URINE. 331 



A positive result from the above manipulation simply indicates 

 ; the possible presence of diacetic acid. Before making a final de- 

 cision regarding the presence of this body make the two following 

 control experiments : 



(a) Place 5 c.c. of urine in a test-tube -and boil it vigorously for 

 3-5 minutes. Cool the tube and, with the boiled urine, make the test 

 as given on p. 330. As has been already stated, diacetic acid yields 

 acetone upon decomposition and acetone does not give a Bordeaux- 

 reel color with ferric chloride. By boiling as indicated above, there- 

 fore, any diacetic acid present would be decomposed into acetone 

 and carbon dioxide and the test upon the resulting fluid would be 

 negative. If positive the color is due to the presence of bodies 

 other than diacetic acid. 



(b) Place 5 c.c. of urine in a test-tube, acidify with H 2 SO 4 , 

 to free diacetic acid from its salts, and carefully extract the mix- 

 ture with ether by shaking. If diacetic acid is present it will be 

 extracted by the ether. Now remove the ethereal solution and 

 add to it an equal volume of dilute ferric chloride; diacetic acid is 

 indicated by the production of the characteristic Bordeaux-red color. 

 This color disappears spontaneously in 24-48 hours. Such sub- 

 stances as antipyrin, kairin, phenacetin, salicylic acid, salicylates, 

 sodium acetate, thiocyanates and thallin yield a similar red color 

 under these conditions, but when due to the presence of any of 

 these substances the color does not disappear spontaneously but 

 may remain permanent for .days. Many of these disturbing sub- 

 stances are soluble in benzene or chloroform and may be removed 

 from the urine by this means before extracting with ether as above. 

 Diacetic acid is insoluble in benzene or chloroform. 



2. Arnold-Lipliawsky Reaction. This reaction is somewhat 

 more delicate than Gerhardt's test ( i ) and serves to detect diacetic 

 acid when present in the proportion of I 125,000. It is also negative 

 toward acetone, /?-oxybutyric acid and the interfering drugs men- 

 tioned as causing erroneous deductions in the application of Ger- 

 hardt's test. If the urine under examination is highly pigmented 

 it should be partly decolorized by means of animal charcoal before 

 applying the test as indicated below. 



Place 5 c.c. of the urine under examination and an equal volume 

 of the Arnold-Lipliawsky reagent 1 in a test tube, add a few drops 



1 This reagent consists of two definite solutions which are ordinarily preserved 

 separately and mixed just before using. The two solutions are prepared as 

 follows : 



(a) One per cent aqueous solution of potassium nitrite. 



