Prof. Thomson on the Measurement of Electric Resistance. 159 



meter, and primary and secondary conductors of sufficient resist- 

 ances to fulfil the condition of doing away with sensible error from 

 imperfect connexions in the manner explained above, but yet of 

 resistances either less than or not many times greater than the re- 

 sistance of the galvanometer coil, it is easy to test to Yrfao the resist- 

 ance of a copper wire or bar not more than 3 inches long. The 

 current we have found to be sufficient for this object would only 

 produce a heating effect of '14° in two seconds, which, with good 

 apparatus, is more than enough of time, as I shall show presently. 

 The influence of this heating effect may be regarded as nearly 

 insensible, since even as much as *2° only alters the resistance of 

 copper by about -j-o^o" • 



In all measurements of electric resistance, whatever degree of gal- 

 vanic power is used, a spring "make and break" key* ought to be 

 placed in one of the battery electrodes, so that the current may 

 never flow except as long as the operator wills to keep it flowing, 

 and presses the key. I introduce a second similar spring key in one 

 of the galvanometer electrodes (that is between either Q or P and 

 the galvanometer coil), so arranged that the pressure of the ope- 

 rator's finger on a little block of vulcanite attached to either spring 

 shall first make the contact of the first spring (completing the bat- 

 tery circuit), and when pushed a little further, shall make the con- 

 tact of the second spring and complete the galvanometer circuit. 

 The test for the balance of resistances will then be that not the slightest 

 motion of the needle is observable as a consequence of this action on 

 the part of the operator. The sensibility of the arrangement is 

 doubled by a convenient reverser in the galvanometer circuit, by 

 which the current, if any, may be reversed easily by the operator 

 while keeping the two connexions made by full pressure on the 

 double spring key just described. Another convenient reverser 

 should be introduced into the battery circuit, to eliminate effects of 

 thermo-electric action if sensible. 



It may often happen, unless the galvanometer is at an inconve- 

 niently great distance from the conductors tested, that its needle will 

 be directly affected to a sensible extent by the main testing- current ; 

 but with the arrangement I have proposed the observer tests whether 

 or not this is the case by pressing the double spring-key to only its 

 middle position (battery contact alone made), and watching whether 

 or not the needle moves perceptibly. If it does not move perceptibly, 

 he has nothing more to do than immediately to press the double key 

 home, to test the balance of resistances. If the needle does move 

 when the key is pressed to its middle position, he may, when in 

 other respects allowable, keep the current flowing by holding the 

 key in its middle position till the needle comes to rest, or at least 

 till it shows the point towards which its oscillations converge, and 

 then press home to test the balance of resistances. When the 

 very highest accuracy is aimed at, or when, for any reason (as, for 



* Morse's original telegraph key, which instrument-makers have " improved" 

 into the in every respect worse form in which it is now commonly made — a 

 massive contact-lever urged by a spring. 



