286 Mr. O. M. Mitchel on the Velocity of the 



2. To reduce the pass of the receiving magnet to a mini- 

 mum, and to keep it unchanged. 



3. To adjust the recording pens to exact radiation of records, 

 the disc being quiescent. 



To ascertain the effect of intensity and pass, the following 

 experiments were performed. The connexions having been 

 made as above described, four circumferences of second-dots 

 were recorded by the two pens, under the following circum- 

 stances : — 



No. 1. The pass of the receiving magnet a minimum. 



No. 2. The pass of the receiving magnet a maximum. 



No. 3. The battery reduced to one-half its former power, 

 the pass a maximum. 



No. 4. The reduced battery, the pass a minimum. 



On circumference No. 1, — minimum pass, and strong bat- 

 tery, — the variable pen fell behind the standard pen 0^'091 on 

 a mean of many measures. The uniformity of these records 

 and the accuracy of the measures are best exhibited by the 

 measures themselves. I give as a specimen the first ten mea- 

 sures out of thirty : — 



0-091 0-091 0-092 



0-092 0-091 0-091 



0-090 0-090 Tv/r^„„ 73^77: 



0-090 0-092 Mean 0-091 



On circumference No. 2, — maximum working pass, strong 

 battery, — the same interval became 0^'2628. 



Circumference No. 3, — battery reduced one-half, maximum 

 pass, — the same interval measured 0^-310. 



Circumference No. 4, — weak battery, pass a minimum, — 

 same interval measured 0^-104. 



From these experiments, it becomes manifest that the adjust- 

 ment of the receiving magnet may give variations of record 

 far greater than the anticipated value of wave time on the 

 longest available circuit. It is further shown, that the effect 

 of different intensities is such as to entail an error so great as 

 to render all experiments useless, from which this effect is not 

 strictly eliminated. 



Two difficulties were now to be overcome. The batteries 

 must be reduced to equality, and the evidence of that equality 

 must be obtained. The following plan was adopted to accom- 

 plish these objects. The handles of the recording pens are 

 flexible, and vibrate at every stroke of the pen. Half the 

 length of this vibration is the armature time, as I will show 

 hereafter. Now the armature time was found to depend on 

 the intensity of the currents which operated on the receiving 



