452 KNIGHT DUNLAP 



with resistance, or 60 cycle alternating current with rectifier, 

 starting the motor is simple. With the thumb and forefinger 

 the armature shaft is given a spin in the proper direction. Wrap- 

 ping a string around the shaft, to get a pull, is not satisfactory. 

 On first trial with the thumb and finger method, a great many 

 spins will perhaps be necessary before the proper knack is acquired. 

 The commonest error is spinning too fast. Another common 

 error is taking hold of the shaft for a fresh spin, before noting 

 whether the preceding one was effective; thus spoiling a good 

 start. After a little practise facility is acquired, so that the 

 motor may be started with little difficulty; usually on the first, 

 second or third spin. 



On the 60 cycle unrectified current, since the speed of rotation 

 is relatively high, the finger and thumb method may be difficult. 

 In that case, wrap a strip of adhesive tape electrician's or 

 surgeon's around the shaft spirally to make a smooth single 

 thickness, and start by drawing the fingers across the wrapped 

 part of the shaft. It may be easier to spin the armature with 

 current off, closing the main switch as the fingers leave the shaft. 



THE CLUTCH CIRCUIT 



The clutch circuit must always be direct current, and should 

 have an amperage just sufficient to give sharp, not violent, action 

 to the clutch. The exact voltage will necessarily vary with the 

 accessory instruments used, and it is well to have a rheostat 

 in the circuit, as in figure 2, so that the current may be properly 

 controlled. 



The essential feature of the Johns Hopkins chronoscope, and 

 the Hipp without armature springs is, that within certain limits, 

 variations in the current strength do not alter the measurements, 

 since the clutch is moved in by one electromagnet, and moved 

 out by another which is the exact mate of the first in core and 

 windings. The two magnets are connected in parallel so that 

 variations in the current affect both equally. In consequence 

 of this feature, it is necessary to avoid differences in resistance 

 and in inductance between the two branches of the clutch circuit. 



