242 
and thus the maximum sensitivity : 
i (m—1)’ 
4n* (K—K,)(r—r,) m2? 
From this we derive that m should be made as large as possible. 
K—K, is the moment of inertia of the mirror, 7—r, is the given 
external resistance (rather + the resistance of the flexible leads 
which is a known quantity in any special case). The greatest though 
in practice unattainable sensitivity is thus: 
as 
= 
Puax’ ST FOI gat, Wo cm ie ner Uns (8) 
whilst the ratio P/Pnax might be called the efficiency of the galvano- 
meter. Hence one generally finds for this efficiency : 
ae rey (ee 1 9 
= & (F- ). Ne RU 
As m can be eg. 10 the conditions (7) mean that one should 
not only — as is known from former researches — reduce the 
resistance of the galvanometer compared with the external resistance, 
but that also the moment of inertia of the coil should be small in 
comparison with that of the mirror. 
Technical construction. We are going to make use of the above 
mentioned formulae for the further calculation of galvanometers 
with two different periods of 3 resp. 8 sec. For the circular mirrors 
which may be used we have: 
Diameter 12 10 8 millimeters 
Moment of inertia 0,0055 0,0026 0,0011 
for a thickness of 0,20 mm. Mirrors thinner than this are mostly 
insufficiently plane, besides they warp too easily in mounting. 
The attainable value of H depends not only on the size of the 
permanent magnet which is used but also on the dimensions of the 
airgap. For various existing galvanometers I found for A values 
near 700; once I found 1100. The small coils with only few turns 
of wire, which are needed according to our calculations, allow to 
increase H considerably, provided one places an iron core inside 
the coil. | use for example a core of 6.8 mm. diameter and 15 mm. 
height, and an airgap of 1.2 mm. round it. The coil then consists 
of rectangular turns of wire of 8 X 16 mm. With a simple steel- 
magnet the magnetic field proved to be ; 
== 2000 
I am going to accept these values for the following. From (6) we 
