SEPARATION AND IDENTIFICATION OP COCAIN. 43 



marked in its character and can be readily distinguished, though 

 until one is familiar with it a parallel test should always be run, 

 using pure cocain. 



The color reaction is also of interest and will prove valuable in 

 detecting the presence of other alkaloids. A purple color would 

 indicate that atropin, strychnin, or yohimbin were present, though 

 it is well known that a residue obtained by the above method from 

 the coca leaf will in some instances also give a purple color. Tropa- 

 cocain, benzoylecgonin, and aconitin will also give the ethyl benzoate 

 test, but the possibility of the presence of the first two can be elimi- 

 nated by a subsequent microscopic test, and from the fact that the 

 benzoylecgonin is not removed from the aqueous solution to any 

 great extent by petroleum ether. The latter fact is also true of 

 aconitin, but the presence of this very poisonous alkaloid would be 

 readily apparent when performing the physiological test. 



The portion of the material which had been reserved should be 

 tested physiologically. The petroleum ether is evaporated, a part 

 of the cool residue is placed on the tongue and rubbed gently with the 

 finger for about one minute, when, if cocain is present, a sensation of 

 numbness gradually develops, the tongue and often the lips which 

 have been touched by the finger will have a smooth, ivory-like feeling, 

 and the effect will be apparent for some time. This is another test 

 which should be performed in conjunction with one on a known 

 sample, unless the analyst is perfectly familiar with the sensation. 



.This residue may also be used for a microscopic test by removing 

 a small portion, placing it on a microscope slide and noting, under 

 the lens, the character of the crystals found with a drop or two of 

 gold chlorid solution. Cocain forms a definite crystalline compound 

 with gold chlorid, and the product obtained should be compared with 

 that given by a solution of the pure alkaloid. 



With a residue obtained originally as described, these tests will 

 substantiate the presence of cocain even though it occur in a minute 

 quantity, and the four reactions can be applied to very small residues, 

 although if it is desired the first mentioned with potassium-mercuric 

 iodid may be omitted. 



Another reaction for cocain which is very characteristic when con- 

 sidered in conjunction with the ethyl benzoate test is obtained by 

 transferring some of the residue to a pressure flask, adding a few 

 crystals of salicylic acid and about 15 cc of dilute hydrochloric acid, 

 and heating the flask, stoppered, for about an hour and a half over the 

 steam bath. On opening the flask the odor of wintergreen will be 

 very marked if cocain was originally present. Tropacocain does not 

 respond to this reaction, though it might indicate the presence of 

 cinnamyl cocain and the truxillins; all of the latter, however, are pre- 

 cluded if the mixture gives the ethyl benzoate test. 



