74 P. C. Ray — Ohemiml Eosamination of [No. 1, 



N 

 .-. 250 c, c. CaO sol. =15 c. c. YqKMiiO^ 



or 50 glycerin solution =15,, „ ,, 

 or 250 „ ., =75,, ., 



N 

 But 1 c. c. — KMnO.^ =00046 gm. glycerin. 



J, , . 1 . 75x0-0046 x102 ^.. 

 Percentage ot glycerin = — — = j0"9. 



2, 1*704 gm. oil were treated exactly as above, the glycerin solu- 

 tion made up to 500c. c. of which 100c, c. were oxidised to oxalic acid. 

 The CaO dissolved in dilute HgSO^ was made up to 250 c. c. 



N 

 50 c. c. CaO sol. = r6c. c, — KMnO, 



10 * 



.-, 250 c. c. „ „ =8-0 „ „ 



or 100 c. c. glycerin sol, =8*0 „ ., ,, 



.-.500 „ '„ „ =40-0 „ „ 



„ , „ , . 40x0-0046x102 _- 

 Percentage 01 glycerin = — - — = 10*8. 



The percentage of glycerin in niger-seed oil is thus practically the 

 same as in mutton-fat. 



Estimation op Glycerin in Butter-pat. 



The oxidation of glycerin by the alkaline jDermanganate is not 

 applicable in the case of butter-fat, as the soluble fatty acids, e, g., 

 butyric, caproic, &c., it contains yield notable quantities of oxalic acid 

 under the same treatment. The same remarks apply to the case of 

 cocoanut oil (See Chem. Neics, Vol. LXIII, p, 251). 



NOTE ON THE FOX AND WANKLYN METHOD OF 

 ESTIMATING GLYCERIN. 



This method, although it yields accurate results, can scarcely be 

 made use of by the ordinary commercial analysist on account of its 

 tedious and troublesome natui'e. Moreover, the manganese precipitate, 

 sometimes bulky, cannot be properly Avashed without the aid of a 

 Bunsen's filter-pump. The details recorded above will show that each 

 determination of glycerin involves steady work of several hours. If 

 oi'dinaiy alcohol be used as a solvent for the fats and oils, saponification 

 is easily effected, but there is considerable risk of the loss of glycerin 

 during the evaporation of alcohol. 



