190 



Mr. (X V. Boys. 



[Mar. 24, 



Deprez (on the principle of Sir W. Thomson's siphon recorder) 

 avouIcI have many advantages. It occurred to me that if an active 

 bar were made with two pieces of metal, antimony and bismuth, 

 as thin as possible, soldered edge to edge, and if the outer edges were 

 joined by an arch of copper, and if this were hung in a strong magnetic 

 field, all the advantages of an ideal thermopile and of the Deprez 

 galvanometer would be gained, while the impossibility of the former, 

 and the resistance and tension of the stretched wires of the latter, and 

 the resistance of the wires connecting the thermopile with the galvano- 

 meter would be abolished. I anticipated that a greater delicacy would 

 be obtained than has been considered possible, that the instrument 

 would be convenient in that it would be of necessity dead beat, that 

 its indications would be proportional to the received radiations, that 

 it would have the constant and definite zero which mechanical control 

 such as torsion gives, that it would not be affected by the magnetism 

 of objects near it, and that as the circuit could be suspended in a hole 

 in a mass of metal with one window for the radiation to enter, the 

 instrument should be very insensitive to temperature changes other 

 than those in the line of action. It has the further advantage that 

 for spectroscopic work the radiation may be limited by a narrow slit 

 without much reducing its sensibility.* It is, however, a fixed instru- 

 ment that cannot be handled and pointed like the thermopile or the 

 bolometer. 



I have made some preliminary experiments to test these conclusions. 

 One instrument, made by my friend Mr. Cunynghame very roughly 

 with a torsion wire support only a few inches long, is fairly sensitive. 



One which I made with the active bar about 6 x 8 X^- mm., and 

 with a circuit 1 cm. square, of one turn of which the three sides are 

 made of copper wire about ^ mm. in diameter, and in which the 

 motion of the circuit is resisted by a torsion fibre of the finest spun 

 glass about 10 cm. long, gives when placed on the poles of a permanent 

 magnet results so good that I have ventured to submit the instrument 

 for inspection to the Royal Society. This particular instrument is 

 capable of showing the heat which would be cast by a candle flame 

 on a halfpenny at a distance of 200 yards. t As I have not yet em- 

 ployed what I know to be the best proportions, it is evident that very 

 great sensibility is possible. 



* I shoxdd liaTe mentioned that much larger angles could be measured if the. 

 fibre were suspended by a torsion head and the light brought to a definite position 

 for every observation. — [March 2, 1887.] 



f This same instrument with a 38 cm. fibre in a field of about 100 units only, 

 gave a deflection of 4'5 cm. when the heat cast by a candle flame 12 feet off fell on 

 the active bar which was exposed over a surface of 4 x 6 mm., therefore a candle 

 254^ feet distant would produce a visible effect, or the heat which would fall on a 

 halfpenny 1168 feet from a candle flame would be sufficient to be observable if it 

 were directed on to the active bar. — [March 2, 1887.] 



