ESTIMATION 



eliminated interfering substances by precipitation with lead acetate, 

 and then recovered the riboflavine by adsorption on fuller's earth and 

 elution with aqueous alcoholic alkali. The fluorescence was measured 

 after treatment with potassium permanganate solution, and again 

 after destruction of the riboflavine by treatment with hot o-i iV sodium 

 hydroxide. The value thus obtained was subtracted from the first 

 value, the difference being proportional to the riboflavine concentration. 



M. Fujiwara and H. Shimizu *0" removed fluorescent impurities with 

 a higher adsorption affinity than riboflavine on a column of pyridine- 

 treated zeolite and then adsorbed riboflavine and impurities with the 

 same adsorption affinity on a column of a phenolsulphonic acid resin, 

 KH9, from which the riboflavine was eluted with a mixture of pyridine 

 and acetic acid. 



Some inorganic ions were said ^^ to have a quenching effect on the 

 fluorescence of riboflavine ; the fluorescence was found to be maximal 

 within the pK range 5-9 to 77. E. C. Slater and D. B. Morell ^^ 

 modified Najjar's method by introducing an " internal standard " to 

 correct for the quenching of the fluorescence by other constituents of 

 biological extracts and also to check the specificity of the method by 

 exposing test solutions for short controlled periods to sunlight. The 

 rate of photochemical decomposition was dependent on the concentra- 

 tion of unidentified degradation products . Values obtained for urine 

 by this method were in agreement with those obtained by biological 

 assay. The addition of known amounts of riboflavine to test solutions 

 was also used by Scott et al.,^^ who also corrected for the presence of 

 stable fluorescent substances by reducing riboflavine with dithionite, 

 a procedure which Slater and Morell claimed to give high results. 



The treatment of biological materials preparatory to fluorimetric 

 assay generally comprises some form of hydrolysis. Meat samples 

 may be digested with clarase ** or with papain or taka-diastase,^^ and 

 cereals and other starchy foods with acid or taka-diastase. 



J. S. Andrews *^ published results obtained in collaborative assays 

 of different flours, including flours fortified with riboflavine. A 

 wide spread was noted in the results reported from different labora- 

 tories with both the fluorimetric and microbiological methods of assay. 

 Good recoveries were obtained in both methods, but the absolute 

 values obtained by the microbiological method were higher than those 

 obtained by the fluorimetric method. With enriched flours, direct 

 readings of the fluorescence of the extracts gave results as satisfactory 

 as those obtained when potassium permanganate or Florisil were 

 employed for the removal of impurities. 



Riboflavine nucleotides (page 191) can also be estimated fluori- 

 metrically. At high salt concentrations, the dinucleotide has 15 % 

 of the fluorescence of a corresponding amount of free riboflavine, but 

 II 161 



