1889.] 0)i a pure Fermentation of Mannite and Glycerin. 351 



the large quantity of sodium chloride. This separation was partially 

 accomplished by means of strong alcohol in which mannitol is 

 considerably more soluble than sodium chloride; 17*135 grams of 

 mannitol, containing 36"2 per cent, of ash (sodium chloride), were 

 obtained, corresponding to 10'93 grams of pure mannitol. 



Fermentation of Mannitol No. 2. 



A second fermentation of mannitol was conducted on the same 

 lines as above, the liquid remaining upwards of three months in the 

 incubator at 38 — 40° C, and then for another month at the ordinary- 

 temperature of the air before it was submitted to examination. 



Alcohol. — In the alcoholic distillate there was obtained after 

 repeated rectification a liquid of specific gravity indicating the pre- 

 sence of 9"519 grams of absolute alcohol, on further rectification the 

 specific gravity indicated 9*094 grams alcohol, and after dehydration 

 with fused carbonate of potash, 7*75 grams of liquid boiling at 

 79 — 80° C. were actually recovered. The amount of ethyl alcohol 

 actually produced in the fermentation may thus be estimated at 

 about 9 '3 grams. 



Volatile Acids. — From the several fractions of the volatile acids, 

 the following barium salts were prepared : — 



I. 6*3435 grams Ba-salt giving 91 '61 per cent. BaS0 4 . 

 II. 2*9525 grams Ba-salt giving 91 35 per cent. BaS0 4 . 



III. 0-8080 gram Ba-salt giving 93*88 per cent. BaS0 4 . 



IV. 2*0385 grams Ba-salt. (The Ba-determination in this salt 



yielded 116*2 per cent, of BaS0 4 , an obviously erroneous 

 figure : unfortunately the determination could not be 

 repeated, the remainder of the salt having been lost.) 



Remembering the calculated proportions of barium sulphate for — 



Barium acetate 91*37 per cent. 



,, formate 102*64 „ 



it is obvious that the fractions I and II consisted of pure barium 

 acetate, fraction III of barium acetate with a small admixture of 

 barium formate, whilst fraction IV, in which unfortunately no satis- 

 factory Ba-determination was obtained, doubtless contained a larger 

 proportion of formate. 



The presence of formic acid in fractions III and IV was further 

 demonstrable by means of the reaction with silver nitrate. 



The amount of barium acetate and formate respectively present in 

 mixtures of the two salts may be calculated from the following 

 formula — 



Bx = 100P + &aj-100&, 



VOL. XLVI. 2 B 



