The stem of the thermometer protrudes from the chamber 
as in the figure. A screw S somewhat larger in diameter 
than the bulb of the thermometer is made use of to attach 
the thermometer to its enclosure, and a smaller screw S' 
pressing home upon india rubber v/ashers enables the ther-= 
mometer to be properly adjusted and kept tight when in 
adjustment. 
In the present instrument the internal diameter of the 
chamber is 2 inches, while the bulb of the thermometer is 
about inches in diameter. 
The scale of the thermometer is very open, more than an 
inch going to one degree, I have generally allowed the 
image of the sun given by the lens to heat the thermometer 
bulb for one minute, during wdiich time an increase of tem= 
perature, not exceeding in any case two degrees, has been 
produced. 
As far as principle is concerned there appears to be no 
objection to the present instrument, nevertheless it is open 
to a very serious practical objection. The scale being so 
very open, the stem comprehends only a few degrees; 
frequently, therefore, the temperature is such that the 
extremity of the mercurial column is either below or above 
the stem. Now the thermometer has a small upper cham- 
ber, and by means of a method of manipulation well known 
to those who work with thermometers, it is possible to add 
to or take away from the main body of mercury in the bulb 
so as to keep the end of the mercurial column always in the 
stem. But experience has convinced me that for a ther- 
mometer with such a large bulb, frequent manipulation of 
this kind is not unattended with danger to the bulb. On 
this account the instrument in its present form is, I conceive, 
unsuited for steady work in an observatory from year to 
year. 
It is however possible without any appreciable sacrifice 
of the scientific principle of the instrument to alter it in 
