164 PROCEEDINGS OF THE AMERICAN ACADEMY. 



resistance of the circuit with any great accuracy from these numbers, 

 any one of which might possibly be in error by 0.5% of its own value 

 and thus introduce a much greater fractional error into the result; 

 but it is interesting to find that if we compare the first number with 

 the others in succession, we get for the apparent resistance of the 

 circuit about 13060, 13070, 12915, 12990, and these numbers point 

 to a value 50% above the real value. The resistance of the "magneto- 

 inductor " alone was 40.3. 



(D) When another similar ballistic galvanometer of the same gen- 

 eral type was substituted for the one mentioned in (C) so that the 

 resistance of the whole circuit was 4270, Mr. John Coulson obtained 

 throws of 116.0, 100.1, 88.6, 72.5, or 65.8, according as the rheostat 

 resistance was 0, 1000, 2000, 4000, or 5000. If we compare the first 

 of these numbers with each of the others in succession, we get as 

 values of the apparent resistances of the circuit, 6295, 6470, 6665, 

 6550, and these indicate a value about 50% above the true one. 



(E) A d'Arsonval galvanometer such that the resistances of the coil, 

 suspending gimp, and connections amounted in all to only 4.8, was 

 connected in series with a rheostat and the " magneto-inductor " 

 already mentioned. The coefficient of air damping in this galvanom- 

 eter was considerable, and the field due to the permanent magnet ex- 

 tremely weak. The total resistance of the circuit was 45.1, plus the 

 rheostat resistance. By fastening the stops which limit the journeys 

 of the sliding coil on the inductor rod near together at the centre of 

 the rod or further apart near an end, the electromotive force induced 

 in the sliding coil as it was moved quickly fi-om one stop to the other 

 could be varied within wide limits. Three different positions of the 

 stop were used, and in each case a large number of observations were 

 taken and the mean used, as in all the results here given. For the 

 first position the throw of the galvanometer corresponded to a scale 

 reading of 77.8 mm. or 36.9 mm., according as the rheostat resistance 

 was or 50 ; these numbers point to a circuit resistance of 45.2, which 

 is almost exactly the true value. The readings for the second position 

 were 129.3 and 61.4, and the calculated resistance was again 45.2. 

 The readings for the third position were 170.4 and 81.7, and the 

 calculated resistance was 46.0 units. The results obtained in this 

 extreme case illustrate what one may expect if the field is extremely 

 weak and the number of turns in the coil small. 



(F) Another d'Arsonval galvanometer of very low coil resistance but 

 furnished with much finer gimp so that its total resistance was about 

 21, and with a strong permanent magnet, was treated in much the 

 same way as the galvanometer mentioned in (E). Two positions of its 



