828 Proceedings of Royal Society of Edinburgh. [sess. 
galvanometer mirror at once swings from the zero position. The 
gratings may conveniently be cut out of very thin tinfoil, and 
mounted on three-quarter inch or seven-eighth inch microscope 
cover-glasses, which are cemented on wooden blocks, the 
terminals being screwed into the blocks through lugs of the 
tinfoil proceeding from the ends of the gratings. The galvan- 
ometer employed was a Wiedemann with low-resistance coils 
(1 ohm). It was found that the construction of a sensitive 
instrument by experimental observation was by no means easy. I 
had not read Langley’s Annals of the Astrophysical Observatory 
for 1900, in which, with the help of Mr Abbot, he gives the 
mathematical theory of his instrument. I had used short coils 
of high-resistance wire for my connections, and, in turn, gratings 
of tinfoil, leadfoil, aluminium, silver, iron, copper were used, with 
disappointing results. The movement of the spot of light for 
radiant heat from the hand was only about three feet on our 
class-room scale (about twelve feet from the mirror of the 
galvanometer), a result rather inferior to that obtained by our 
delicate thermopiles. I then investigated the question mathe- 
matically on the supposition that A and B were kept at a fixed 
difference of potential by a shunt current from the mains of 
reduced voltage, and I found that the most advantage was gained 
by making the arms A C and C B of as low resistance as possible, 
and of equal resistance. 
Thick copper wires were substituted for the connections, and 
the gratings were made of a resistance of from one to three ohms, 
with six or seven bars to the grating and very fine inter-channels. 
The result was a great improvement in sensitiveness. Radiant 
heat from the hand now gave a twenty-foot swing, and touching 
a grating caused the heavy ring magnet to spin round and round. 
The spot of light was caused to swing with every draught in the 
room, and the instrument was already so sensitive that its value 
was much lessened unless special precautions were taken to enclose 
it from accidental influences. In the complete instrument all 
connections as Avell as the gratings must be covered in and 
protected from such influences. I would draw the attention of 
physiologists to the fact that no especially delicate galvanometer 
is required for our purposes. Langley uses special galvanometers, 
