THE CARBOHYDRATES. 279 



glucuronates and homogentisinic acid likewise give a positive reac- 

 tion, and ammonia, if present beyond traces, may hold in solution 

 any cuprous oxide that may be referable to very small amounts of 

 sugar. 



FERMENTATION TEST. This test, when controlled by Nylander's 

 test, is the most satisfactory one. To this end, a little compressed 

 yeast is shaken with about 20 c.c. of urine, and the mixture is 

 placed in a saccharimetric tube, such as that devised by Lohnstein 

 or Einhorn. On standing at the ordinary temperature of the room, 

 or, still better, at 37 C., fermentation occurs if glucose is present, 

 and the liberated carbon dioxide collects at the top of the tube. 

 In any case, however, two controls should be made, viz., one to 

 determine that the yeast is active, and another with normal urine. 

 If a small amount of sugar is present, it may happen that the 

 resulting carbon dioxide is absorbed. If in such a case Nylander's 

 test first gave a positive reaction, but no longer reacts after fermenta- 

 tion is complete (in twelve to twenty-four hours), the presence of 

 sugar may be inferred. If, on the other hand, no fermentation 

 occurs, and Xylander's test still gives a positive result, we may con- 

 clude that the reaction is due to the presence of a non-fermentable 

 reducing substance. 



PHENYLHYDRAZIN TEST. As has been pointed out, all mono- 

 saccharides and some of the disaccharides, such as maltose, isomal- 

 tose, and lactose, form compounds with phenylhydrazin which are 

 known as osazons (see page 55). The resulting bodies are all very 

 similar, but may be distinguished from each other by the melting- 

 point of their crystals, and to some extent also by their microscopical 

 appearance. With free glucuronic acid a similar compound may 

 be obtained, according to Thierfelder, which may also be recognized 

 by its melting-point, while the conjugate glucuronates are inactive in 

 this respect. Pentoses likewise give rise to the formation of osa- 

 zons, but the melting-point of the resulting crystals serves to distin- 

 guish these also from the osazons of the hexoses. As a general 

 rule, however, neither the pentoses nor glucuronic acid interferes 

 with the reliability of the test. If doubt should arise, a special 

 examination should be made to ascertain whether pentoses or glu- 

 curonates are present in amounts sufficient to react with the reagent. 

 A further objection to the phenylhydrazin test has been urged on 

 the basis that its delicacy is such that a positive reaction is obtained 

 even under normal conditions. This, how r evcr, I must deny. 



The test is conveniently conducted as follows s 5 drops of pure 

 phenylhydrazin are mixed in a test-tube with 10 drops of glacial 

 acetic acid and 1 c.c. of a saturated solution of common salt. To 

 this are added 3 c.c. of urine, when the mixture is boiled for two 

 minutes and is then set aside to cool. In the presence of more 

 than 0.5 per cent, of glucose, crystals of phenyl-glucosazon begin 

 to separate out after one or two minutes. Should smaller amounts 

 be present, it is necessary to wait. The sediment is then exam- 



