylmethane, and certain of their salts have been described.2. The effects of p-phenylenediguanidine, 4,4'-diguanidobiphenyl and 4,4'-diguanidodiphenylmethane upon the blood sugar of normal rabbits have been studied and compared directly with the hypoglycemic properties and toxicity of a compound of the synthalin type, 1,6-diguanidohexane, and the results are briefly discussed from the point of view of hypoglycemia and chemical constitution in the guanidine field.

No abstract

We have studied the polymerization of ketene in the gaseous and liquid phases and in solution.Ketene does not appear to polymerize appreciably in the gas phase at room temperature. Polymerization only sets in at one or more points on the surface of the vessel after the gaseous ketene has stood for several hours.The pure liquid is stable at -80°but at 0°p olymerization is rapid and is complete in about one and one-half hours.In solution, ketene polymerizes according to a bimolecular law; within the limits of our experimental error (±20%) antioxidants such as hydroquinone do not inhibit the reaction and peroxides such as ascaridole or acetone peroxide do not accelerate it. There is a wide variation in the rate of polymerization depending on the solvent used and there seems to exist a rough parallelism between the rate and the dielectric constant of the solvent; solvents of high dielectric constant favor polymerization. The activation energy of the polymerization process in acetone is 11,000 cal.