60 
The former process consists in treating acetic anhydride 
with peroxide of barium, when the two gases di-methyl and 
carbonic acid are said to be given off ; in the latter process 
the action of zinc on methyl iodide in sealed tubes heated 
to about 150° C. yields di-methyl. 
These two methods were tried for the preparation of 
di-methyl, but were abandoned for the electrolysis method 
for reasons which will be noticed further on. 
I. Preparation of Di- Methyl by Schiitzenberger’s 
Process. 
When peroxide of barium acts on acetic anhydride, a gas 
stated to be di-methyl is given off. 
If the peroxide is heated with the anhydride, as Schutzen- 
berger directs, violent explosions occur. This can be avoided 
by mixing the peroxide with dry sand provided the flask 
containing the mixture is not cooled; if it is cooled, the 
reaction after a time becomes so violent, probably owing to 
the formation of peroxide of acetyl, that explosions occur. 
20 grms. of peroxide of barium mixed with 40 grms. of 
dry sand were poured into a flask containing 20 grms. of 
acetic anhydride and mixed by shaking ; the gas evolved 
was collected in a Pepy’s gas holder,. and afterwards dis- 
placed by pressure ; purified by passing through a solution 
of caustic potash and afterwards through concentrated sul- 
phuric acid. The treatment with chlorine, exposure to the 
sun’s light, and final displacement, was effected in the 
method described in the previous paper. 
The total volume of liquid obtained from acetic anhydride 
prepared from one pound of phosphorus did not exceed 
25 cc., and began to boil at 40° C., the temperature rising 
up to 80° C. This quantity proved too small to admit of 
being fractionated, in order to separate any more highly 
chlorinated products. 
In order to ascertain the cause of so small a yield of 
