A PROBLEM which is occasionally presented to the chemist engaged in the examination of various tar products is the determination of actual phenol (C6HsOH) in the substance under consideration. Usually all that is required is the estimation of total phenols, generally called "tar acids." These tar acids consist of a complex mixture of phenol, cresol, and various higher homologues, and their quantitative estimation in toto is always based on their common chemical property of solubility in a solution of caustic alkali. Sometimes, however, it is necessary to proceed further, and ascertain how much of these tar acids is real phenol.
At present, so far as the writer has been able to learn, there is no satisfactory solution of this problem in the li'terature. Many tests for qualitative distinction of phenol, cresol, and the higher homologous oxygenated aromatic hydrocarbons have been proposed. Some of these are of very doubtful value, even qualitatively, but none are all susceptible of even approximate quanlitative application.
The only real attempt in this field was that of Charles Lowe (see Lunge's Coal Tar and Ammonia, Fourth Ed.,. This is proposed for crude .carbolic acid, and is more a specification for material than a method of analysis. Briefly, it consists in plain distillation in a retort, with no attempt at fractionation, and subsequent determination of the melting point of an arbitrary fraction. This fraction is a definite volume per cent. after removal of the water. The percentage of phenol is determined by adjusting a mixture of phenol and cresol (mixture of the three isomers) to have the same melting point. This is a method 1 Communicated by the author. See also "Note on the Rideal-Walker