86 SUBSTJ^^CES SOLUBLE LN WATER. 



by means of an indiarubber tube with the bent glass tube a. The 

 longer leg of the latter is similarly connected with a second bent- 

 glass tube c, communication being regulated by a clip at h. The 

 clip being opened, the atmospheric air in the flask is completely 

 expelled by boiling the liquid down to at least one-fourth of its 

 original bulk. The end of the glass tube c is then introduced 

 into a i^recii^itating glass containing about 30 cc. of a concentrated 

 solution of ferrous chloride, and, after allowing a little steam to 

 escape, the clip at h is closed and the lamp removed. As soon as 

 a partial vacuum has been produced in A, the clip is cautiously 

 opened, and about 20 cc. of the iron solution allowed to enter. 

 The precipitating glass is then filled with hydrochloric acid 

 (sp. gr. 1'12), and 25 to 40 cc. introduced in a similar manner, so as 

 to sweep the iron solution out of the tubes into the flask. After 

 closing the end of the tube c with an indiarubber stopper, it is 

 introduced into a mercury-bath and brought under a cylinder B, 

 j)reviously filled with mercury. The stopper is now removed, 

 and the flask again heated until the pressure in the interior is 

 nearly equal to that of the atmosphere. By opening the clip h, 

 and regulating the pressure with the finger and thumb, the 

 mercury is allowed to rise in the tube so as to drive most of the 

 hydrochloric acid into the flask ; it must, however, itself be 

 carefully prevented from passing into the latter. After the 

 external pressure has been overcome, the heat is so regulated that 

 half the liquid in the flask distils over in eight to ten minutes. It 

 is then certain that all the nitric oxide that has been formed has 

 been driven into B. The latter is provided with a glass-tap d, 

 and can be connected air-tight with a measuring tube /. After 

 cooling, the measuring tube filled with mercury is fitted on to 

 the cylinder, and the nitric oxide transferred to it by opening the 

 tap and sinking the cylinder. The amount of nitric acid may 

 then be calculated from the volume of nitric oxide found. ^ 



§ 101. Sclerotic, Cathartic Acid, etc. — If the total amount of 

 nitrogen in the aqueous extract (§ 96) is now found, on com- 

 parison, to be larger than that present as albuminoids, alkaloids, 

 ammonia and nitrates already estimated, the excess may be 

 reasonably ascribed to certain albuminous substances not pre- 



^ On the estimation of nitric acid in cultivated plants, see also Schloessing, 

 Journ, f. pract. Chem. lii. 142 ; Frühling und Grouven, Landwirthsch. 

 Versuchs.st. ix. 9, and 150 (1867) ; Keichardt, Zeitschr. f. anal. Chem. ix. 24 

 (1870) (Journ. Chem. Soc. xxiv. 439). 



