of Large and Small Alternating Currents. 101 
thermogalvanometer, and the results obtained are given in 
Table II. 
The currents, P.D.’s, and power are those required to give 
a deflexion | 1 the scale-distance, which forms a very convenient 
basis on which to compare square law instruments. I have 
also included the Ayrton-Perry twisted strip instrument in 
the table for comparison. 
TABLE II. 
For deflexion of 25 em.at 100 em. seale-distance. 
es | | i 
Resistance _ Current PLD, Power | 
Instrument. | in in micro- in in | 
| ohins. amperes. | milli-volts. | micro-watts. | 
| omen | | | 
ie ade Heater ....:.... 18 800 144 | ts | 
_ ‘Thermogalvanometer, | 
_ platinum on glass | / 
LEE FG | 108 346 39°6 123 
Do. er e:.: ! 202°5 275 55°6 | 15°3 
Do. Mow Li). s<. 363 231: | 84 19-4 
Do. eet: | 1071 121 130 157 | 
Do. 2 a | 3367 88 296 | 26°0 
Do. ar i... | 138910 dl 43] 13°9 
_ Ayrton-Perry twisted | 
fe SERED <=). ne scene 20 22000 440 9680 | 
The small power taken to deflect this instrument combined 
with the fact that the self-induction and capacity of the in- 
strument are so extremely minute, and that it obeys a square 
law, should render it very serviceable for many measurements 
which I need not detail. One instrument can be easily used 
for various sensibilities by simply changing the heater; and 
should at any time a very excessive current be put through 
the instrument and destroy the heater, it can be readily 
replaced without touching the delicate suspended system. 
With the highest resistance heater so far tested, namely 
13910 ohms, a detectable deflexion of 0°1 mm. at one metre 
is given by a current of 0°6 micro-ampere. This resistance 
is not the highest which I think can be made, as I succeeded 
in making one of 25,000 ohms, which unfortunately got 
broken before the sensibility was tested. The radio-micro- 
meter part can also be easily made more sensitive. I have, 
therefore, reason to hope that it may be possible to improve 
this thermogalvanometer so as to detect small fractions of a 
microampere with it. 
The present high sensibility combined with the suitability 
for use with high-frequency currents, leads me to think that 
it might be applied to measure the current in the vertical 
