932 
turer in general physics happens to have had 
considerable experience in the use of these 
instruments they are likely to cause annoy- 
ing delays in the generally hurried prepara- 
tion for an experimental lecture. 
For demonstrative purposes the simpler an 
instrument is the better, provided it may be 
made sufficiently sensitive, because in an ele- 
mentary lecture course there is danger of the 
student losing sight of the very phenomenon 
which an instrument is designed to render 
apparent if his attention is divided between 
the instrument and the phenomenon. I 
have recently devised a substitute for the 
bolometer or other detecting instrument which 
meets the requirement of simplicity in an 
eminent degree, as it consists only of an india- 
rubber band, a glass fiber, a small mirror and 
a small piece of plate glass. It may be put 
together and adjusted in less than a minute 
and may be mounted anywhere on the lecture 
table, as it does not require a very stable sup- 
port. 
It is well known that the elasticity of 
SCIENCE 
[N.S. Vou. XXXV. No. 911 
stretched india-rubber increases with increas- 
ing temperature so that on heating the — 
stretched piece it shortens. An india-rubber 
band about one millimeter wide is stretched 
to nearly double its length about the rec- 
tangular frame of plate glass FGHI and a 
glass staff DH (which should be as straight as 
possible and of a diameter such that it is just 
stiff enough to support the small mirror # of 
silvered microscope cover-glass of about two 
or three square millimeters area) is placed, as 
shown in the figure, under the rubber band at 
its middle point B. If the portion AB be 
heated the staff D# will roll on the frame 
towards the edge FG, and a beam of light re- 
flected from the mirror H may be used to indi- 
eate (and magnify) this contraction of the 
india-rubber band. A change of temperature 
of both portions of the band, AB and BC, will 
be without effect on the position of the mirror 
H. This ensures steadiness of “zero.” The 
inertia of the moving parts is small and the 
change in elasticity of the rubber follows the 
changes of temperature without any very 
noticeable lag, so that the instrument is both 
“dead beat” and sufficiently quick to respond 
to differences in temperature between the sec- 
tions AB and BC, and insensitive to changes 
in temperature which affect both of the sec- 
tions. 
For use as a substitute for the bolometer it 
is sufficient to clamp the rectangle FGHI in 
a vertical plane and to cause the spectrum to 
pass across the portion AB—the rise in tem- 
perature of the strip AB is approximately 
proportional to the energy present in the por- 
tion of the spectrum which covers AB and, the 
change in elasticity being approximately pro- 
portional to the change in temperature, the 
deviation in the beam of light from the mirror 
is approximately proportional to the energy 
present in the particular part of the spectrum 
which covers the strip. 
For demonstration in the spectral region 
between the red and wave-length j= 3 
where glass begins to be opaque the arrange- 
ment shown in the figure is convenient—J is 
a Nernst filament mounted in a vertical plane 
V —h, and h, large short-focus glass lenses and 
