Excretion of Creatine in Carbohydrate Starvation. 207 



to eight readings were made and the mean was taken. The total nitrogen 

 was estimated by Kjeldahl's method. The aceto-acetic acid + acetone was 

 estimated by the Messinger-Huppert method. 



The substances tested were prepared in the following manner, and we wish 

 to thank Dr. Hurtley for very kindly supplying us with them. The 

 /3-oxybutyric acid was extracted from urine and the strength of the solution 

 was accurately known ; the solution was nearly colourless. The acetone 

 was chemically pure. The ethyl ester of aceto-acetic acid was obtained 

 by distilling the pure commercial ester under reduced pressure, and the 

 product boiled constantly at the correct boiling point for the pure substance. 



The aceto-acetic acid was obtained from the ester, which was hydrolysed 

 by adding the theoretical amount of normal caustic soda, and allowing it to 

 stand at the room temperature for 36 hours, when the hydrolysis was 

 complete. It will be seen that the solution used consisted of the sodium salt 

 of aceto-acetic acid and an equivalent amount of ethyl alcohol. The mixture 

 was diluted and the amount of acetone present was determined by Folin's 

 method (9). The amount of aceto-acetic acid corresponding to the acetone 

 present was deducted from the theoretical amount of aceto-acetic acid in 

 order to get the correct value for the aceto-acetic acid. The aceto-acetic 

 acid was kept in an ice chest in order to prevent its decomposition, and was 

 tested from time to time by means of the Folin and Messinger-Huppert 

 methods. 



/3-oxybutyric acid, acetone, the ethyl ester and the sodium salt of aceto- 

 acetic acid all give an orange colour when treated alone with picric acid and 

 soda, but on dilution the solution is much paler than the usual colour 

 obtained with urine under these conditions. These substances were then added 

 to urine in varying concentrations, and the colour produced by the addition 

 of picric acid and soda was compared with the colour produced by the urine 

 alone with the picric acid and soda, without the addition of these substances. 

 The addition of /3-oxybutyric acid produces practically no alteration in the 

 colour obtained by adding picric acid and soda to urine. 



Thus when added to the urine (Table I) in amounts corresponding to 

 - 036 grm. per 100 c.c. and 1 grm. per 100 c.c. it caused no error at all 

 in the creatinine determination. When present in amounts corresponding 

 to a 216-per-cent. solution it made the colour slightly lighter, causing a 

 difference of 0'27 mm. scale reading, which is almost within the limit of 

 accuracy of the method. The amounts of the /3-oxybutyric acid added 

 are quite comparable with those found in the urine in diabetes, and as the 

 error caused even by large amounts is so small its effect may be safely 

 neglected. 



Q 2 



