CONSTITUENTS OF COFFEES. 905 
loid at 100° C. as is usual, and claims that it commences to sublime. at 
79°C. The specific gravity! of caffeine at 10° C.is 1.23. The follow- 
ing reactions are given by Luchini:? Wenzell’s reagent (solution of 1 
part KMnO, in 200 parts of hot H.SO,), gives an amethyst color with 
caffeine, which changes to a dark violet, then becomes blood red and 
after twenty-four hours a brown precipitate forms: 1 part caffeine in 
10,000 can be detected by this reagent. With Luchini’s reagent (a 
hot solution of K,Cr,O-; in concentrated H,SO, there is no change on 
standing twenty-four hours. According to Béttiger* caffeine may be 
detected by evaporating an alcoholic extract of the substance to dry- 
ness, treating with hydrochloric acid, again drying, then adding water. 
The presence of the alkaloid is shown by a purple-red color, Schwar- 
zenbach* employs chlorine water instead of hydrochloric acid in the 
above test. The purple residue becomes yellow on heating and red 
when treated with ammonia. 
Kornauth® evidently overestimates the percentage of caffeine in 
coffees. He states that a percentage of caffeine below 1.97 is an indi- 
cation of an adulterated sample. The writer’s analyses and those quoted, 
in the table (p. 902) all show percentages of caffeine considerably below 
the limit given by Kornauth. 
The fat is a prominent constituent of coffee. Husemann® separated 
a white, odorless fat on cooling an alcoholic extract of coffee. The 
melting point of the fat was 37.5° C.; it consists? of the glycerides of 
palmitic acid and of an acid (CyH,,O,). The percentage of fat in raw 
coffee is approximately 15. 
The ecaffetannie acid of coffee has been isolated by Pfaff* and by 
Rochleder2 The latter chemist also reported the presence of traces of 
citric acid. Caffetannic acid as isolated by W. H. Krug in connection 
with this Department’s investigations agrees with the description given 
~by Beilstein.” Itis a light yellow, sticky mass, with a slightly acid and 
astringent taste. According to Hlasiwetz" it is a yellow bitter mass 
of the formula C,;H),Ox. 
Boussingault” found in a sample of coffee 2.21 per cent mannite, 8.73 
per cent invert sugar, and 2.357 per cent sucrose. 
The carbohydrates of the coffee berry have received very little 
1 Ann. Chem. Pharm., 1, 17. 
2 Arch d.Pharm. [3 R.], 23, 684; Ztschr. anal. Chem. 25, 565. 
3 Pol. Notizblitter, 1873, 257; Ztschr. anal. Chem. 1873, 442. 
4Chem. Centrbl., 1861, 989; Ztschr. anal. Chem. 1, 229. 
®Mittl. a.d. Pharm. Inst. n. Lab. f. angew. Chem., Erlangen, Heft. 3, 1-56. 
® Pflanzenstoffe, p. 1367. 
7 Wien. Akad. Ber., 24, 40. 
8’ Schweigg. Journ., 62, 31. 
* Wien. Acad. Ber., 7, 815. 
'0Organische Chemie, Beilstein, 3, 342. 
" Hlasiwetz, Ann. Chem. Pharm., 149, 219. 
2 Compt. rend., 91, 639. 
