working up in a customary manner there were obtained 0.66 g. of triphenylbismuth, and 0.15 g. of a product which melted at 179-180°( with prior shrinking) after crystallization from chloroform-methanol.Anal. Caled, for CisH^OBi: Bi, 45.83. Found: Bi, 40.98.

No abstract

No abstract

No abstract

ANDThe only methods heretofore described for predicting carcinogenic activity have been limited to series of pure aromatic compounds or to high]>-specialized types of petroleum products. After examination of a number of different reactions and properties of polycyclic aromatic compounds, five nonbiological methods have been developed for the prediction of tumor potency of high boiling petroleum products. Each method gave about the same degree of correlation with biological data. These physical-chemical methods are all based upon measurements of certain types of polycyclic aromatic hydrocarbons. The results support the premise t h a t the carcinogenicity is a function qf the concentration of these types of compounds in oils. The methods, therefore, obviate the necessity of isolating or analyzing the actual carcinogenic compounds. The caffeine number method is preferred for routine analytical work. It is based upon a n ultraviolet examination of a n aqueous caffeine extract of the aromatic hydrocarbons. A standard deviation of 8.4 was obtained for the difference between the predicted and the biological tumor potency values of 123 different products. Reproducibility of the caffeine number was better than t h a t of the biological data. Comprehensive analytical data for all five methods are presented for a variety of high boiling petroleum products. These demonstrate t h a t a choice of methods is available, depending upon the particular types of products to be analyzed and the facilities a t hand.