1028 
PROFESSOR J. W. MALLET OX A REVISION 
value of the force of gravity at Paris and at the University of Virginia had to be 
taken into account. In view of the difference of latitude and elevation above the sea 
this constant is, in C.G.S. units, 
For Paris. 9 80-94 
For University of Virginia . . . 979‘95 
and, applying the difference in nominally normal pressure at the two places, Pegxault's 
value for the weight of a litre of hydrogen at 0° C. and 760 m.m., ’089578 grm., 
becomes ’089488 grm. 
In the experiments made in this way the only assignable cause of constant error, 
tending to affect in a particular direction the atomic weight deduced from them, seems 
to be the retention in solution of traces of hydrogen by the alkaline liquid in the bulb. 
The tendency of this is, of course, to make the atomic weight of aluminum appear 
greater than it should be, but I am satisfied that the possible extent of such error 
must be excessively minute, inappreciable within the limits of error of observation. 
Direct results of first set of third series of experiments. —The results obtained were 
as follows 
A.—Hydrogen by volume at 0 C. and 760 m.m. 
I.- 
’3697 grm. of A1 gave 
458’S 
c,c.=-04106 
grm. of H. 
II.— 
’3769 ,, 
467’9 
„ =-04187 
33 
III.- 
••3620 
449T 
„ =-04019 
33 
IV.— 
•7579 
941’5 
„ =‘08425 
33 
V.- 
•7314 
907’9 
„ =-08125 
33 
VI.— 
’7541 
936-4 
„ =-08380 
33 
Second set of experiments of third series.—Hydrogen collected and weighed as water. 
—Details of method used. —As the collection and accurate measurement of larger quan¬ 
tities of hydrogen would be difficult from the great weight of mercury to be dealt with, 
while it seemed desirable to repeat these experiments with a larger amount of aluminum, 
an arrangement was adopted for burning the hydrogen and weighing it as water. 
A bulb like that described above was used for the reaction between the aluminum 
and solution of sodium hydrate, another small tube (which may be designated as (x)) 
being connected, however, with the bulb, for the purpose of sweeping out air at the 
beginning of the combustion and the remainder of the hydrogen at the end by a 
current of other gas. The gas from the delivery tube (c) was carried through a series 
of four drying tubes, containing in the former two pumice stone soaked with pure 
sulphuric acid of full strength, and in the other two well dried asbestos and loose, 
woolly phosphorus pentoxide; thence it passed through a long combustion tube, filled 
for the first two-fifths with pure, finely granular cupric oxide, which had been recently 
ignited, and for two-fifths more with turnings of electrotype copper oxidised on the 
