56 INTRODUCTION. 



The solution may be prepared according to the following 

 formula, reduced to grains from the formula given by Ber- 

 nard : ' 



Of bitartrate of potash, 3 vi. gr. xxiij. 



Of crystallized carbonate of soda, 3 v. gr. ix. 



Dissolve in 3 vss. of water ; add to the solution 3 iij. gr. 

 li. of sulphate of copper, and boil ; allow the mixture to 

 cool and add 3 v. gr. ix. of potash dissolved in 3 iv. of water. 

 Add water till the whole measures 3 xvii. 



Maumen&s Test. Bottger^s Test. The first of these 

 tests is employed by saturating strips of some woollen tissue, 

 such as flannel, with a strong solution of bichloride of tin, 

 and drying. One of these strips is moistened with the sus- 

 pected liquid, and dried quickly by the heat of a fire or lamp. 

 If sugar be present, the strips will assume a brownish or Jblack 

 tint. 



Bottger's test depends upon the reduction of a salt of bis- 

 muth, analogous to the reduction of the copper in Trommer's 

 test. It is employed in the following way : We add to the 

 suspected liquid a few drops of a weak solution of the nitrate 

 of bismuth in nitric acid, render the whole alkaline by tfre 

 addition of a solution of carbonate of soda, and boil for three 

 or four minutes. If sugar be present, the bismuth will be 

 reduced, and form a dark precipitate. Neither of these tests 

 presents any advantage over Trommer's test, which is the one 

 most generally employed. 



Fermentation Test. With the exception of actual ex- 

 traction, this is the most certain test for sugar, and should 

 always be employed when the other tests leave any doubt 

 with regard to its presence. It depends on a property of 

 sugar whereby it is decomposed into alcohol and carbonic 

 acid in the presence of certain ferments, at a moderately ele- 

 vated temperature. The test is applicable to all varieties of 



1 BERNARD, Lemons de Physiologic Experimentalc, Paris, 1855, p. 34. 



