162 poisons : THEIR effects and detection. [§ 194. 
observed after death in the blood of those poisoned by chloroform, but 
it is doubtful whether the bubbles are not merely those to be found in 
any other corpse—in 189 cases, only eighteen times were these gas- 
bubbles observed, 1 so that, even if they are characteristic, the chances 
in a given case that they will not be seen are greater than the reverse. 
The smell of chloroform may be present, but has been noticed very seldom. 
§ 194. The detection of chloroform from organic substances is not 
difficult, its low boiling-point causing it to distil readily. Accordingly 
(whatever may be the ultimate modifications, as suggested by different 
experimenters), the first step is to bring the substances, unless fluid, 
into a pulp with water, and submit this pulp to distillation by the heat 
of a water-bath. If the liquid operated upon possesses no particular 
odour, the chloroform may in this way be recognised in the distillate, 
which, if necessary, may be redistilled in the same manner, so as to 
concentrate the volatile matters in a small compass. 
There are four chief tests for the identification of chloroform :— 
1. The final distillate is tested with a little aniline, and an alcoholic 
solution of soda or potash lye ; either immediately, or upon gently 
warming the liquid, there is a peculiar and penetrating odour of phenyl- 
carbylamine, C 6 H 5 NC ; it is produced by the following reaction :— 
CHC1 3 +3K0H+C 6 H 5 NH 2 C 6 H 5 NC+3KC1+3H 2 0. 
Chloral, trichloracetic acid, bromoform and iodoform also give the same 
reaction ; on the other hand, ethylidene chloride does not yield under 
these circumstances any carbylamine (isonitrile). 
2. Chloroform reduces Fehling’s alkaline copper solution, when 
applied to a distillate, thus excluding a host of more fixed bodies which 
have the same reaction ; it is a very excellent test, and may be made 
quantitative. The reaction is as follows :— 
CHCI3 +5KH0+2Cu0 = Cu 2 0+K 2 C0 3 -f 3KC1+3H 2 0 ; 
thus, every 100 parts of cuprous oxide equal 83'4 of chloroform. 
3. The fluid to be tested (which, if acid, should be neutralised) is 
distilled in a slow current of hydrogen, and the vapour conducted through 
a short bit of red-hot combustion tube containing platinum gauze. 
Under these circumstances, the chloroform is decomposed and hydro¬ 
chloric acid formed ; hence, the issuing vapour has an acid reaction to 
test paper, and if led into a solution of silver nitrate, gives the usual 
precipitate of argentic chloride. Every 100 parts of silver chloride 
equal 27*75 of chloroform. 
4. The fluid is mixed with a little thymol and potash ; if chloro¬ 
form be present, a reddish-violet colour is developed, becoming more 
distinct on the application of heat. 2 
1 Schauenstein {op. cit.). 
2 S. Vidali in Deutsch-amerikan. Apoth.-Zeitung, vol. iii. Aug. 15, 1882. 
