120 
sources of error, especially probable want of uniformity of 
temperature in different parts of the instruments, and also 
too great distance (about 12ft.) between the standard bar- 
ometer and the cathetometer, which was unavoidable in the 
present arrangement of the Laboratory. So that, as far as 
these observations go, the standard barometer of the Labora- 
tory and that of the Kew Observatory may be considered to 
agree, within a small experimental error. 
“On a New Calorimeter,” by J. B. Hannay, F.RS.E., 
Assistant Lecturer on Chemistry, Owens College. 
Having been desirous some time since of making a num- 
ber of calorimetric determinations, I set about constructing 
an instrument for the work, but I found that working in 
the rapid, irregular manner in which I was compelled to do 
by using only odd leisure moments, the difficulty of obtain- 
ing a perfectly constant temperature was very great ; and 
as I have in a great measure overcome that difficulty, I 
think that an account of the instruments which I have con- 
structed may be of some use to workers who may be placed 
in such a position as mine, and may desire to make 
measurements quickly, and without waiting until the tem- 
perature is quite constant. In the first place it is evident 
that as the external or accidental variation of temperature 
of the instrument is generally not more than 02 to 03 of a 
degree; if the internal variation be made greater in one 
case than another, the error arising from accidental varia- 
tion will be smaller. Now if we use a liquid body to 
expand as an indicator we must, if we want a large rise of 
temperature, use a small quantity of the liquid or a large 
quantity of the body experimented with ; either of these 
conditions, however, involves difficulties of a kind nearly as 
bad as the temperature variations ; but if we use a gas as 
our index, that is to say, if we use an air vessel into the 
middle of which we plunge the substance whose specific 
