VII. ESTIMATION 371 



Siinunarij 



The author lias attempted to point to some of the faults of existing meth- 

 ods for riboflavin assay as a stimulus for fiu'ther development in the field 

 rather than a condemnation of the present procedures. Many of the present 

 methods are adequate to obtain the kind of knowledge most generally 

 sought, which is a more or less quantitative idea of the riboflavin content 

 and with special interest in high-potency materials. Methods designed for 

 a special tissue, serum for example, may be quite satisfactory. It is not to 

 be assumed, however, that success with one material, or even several, 

 guarantees the method for general use. The following report may well be- 

 come a classic example of what may befall the unsuspecting worker who 

 makes a simple analysis. Baude et al.^^ reported that the riboflavin content 

 of sow's milk as determined bj^ a fluorometric method designed for cow's 

 milk^^ was quite low, approximately 0.5 7 per millihter, and more stable 

 to light than the riboflavin in cow's milk." Microbiological studies in an- 

 other laboratoiy failed to confirm this figure, and this discrepancy led to a 

 collaborative effort^^ to uncover the difficulty. The same sample of milk 

 was analyzed by several methods with the following results : original fluoro- 

 metric assay, 0.4 7 per milHliter; microbiological assay, ^^ 1.2 to 1.5 7 per 

 millihter; chick assay, 1.2 to 1.8 7 per milhUter; and rat assay, 1.6 to 2.0 

 7 per milliliter. It was provisionally decided that the fluorometric method 

 did not jaeld complete extraction, and the sample was then hydrolyzed with 

 H2SO4 at pH 1 for an hour, incubated with takadiastase, and the assay com- 

 pleted by the method of Slater and Morell. This also yielded low results, 

 but it was found that if the final extraction into butanol-pyridine was omit- 

 ted higher values were obtained. Thus it appeared that the riboflavin did 

 not behave normally and did not completely enter this solvent. Finally, an 

 assay by the method of Bessey, Lowry, and Love was said to yield results 

 comparable to those of the microbiological assay. Unfortunately the story 

 is not complete at the present time, but it appears that at least part of the 

 riboflavin of sow's milk is present in a form which does not behave as does 

 previously known compounds. It should be borne in mind that this example 

 is with a material which does not yield highly colored extracts and that 

 cow's milk has been analyzed successfully for riboflavin many times by 

 fluorometric methods. 



" R. Baude, S. K. Kon, and S. Y. Thompson, /. Dairy Research 14, 414 (1945-46). 



" K. M. Henry, J. Houston, and S. K. Kon, Biochem. J. 34, 607 (1940). 



" R. Baude, M. E. Coates, K. M. Henry, S. K. Kon, S. J. Rowland, S. Y. Thomp- 

 son, and D. M. Walker, Brit. J. Nutrition 1, 64 (1947). 



68 V. E. Davis, R. MacVicar, C. B. Ross, C. K. Whitehair, A. A. Heidebrecht, R. 

 Baude, M. E. Coates, K. M. Henry, S. K. Kon, S. Y. Tompson, and F. Wilby, 

 Nature 165, 522 (1950). 



" H. A. Kornberg, R. S. Langdon, and V. H. Cheldelin, Anal. Chem. 20, 81 (1948). 



