ESTIMATION OF CARBOHYDRATES 225 



(ii) Pavy's Method. 



Owing to the difficulty experienced in determining the end point in 

 Ferrling's method, especially in the case of the estimation of glucose 

 in liquids, such as urine, owing to the formation of ammonia, which 

 prevents the precipitation of cuprous oxide (cf. p. 191), the use of 

 ammonium salts was introduced by Monier. A practical method 

 for estimating glucose was worked out by Pavy. 

 Procedure. 



50 c.c. of Pavy's solution (see p. 615) are measured with a burette 

 into a 200 c.c. conical flask. The flask is closed by a cork with two 

 holes ; through one of these the end of the burette containing the 

 sugar solution is passed and through the other an escape tube to carry 

 off steam and ammonia. To prevent the ammonia fumes coming into 

 the air, Pavy fitted to the escape tube a U-tube containing pumice and 

 sulphuric acid, but it is most convenient to fit. a valve as described 

 by Allen. This consists of a short length of rubber tubing closed at 

 its end by a piece of glass rod and cut near the end with a V-shaped 

 slit. This arrangement is preferable to the valve described in 1904 

 by Kumagawa and Suto. The end of the valve is placed in a dilute 

 solution of sulphuric acid, which is renewed when the acid is neutralised. 



The solution is boiled to drive out the air, which readily oxidises 

 ammoniacal cuprous solutions, and the sugar solution (0*5 to I c.c. at 

 a time) is gradually run in until the blue colour is discharged, the 

 solution being kept boiling throughout to exclude air. Sufficient 

 time must be allowed for the reduction to take place, as it is slower 

 than with Fehling's solution. The valve prevents any liquid being 

 sucked back, if the sugar solution be run in so quickly that boiling is 

 stopped. 



Just as in Fehling's method, the sugar solution must be of such a 

 strength that 10 c.c. = 50 c.c. of Pavy's solution. The titration must 

 be carried out rapidly and must be completed within three minutes, 

 otherwise the ammonia is all evolved before the titration is completed 

 and cuprous oxide is deposited. The boiling must not be interrupted 

 and the sugar solution must be run in at such a rate that the solution 

 is kept boiling the whole time. The final estimation should be made 

 by running in rather less (-5 c.c.) than the amount required and finishing 

 off more slowly. The minimal amount which is found to reduce the 

 Pavy's solution completely, when added at one time, is the exact volume 

 of the solution required. 



The amount of sugar in the solution is calculated from 

 10 c.c. Pavy's solution = 0*005 gm. glucose. 



The method has been compared against other methods by Kinoshita 

 who finds it very accurate. 



