56 MEMOIRS OF THE NATIONAL ACADEMY OF SGIENCES. 



readily determiued — should only equal Tb%o> or ^ Jo "f ^ vibration in one second, a variatiou of that 

 amount would be produced by a cliauge of temperature of only one-fourth of a degree F. in the 

 Theatre Feydeau fork, and if measured in beats would amount to the difference in the pitch of two 

 forks, which, when sounded together, would give oue beat in 200 seconds. 



On the uses op the tuning-fork as a chronoscope. — Various forms of chronoscopic 

 apparatus contain a vibrating foi'lc as a register of time. The majority of these are costly, by 

 reason of the attempts of the iuventons to obtain regular rotations of cylinders or disks by means 

 of clock-work, when really all such appliances are useless. The fork itself, if only allowed to register 

 its own trace on the revolving cylinder or disk, will give all that is desired without such adjuncts, 

 for the accuracy of its registratiou has no connection with the rotation of the cylinder on which it 

 leaves its record, and it matters not whether the latter be revolved quicker or slower, regularly or 

 irregularly, so long as the motion is appreciably uniform during the trace of one flexure of the fork ; 

 this duration in the case of an VTi fork would be oidy the oi,y of a second, and in that minute 

 interval it would not be possible to get a measurable variation in velocity unless we did our best 

 to attain it. Any ordinary care in the rotation of the cylinder by hand will give waves which at 

 and near the spark-mark will be found to be similar and equal, and therefore no error can be made 

 in the measure of the fraction of a wave. 



The numbers of vibrations of a fork per second can be determined to ^^-^ of a vibration, or, to 

 be surely within bounds, say to the '^Jii^ of a vibration, by the method we have described in this 

 paper. This will give the time record with an A fork of WO vibrations per second to ttoto of a 

 second. 



It is not necessary to make any correction for the effect of the scrape or weight of tracing 

 style or for the effect of the kind of support of the fork, for the number of the fork's vibrations per 

 second is determined while the fork is on the same support it has when used as a chronoscope and 

 while the fork is making its record; in other words, the number of the fork's vibrations per second 

 are determined in the cruet conditions in which it is used as a chronoscope. 



The arrangement of such a chronoscope is of the sinqjlest character. Fig. 4 shows it. As 

 an example, we will suppose that we are to determine the initial velocity of a rifle-ball. 7> is a 

 voltaic cell, whose current goes thi-ough the primary coil of the inductorium /, then to the target 

 T formed of a metal plate (or a screen of wire, if we are determining the velocity of a caunou ball). 

 This plate is very slightly inclined forwards, so that its upper edge presses very slightly against 

 an adjusting screw at A'. The abutting surfaces of this screw and the plate are amalgamated to 

 insure good elastic contact. The bottom of the plate rests in a siuall trough of mercury. The 

 current passes to this trough and out of the plate at the adjusting screw S, thence to the make- 

 circuit lever J\r(\ and back to the battery 7>. Oue pole of the secondary circuit of the inductorium 

 is connected with the fork F, the other pole with the rotating cylinder C. The make-circuit lever 

 is formeil in this manner: It moves around a center at 0. On its lower side are two platinum 

 lugs. By the motion of the lever around O, eitlier one or the other of these lugs are brought in 

 contact with two platinum contact-pieces, e and c, which are insulated from the plate and standard 

 on which the lever is supported. 



The chronoscopic apparatus haviug been arranged as in the diagram, the fork is raised on the 

 hinge h (see Fig. 1) and vibrated with a bow. The cylinder is revolved and the fork brought 

 down on its smoked paper surface. At the word "Are," the rifle is discharged. The fine wire or 

 thread ir is cut by the ball, and the weight j) which it supported and which brought the left 

 hand platinum lug onto the left hand insulated contact-piece, falls; then the spring ,v (or, better, 

 a rubber band), which opposed the action of the weight, swings the right hand lug on to the 

 right hand contact-piece. When the ball cut the wire, the primary circuit of the inductorium 

 was broken, and a spark, at that instant, passed from the style of the fork and made a spark- 

 hole in its sinuous trace. But the spring .v at once made contact again, and the circuit was 

 made through the right-hand lug c. The ball, therefore, reaches tlie target-plate T with the 

 circuit closed, and when it strikes 2* the plate is thrown from the contact-screw 8, and a second 

 break takes i)lace in tlie primary circuit and another spark passes from the style of the fork. By 

 counting the number of waves and measuring with a microscope-micrometer the fraction of the 



