GLUCOSE TESTS. 185 



388. Test diabetic urine with 



1. Trommels test (26). 



2. Fehling's test (27). 



3. Bismuth subnitrate test or Nylandar's test (30). 



4. Phenyl-hydrazin test (31). 



5. Fermentation test (32). 



Notice that the other urinary constituents may mod- 

 ify the results obtained with the solution of pure glucose. 



For the detection of glucose in urine the tests above given 

 may be employed. No one of these, however, is an absolute proof 

 of the presence of glucose. Other constituents of urine have a 

 slight reducing power, and may respond to the tests with alkaline 

 solutions of copper or bismuth, where the action is that of re- 

 duction, for example, uric acid and its salts; creatinin, mucin, and 

 others occurring in smaller amounts have this power of reduction 

 and will reduce Trommer's and Fehling's reagents. The same 

 is true of many medicines which pass into the urine. Trommer's 

 and Fehling's tests are very sensitive under ordinary conditions, 

 but they may fail in some decomposing urines, the ammonia which 

 is present keeping the cuprous oxid in solution. Long boiling will 

 expel the ammonia, and the test may then succeed. Large amounts 

 of uric acid, creatinin, or albumin may act in the same manner, 

 keeping the red oxid from precipitating. It must be borne in mind 

 that the earthy phosphates will always be precipitated when the 

 urine is made alkaline, and consequently, appear in many of the 

 glucose tests. They are always colorless, as can be shown by 

 washing them on a filter, whereas the oxid of copper or bismuth 

 reduced by the sugar are colored. 



In the test with the subnitrate of bismuth the salt is not so 

 easily reduced by other compounds than glucose. Consequently 

 there is not so much danger of mistaking these for sugar. With 

 a very large excess of the alkali this reduction may occur. This 

 is said not to be the case with Nylander's modification of the test 

 (Experiment 30). Albumin is, however, decomposed under such 

 circumstances, giving a black precipitate. It must, therefore, be 

 removed from the solution before the test is made. With this test 



