18 
I have attempted to control the above empirical formula 
(Cr 3 0 6 Cl 2 ) by heating a weighed portion of the substance in 
hydrogen. The action of hydrogen upon the new chloride 
when heated is extremely energetic. At a comparatively 
low temperature it takes fire, combustion proceeds rapidly 
throughout the mass, and ultimately the substance is con- 
verted into chrome sesquioxide, hydrochloric acid, and 
water. Care must be taken to regulate the current of the 
hydrogen, since, if too rapid, particles of the finely divided 
sesquioxide are apt to be mechanically carried away, From 
an experiment, in which the gas was carefully purified from 
oxygen by passing it through strongly alkaline pyrogallate 
solution and over heated metallic copper, and then dried by 
transmitting it through tubes containing pumice moistened 
with strong sulphuric acid, the following numbers were 
obtained - 
0*8715 grm. substance gave 0*6150 grm. chrome sesquioxide. 
Found 70*58 °/ o Cr 2 0 3 . 
Cr 3 0 6 Cl 2 gives by calculation 70*72 °/ Q 
I had an additional object in thus studying the action of 
hydrogen upon the new chloride. I considered that this 
action might possibly throw some light on the constitution 
of this compound. The new oxichloride may in conformity 
with the analytical results be regarded as a compound of 
chromous chloride with two equivalents of chromium 
trioxide. Now, chromous chloride, according to Moberg, 
may be heated in hydrogen to the softening point of glass 
without suffering decomposition, and if it were found that 
water was the only volatile product of the reaction, we 
should possess a certain amount of evidence for supposing 
that the formula CrCl 2 , 2Cr0 3 represents the constitution 
of this substance. Experiment showed, however, that the 
chlorine was not so firmly united in this compound as in 
the chromous chloride : on gently heating the substance in 
hydrogen, hydrochloric acid was immediately evolved. 
