18 
is slowly bubbled through the apparatus for several hours 
(both before and after attaching the sodium tube) until a 
sample of gas issuing from the last wash -bottle is entirely 
absorbed in water. The tube containing the sodium is also 
heated gently whilst the gas is passing through to drive off 
traces of moisture possibly adhering to the sides. The 
sodium tube is now sealed off at both ends, and the bulb 
broken by allowing the hooked ends to fall sharply against 
the ends of the tube. The following results were then 
noted in two experiments. 
1. The sodium retained its metallic appearance for a 
few weeks, and it slowly assumed a dark grey colour, finally, 
after several months, a deeper violet grey. 
2. The sodium lost its metallic appearance much more 
quickly, and after a few weeks became dull black, like 
charcoal. This blackness did not extend far below the 
surface, that portion of the metal attached to the glass 
retained its mirror-like appearance. 
The composition of this black compound has not been 
further investigated ; but it may possibly be a sub-chloride 
of sodium. 
Experiments with metallic aluminium point to the fact 
that it is unacted upon by the gas when dry, whereas, as is 
well known, it is readily tarnished by moist hydrochloric 
acid gas, and dissolves in the liquefied acid. 
These experiments, which have for the present been dis- 
continued owing to want of time, will be taken up again, 
and a larger number of facts, if possible, collected. 
‘‘Capillary Constants of Benzene and its homologues 
occurring in Coal-Tar,” by J. B. Cohen, Ph.T)., F.C.S., com- 
municated by Dr. A. Schuster, F.B S. 
The object of the present paper is to bring before you a 
subject which promises to have special interest to the 
