BUNSEN ON THE CACODYL SERIES. 283 
The analysis of the other constituents of these substances, 
whether volatile or not, can be determined without difficulty by 
any of the methods already known. As many of.these sub- 
stances become solid by absorption of oxygen, it is dangerous 
to cut off the points of the glass balls which contain the fluid 
destined for analysis outside of the combustion tube, for the soli- 
difying of a portion in the tube might cause an explosion during 
the course of the experiment. In weigbing out these fluids I 
have therefore used small glass tubes, one end of which has 
been drawn out into a fine point three or four inches long. 
They also possess the advantage, that the weight and contents 
can be marked on the outside by means of a diamond, and a great 
many filled at one time, which is very necessary, in consequence 
of the ease with which these substances are decomposed by con- 
tact with the air. Ten or fifteen of these small tubes are then 
filled with carbonic acid, and all connected together by a piece 
of platinum wire. The larger end is heated and immediately im- 
mersed in a tube, (previously hermetically sealed but one end 
. broken off at the moment,) to allow all the fine points to dip 
into the fluid, which then immediately rises up in these small 
tubes as far as the heated part. As soon as this takes place 
the fluid must be made to enter the body of the tube by a gen- 
tle jerk. The fluid then boils and drives out all the carbonic 
_ acid, and on cooling the tube is filled with fluid for analysis. 
_ This operation may be performed in thirty seconds, and it is 
impossible to fill the tubes one by one, as the fluid usually be- 
comes so muddy in the course of three or four minutes as to be 
quite unfit for analysis. These small tubes can be easily opened 
inside of the combustion tube, {which has been bent and then 
again drawn out in the horizontal direction,) by means of a wire 
bent in the form of a cork-screw. The estimation of the hydrogen 
and oxygen is more difficult, if other substances are present. It 
can however be accomplished with oxide of copper by using a 
longer combustion tube, the fore-part of which is filled with 
chromate of lead or copper chips. Some of these compounds 
cannot be analysed by means of oxide of copper, as they are 
suddenly decomposed before the termination of the combustion ; 
I have therefore employed chromate of lead. 
In concluding these remarks, I ought to notice one thing 
more, which is of great importance in preparing these substances 
in a state of purity, viz. never to use larger vessels in the ex- 
uU 2 
