Maech 14, ]913] 



SCIENCE 



399 



respect are not essentially different from 

 those of Weitzel.^" In the clinical feeding 

 of infants sodium benzoate has been fre- 

 quently added to milk. 



According to Amberg and Loevenhart^^ 

 the activity of lipase, as measured by the 

 splitting of ethyl butyrate, is not dimin- 

 ished by the presence of 0.1 per cent, of 

 sodium benzoate. Dakin, working in Her- 

 ter's laboratory, has made a careful rede- 

 termination of many of the digestive con- 

 stants in presence of benzoate. ^^ 



(c) We now come to the most important 

 part of the subject, the behavior of benzoic 

 acid with reference to general health and 

 metabolism. At the time when this sub- 

 stance was extensively used in medical 

 practise, that is, from 1875 to 1880, it was 

 recognized by some physiologists that large 

 doses were followed by increased elimina- 

 tion of nitrogen, which, it was assumed, 

 must come from the breaking down of body 

 proteins. B. Salkowski, especially, from 

 experiments on dogs,^^ concluded that high 

 doses might occasion a considerable loss in 

 man. But in the dog experiments the in- 

 gested benzoate amounted in the mean to 

 about one third of a gram per kilogram of 

 body weight, which proportion if applied 

 to a man of 50 kilograms weight would call 

 for nearly 17 grams of benzoate, or 25 

 grams for a man of 75 kilograms weight. 

 Somewhat similar observations were made 

 by other physiologists, but on the other 

 hand the reports from clinical practise 

 failed to show any such losses. To follow 

 these discrepant observations farther is not 

 necessary in this place, as the question of 

 the increased nitrogen excretion has been 

 pretty fully handled in the investigations 



" Arbeiten aus dem Tcais. Gesundheitsamt, 19 : 

 1902. 



'^Jour. Biol. Chem., 4: 1908. 



"Herter, Jour. Am. Med. Assoc, 54: 1774. 



^Virehow's Archiv, 78. 



of Magnus-Levy,'* Ringer, and Epstein 

 and Bookman,^'' and others already re- 

 ferred to. 



But these early reports have had one 

 very important effect, which must be re- 

 called here. They left the impression that 

 the ingestion of sodium benzoate is in gen- 

 eral followed by increased protein metabol- 

 ism, tissue metabolism possibly, an unde- 

 sirable result, and this statement is fre- 

 quently repeated as applicable to all doses 

 of benzoate. A number of lengthy metab- 

 olism experiments have shown that for 

 ordinary ingestions of benzoate this in- 

 creased protein metabolism does not occur. 

 In the last few years the results of several 

 such investigations have been published. 

 One of these investigations was conducted 

 under the auspices of the Bureau of Chem- 

 istry of the United States Department of 

 Agriculture, and from it the conclusion 

 was drawn that small doses of benzoic acid 

 or benzoates exert a harmful action on 

 man, a slight loss in weight being affirmed 

 in some cases. It is not my purpose to 

 criticize this work here beyond saying that 

 the published data do not seem to warrant 

 the conclusions drawn, which opinion is 

 shared in a lengthy review of the work by 

 K. B. Lehmann, recently published.^' 



I wish to speak more particularly of the 

 results of the extended studies carried out 

 by Chittenden, Herter and myself, as mem- 

 bers of a commission appointed by the Sec- 

 retary of Agriculture to investigate the 

 question anew.'^ In Herter's work four 

 men were observed through periods of four 

 months, while in the investigations of Chit- 

 tenden and myself six men on a definite 



" Loc. cit. 



^^ Jour. Biolog. Chem., 10. 

 " Chemiker Zeitung, November 28, 1911. 

 "Report No. 88, U. S. Department of Agricul- 

 ture, 1909. 



