206 Messrs. G. Graham and E. P. Poulton. The Alleged 



slightly darker colour than was obtained with the urine alone, but that 

 the colour soon faded and caused no error in the determination of 

 creatinine. Krause (16), Wolf and Osterberg (22), and Eose (20) found 

 that the addition of the ethyl ester of ace to-acetic acid to urine did not 

 produce any error in the estimation of creatinine, unless large amounts 

 {i.e. over 1 per cent.) were added. They seem to have assumed that the 

 action of aceto-acetic acid would be the same as that of the ester. 



Recently, however, Greenwald (12), working on diabetic urines, has shown 

 that if aceto-acetic acid is added to urine directly a considerable error is 

 introduced into the estimation of creatinine. 



This observation may possibly explain why we did not find any creatine in 

 our experiment on the fat diet. Folin (8) originally stated that acetone 

 and aceto-acetic acid gave the orange colour with picric acid and soda, 

 but remarked that they could easily be removed from the urine. In our 

 experiment we removed the aceto-acetic acid as far as possible from the 

 urine before making the estimation, in order to get rid of any disturbing 

 effect that the acetone bodies might have on the creatinine figures. 



In the experiments described in this paper we have studied this question 

 in greater detail and also the means of overcoming the difficulty. 



I. The Effect of Aceto-acetic Acid on the Estimation of Creatinine. 



The different intensities of colour produced by /3-oxybutyric acid, acetone, 

 aceto-acetic ester and aceto-acetic acid when treated with picric acid and 

 soda were first investigated. 



The importance of aceto-acetic acid is emphasised by the experiments of 

 Arnold (1), Emden (7), and Hurtley (14). These investigators have, 

 independently, pointed out that in cases of acidosis the fresh urine contains 

 only small amounts of acetone, while aceto-acetic acid may be present in 

 large amounts. 



Throughout our experiments the estimation of the creatinine was performed 

 in the usual way ; 15 c.c. of a saturated solution of picric acid and 5 c.c. of 

 10-per-cent. caustic soda were added to 10 c.c. of urine, the mixture was 

 allowed to stand for seven instead of five minutes, and then diluted to 

 500 c.c. with water. Polin (8) stated that the maximum intensity of colour 

 occurs in five to nine minutes after mixing. We have always waited seven 

 minutes because we found that five minutes was not always sufficient if 

 the urine was slightly diluted, as occurs in the estimation of the 

 creatinine + creatine by this method. The matching was done with a 

 Duboscq colorimeter against an N/2 potassium bichromate solution. All 

 the matching was done by E. P. P. while the scale was read by G. G., six 



