448 TRANSACTIONS OF THE AMERICAN INSTITUTE. 



suring and marking the resistance of all the magnets I put upon 

 the market, which enables telegraph superintendents to arrange 

 and adapt their magnets to the several parts of lines so as to 

 secure the greatest economy in their use, which is a matter of 

 veay great importance. I make the rheostat daily useful also 

 in comparing magnets and ascertaining their relative working 

 qualities. 



The subject of a uniform standard of resistance has engaged 

 the attention of electricians considerably, but they have not as 

 yet arrived at anything reliable. Wire of a given number is not 

 only more or less variable in its dimensions, but is also variable 

 in the specific resistance of the metal of which it is composed. 

 The standard unit of resistance of this instrument approximates 

 that of one mile of No. 8 iron wire. It consists of coils of diifer- 

 ent resistances, from the one-quarter mile to 150 miles, which are 

 so connected with switches that any amount of resistance up to 

 1,200 miles can be introduced at pleasure; and the graduated 

 sliding bar subdivides the quarter of a mile into hundredths of a 

 mile. The true Tangent Galvanometer (Eheometer) measures 

 correctly the strength or actual force of a current in circulation, 

 which it may be demonstrated is directly proportional to the tan- 

 gent of the angle of deflection. Common galvanometers do not 

 fulfill the requisite condition for this, viz: that the adventitious 

 force which is sent through the galvanometer coil shall act with 

 the same uniformity upon the needle in all its deviations as the 

 terestrial magnetism does. Where the coil is narrow and the 

 ne43dle long, the inductive influence upon the needle is very great 

 while it is at or near the meridian; but, as it deflects, its extremi- 

 ties pass more and more away from the rays of induction, and 

 consequently its deflections are less and less, so that the tangents 

 of deflection are not at all proportional to the strength of current. 

 To obviate this difiiculty I made a coil of few la^'ers carefully 

 wound, whose width was equal to the length of the needle, but 

 upon trial a difficulty in the opposite direction was manifest. 

 When the needle is on the meridian over such a coil it is under 

 the influence of but few convolutions of the coil wire, but as it 

 deflects more and more, it comes under the influence of more and 

 more of the convolutions, so that the eflect, instead of diminish- 

 ing, as in the former case, is more and more increased. Being 

 'now convinced that the truth lay somewhere between the extremes 

 >of these experiments, I resolved to find it if possible, and upon a 



