1843] WRITINGS OF JOSEPH HENRY. 213 



the hair spring of a watch, coiled around the centre pin 

 which supports the needle, and having an elastic force a 

 little less than the deflecting power of the electrical current. 

 The relative position of the dots thus formed gives the time 

 of the passage of the ball through the space between the 

 screens, and indicates the velocity at this part of the course. 



The degree of deflection of the needle can be increased or 

 diminished by turning a screw, which alters the position of 

 the " steady pin," and the tension of the spiral spring can 

 also be changed by an arrangement like that of the regu- 

 lator of a watch. 



In order that the position of the dots on the surface of the 

 cylinder may exactly indicate the required interval of time, 

 it is necessary that the time occupied by each needle in 

 starting from rest and moving across the small arc to strike 

 against the cylinder, should be precisely equal. If this be 

 not the case, then the difference of these times will be the 

 error of the instruments. This must however be exceed- 

 ingly small, since the whole range of the end of the needle 

 need not be more than the one-twentieth of an inch — and the 

 precise amount of error can readily be determined by ex- 

 periment. 



To adjust the apparatus for use, the galvanometers must 

 be so placed that the two dots may be impressed on the 

 cylinder, diametrically opposite each other when the instru- 

 ment is at rest. The cylinder being then put in motion, the 

 two circuits of long wire are placed together, so that they 

 can be broken at the same instant by lifting a wire common 

 to both from a cup of mercury. If, after breaking the cir- 

 cuits, the dots are still found in the same relative position, 

 no further adjustment or correction will be required: but if 

 this is not the case, then the springs may be altered until 

 the dots are found in their proper positions; or the differ- 

 ence may be noted, and this constantly applied in each 

 actual experiment as an index error. 



To prevent the dot from the first galvanometer being con- 

 founded with that from the second, the two instruments are 

 placed one below the other in different horizontal planes. 



