( 403 ) 
tried to show it experimentally by the measurements of ‘which 
table V gives the results. 
TABLE V 
4: | 
Coils Mean deviation in Mean deviation in bes 
i 
minutes. minutes. 
3A 470 | 
1.036 
3B 504.5 Rou 
15A Wl Fe) Gia a alae 
| 1.539 
15B 263 | 
The method followed for getting these results is shown in Fig. 11. 
T is the translator without iron, already mentioned on page 369. 
G is the electro-dynamometer, which for this experiment however is 
provided with thicker wire. The thickness of this wire is 0,2 mm., 
the resistance 42 Ohm. « is the coil with the iron core in it, which 
we used before. Before the microphone are the resonator and the 
tuning fork as before. If the 4 keys are pressed down, the primary 
of 3 A is connected with the electro-dynamometer and the translator; 
and the secondary of 3 A is connected with coil a. If the keys 
are on the contact of rest, 3 B has changed places with 3 A. 
These comparative experiments were made with the coils 3 and the 
coils 15; it is true, that table V shows but a very trifling influence 
of the iron, but yet it is clear that the influence is greater for 15 
than for 3. These numbers cannot have an absolute value, the 
introduction of the translator and of the electro-dynamometer in the 
primary circuit greatly increasing the self-induction. On purpose to 
make this increase as small as possible, we provided the electro- 
dynamometer with thicker wire. If this experiment could be made 
with a measuring-instrument and a translator without self-induction 
we would of course have got 1,17 for coil 3 and at 2,66 for coil 
15 (see row 4, table page 402). The proportion of those numbers, 
2,66 
1,17 
bers of table V. This gives however for that proportion only 1,5. 
= 2,3, should correspond with the proportion of the final num- 
