PRACTICAL EXERCISES 489 



2 decigrammes of sodium acetate. It is sufficiently accurate to add 

 as much phenyl-hydrazine as will lie on a sixpence (or a dime) and 

 twice as much sodium acetate. Heat the test-tube in a boiling water- 

 bath for half an hour. Then cool at the tap and examine the deposit 

 under the microscope for the yellow phenyl-glucosazone crystals 

 (Fig. 181). Sometimes the osazone precipitate is amorphous. Lest 

 this should be the case, the precipitate, if no crystals can be seen, 

 must be dissolved in hot alcohol. The solution is then diluted 

 with water and the alcohol boiled off, when the osazone, if any be 

 present, will crystallize out. Very minute traces of sugar can be 

 detected in this way (as little as OT per cent, in urine). Often 

 in normal urine yellow crystals are deposited during the first fifteen 

 minutes' heating. They must not be mistaken for glucosazone. 

 They probably consist of a compound of glycuronic acid and phenyl- 

 hydrazine. They are changed as the heating goes on into an amor- 

 phous brownish-yellow precipitate (Abel). 



(c) The Yeast Test is an important confirmatory test for dis- 

 tinguishing the fermentable sugars from other reducing substances, 

 but it is not very delicate, and will with difficulty detect sugar 

 when less than 0^5 per cent, is present. It can be performed thus : 

 A little yeast (the tablets of compressed yeast do very well) is added 

 to a test-tube half filled with urine. The test-tube is then filled up 

 with mercury, closed with the thumb, and inverted over a dish 

 containing mercury. The dish may be placed on the top of a water- 

 bath whose temperature is about 40 C. After twenty-four hours 

 the sugar will have been broken up into alcohol and carbon dioxide. 

 The latter will have collected above the mercury in the test-tube, 

 and the former will be present in the urine. The tests for sugar 

 will either be negative or will be less distinct than before. A con- 

 trol test-tube containing water and yeast should also be set up, 

 as impurities in the yeast sometimes yield a small amount of carbon 

 dioxide. Specially constructed tubes are also often used for per- 

 forming the test. 



(2) Quantitative Estimation of Sugar in Urine. (a) Volumetrically, 

 the sugar can be estimated by titration with Fehling's solution. As 

 this does not keep well, two solutions containing its ingredients 

 should be kept separately and mixed when required. Solution I. : 

 Dissolve 34*64 grammes pure cupric sulphate in distilled water, and 

 make up the volume to 500 c.c. Solution II. : Dissolve 173 grammes 

 Rochelle salt in 400 c.c. of water, add to this 51 '6 grammes sodium 

 hydroxide, and make up the volume with water to 500 c.c. Keep in 

 well-stoppered bottles in the dark. For use, mix together equal 

 volumes of the two solutions. Ten c.c. of this mixture is reduced by 

 0*05 gramme dextrose. To estimate the sugar in urine, put 10 c.c. 

 of the mixture into a porcelain capsule or glass flask, and dilute it 

 four or five times with distilled water. Dilute some of the urine, 

 say ten or twenty times, according to the quantity of sugar indicated 

 by a rough determination. Run the diluted urine from a burette 

 into the Fehling's solution, bringing it to the boil each time urine is 

 added, until, on allowing the precipitate to settle, the blue colour 

 is seen to have entirely disappeared from the supernatant liquid. 

 The observation of the colour must be made while the liquid is still 

 hot. 



Suppose that 10 c.c. of Fehling's solution is decolourized by 20 c.c. 

 of the ten-times diluted urine. Then 2 c.c. of the original urine 

 contains 0*5 gramme dextrose. If the urine of the twenty-four hours 



