THE URINE. 411 



only very small amounts may make its detection sometimes very 

 difficult and laborious. A urine containing albumin must first 

 have the albumin removed by coagulation with acetic acid and heat 

 before it can be tested for sugar. 



The tests which are most frequently employed and are especially 

 recommended are as follows: 



TROMMER'S test. In a typical diabetic urine or one rich in 

 sugar this test succeeds well, and it may be performed in the 

 manner suggested on page 409. This test may lead to very great 

 mistakes in urines poor in sugar, especially when they have at the 

 same time normal or increased amounts of physiological constitu- 

 ents, and therefore it cannot be recommended to physicians or to 

 persons little practised in such work. Normal urine contains 

 reducible substances, such as uric acid, creatinin, and others, and 

 therefore a reduction takes place with all urine on using this test. 

 We generally do not have a separation of copper suboxide, but still 

 if we vary the proportion of the alkali to the copper sulphate and 

 boil, we often have an actual separation of suboxide in normal 

 urines, or we obtain a peculiar yellowish-red liquid due to finely- 

 divided hydrated suboxide. This occurs especially on the addition 

 of much alkali or too much copper sulphate, and by careless ma- 

 nipulation the inexperienced worker may therefore sometimes 

 obtain apparently positive results in a normal urine. Also as urine 

 contains substances such as creatinin and ammonia (from the 

 urea), which in the presence of only little sugar may keep the cop- 

 per suboxide in solution, therefore in inexperienced hands small 

 amounts of sugar may be overlooked. 



TROMMER'S test may indeed be made positive and useful, even in 

 the presence of very small amounts of sugar, by using the modifica- 

 tion suggested by WORM MULLER. As this modification is rather 

 complicated, and besides this requires much practice and exactness, 

 it is probably rarely employed by the busy physician. The follow- 

 ing test is to be preferred : 



ALMEN'S bismuth test, which recently has been incorrectly called 

 NYLANDER'S test, is performed with the alkaline bismuth solution 

 prepared as above described (page 410). For each test 10 c. c. of 

 urine are taken and treated with 1 c. c. of the bismuth solution and 

 boiled for a few minutes. In the presence of sugar the urine 

 becomes darker yellow or yellowish brown. Then it grows darker^ 

 cloudy, dark brown, or nearly black, and non-transparent. After a 

 shorter or longer time a black deposit appears, the supernatant 

 liquid gradually clears, but still remains colored. In the presence 

 of only very little sugar the test is not black or dark brown, but is 

 only deeper-colored, and after a longer time we only see on the 

 upper layer of the phosphate precipitate a dark or black edge (of 

 bismuth ?). In the presence of much sugar a larger amount of 



