594 ELECTRICAL MEASUREMENTS 



and the averaging of many observations are necessary. After 

 the experimental work has begun these things may distract the 

 beginner's attention from insidious constant errors and errors of 

 method so that these are scarcely thought of though they may 

 be of vastly greater importance than the errors of reading. For 

 this reason it is necessary to study carefully any proposed test 

 or method of measurement in order that, as far as possible, all 

 constant errors and errors of method may be eliminated before 

 the experimental work is begun. 



Another result of this preliminary study should be a proper 

 division of the labor among the various component measurements. 

 Generally there are some measurements whose effects on the 

 final result are small, and therefore labor expended in making 

 numerous readings is wasted. This, however, does not imply that 

 when they are taken, the readings giving these less important 

 components may be made in a slovenly manner. 



In an investigation, the preliminary study is frequently a very 

 difficult part of the work, involving as it does a careful analysis 

 of the workings of the apparatus, a wide range of theoretical 

 knowledge and an accurate understanding of the behavior and 

 sources of error in many different kinds of instruments. 



In a complicated experimental investigation, to complete the 

 elimination of constant errors and errors of method, a second 

 determination of the quantity under measurement should, if 

 possible, be made by an independent method, using other apparatus. 



The beginner must keep in mind that it is scarcely possible to 

 measure any of the electrical magnitudes as he finds them in 

 engineering practice, without changing the conditions of the 

 circuit in which the measurement is made and thus altering the 

 very thing which is to be determined. An ammeter when intro- 

 duced into a circuit alters the current, and an electromagnetic 

 voltmeter alters the potential difference to which it is applied. 

 In the vast majority of cases, these effects are negligible, but the 

 possibilities of error due to the alterations of circuit conditions 

 must not be lost sight of. 



Calibration Before and After Tests. In making careful accep- 

 tance tests of electrical machinery, the indicating instruments 

 should be calibrated before the test and a check calibration made 

 at the conclusion of the work. This allows the various runs to 



