158 COAL-TAR COLOBS USED IN FOOD PRODUCTS. 



it is only necessary to multiply the weight of color by 5 and by 

 15.432. Doing so, the following data are obtained: 



A. Harmful colors produce bad e fects in doses of a "few" decigrams, i. e., mul- 

 tiples of a half gram; the word "few" is not sharply defined, but the rule means a 

 "few" times 7.7 grains. 



B. Harmless colors produce, when continuously fed, no bad effect in multiples of 

 10 of the doses of A, above, i. e., multiples of 77 grains. 



C. Slight diarrhea and slight, temporary albuminuria are not to count against the 

 color. 



Applying these rules to the preceding colors discloses that they were 

 apparently not so used, in selecting the harmless and harmful colors 

 in the Confectioners' List. 



LEHM ANN'S RULES. 



The Lehmann rules as just interpreted were applied to ths data 

 just given. As a result of such application the conclusion is reached 

 that of the 15 colors classed as harmless (foregoing classes I, II, and 

 III), 6 would be classed as doubtful by the Lehmann rules, namely: 



13. Ponceau 4 GB, 65. Fast Red B, 



8 and 9. Fast Yellow Y and R, 103. Azorubin S, 



55. Ponceau 2R, 105. Fast Red E, 



and the remaining 9 would have been classed as harmless. 



Of the 11 classed as harmful (foregoing classes I, II, III, and IV) y 



2 would have been regarded as harmless, namely: 



398. Naphthol Green B, 138. Fast Brown G, 



3 would have been classed as doubtful, namely: 



197. Bismarck Brown, 88. Diphenylamin Orange, 



11. Sudan I, 



and the remaining 6 would have been classed as harmful. 

 SANTORFS WORK AS A GUIDE TO A RULE. 



San tori regards the following six dyes as harmless: 457 (188), 

 467 (74), 477 (129.5), 480 (156), 599 (64), 692 (467). 



The first is the Green Table number and the bracketed figure the 

 average number of grains per 100 pounds per day for 30 days. 



In the case of Nos. 457 and 599 the animal was wholly normal 

 throughout the test and the autopsy showed only normal conditions. 



In the case of No. 477 the only disturbance was colored feces and 

 the autopsy showed only normal conditions. 



In the case of No. 480 the only disturbance was colored feces, but 

 the autopsy showed a pea-green kidney. 



In the case of Nos. 467 and 692 there was only vomiting; in the 

 case of No. 467 the autopsy showed everything normal; whereas in the 

 case of No. 692 the autopsy disclosed a swollen kidney. 



Therefore, according to Santori, a dye that causes colored feces, 

 even with colored kidney, is harmless. Also a dye that causes vomit- 



