MEASUEING THE VIBRATORY PERIODS OF TUNING-FORKS. 47 



portion of tlie discharge has passed tbere is a period of quiescence lasting about y-jVo second; then 

 follows a shower of minute sparks, which forms the short dotted line already referred to. This 

 si»iirk-shower lasts ^rio of "^ second, and is form(\d of 30 sjiarks; hence the average interval sepa- 

 rating these sparks is ^iiVo second. The intervals sejjarating these sparks arc, however, not uni- 

 form, Imt are smaller in the middle of the spark-shower than at the beginning or at the end of 

 this piit'iiomcnoii. This sjiark shower, indeed, is a miniature of the phenomenon obtained when a 

 Lejden Jar is placed in tlie secondary circuit of the coil, and which will be described in the follow- 

 ing experiments. The above determinations of intervals of time in the discharge are the mean 

 of measures on six disks. 



2. Discharge of large indxjctorium betaveen platinum points one millimeter apart, 



WITH A LEYDEN jar OF 242 square CENTIMETERS SURFACE IN THE SECONDARY CIRCUIT 

 OF THE COIL. 



After this discharge through the disk a very remarkable appearance is presented. The dis- 

 charge in its path around the rotating disk dissipates little circles of carbon. There are 91 of these 

 circles, each perforated by 4, 3, 2, or 1 holes. I have to frame a new nomenclature to describe this 

 complex phenomenon. I call the whole act of discharge of the coil, the (lischargc. Those separate 

 actions which form the little circles by the dissipation of the carbon we will call flashes, and the 

 perforations of these circles we call spa rl- -holes. The dischai-ge in the above experiment lasts -2-4- 

 second. The flashes at the bi'ginniug of the discharge are separated by intervals averaging 5^5 

 second up to the tenth flash ; after this the intervals of the flash rapidly close up, so that during 

 the fourth fiftii of the discharge they follow at each ^^g^ of a second. During the last fifth of the 

 discharge the intervals between the flashes gradually increase, and the last flash is separated from 

 its predecessor by y^oo of ^ second. 



The appearance of the carbon-covered disk, after one of the discharges just described has 

 passed through it, is giveu in Fig. 3. 



On diminishing the current in the primary coil of the inductorinm I found that the number of 

 flashes in the discharge diminished, so that at last I obtained a discharge which consisted of but 

 one flash perforated by one minute spark-hole. Also, if the current remain the same and a i)ortion 

 of the secondary circuit be divided, and gradually separated more and more, the number of flashes 

 in the discharge will be diminished and the whole energy of the discharge concentrated in time. 

 l>ut no rule can be given for any special coil to obtaiu from it such a discharge as is alone useful 

 in the work on the forks, and the current must be gradually varied by resistances in the primary 

 circuit of the inductorinm, and the area of the condenser in the secondary circuit, till the conditions 

 for any special coil are obtained which cause it to give a s[)ark which makes a minute circular and 

 well-defined mark directly in the trace of the style of the fork. In the inductorinm used there is 

 150 feet of wire in the primary circuit and eight miles in the secondary. The condenser in the 

 secondary circuit was formed of tin-foil separated by panes of glass, and had an area of 50 square 

 inches. 



STUDY OF THE EFFECT OF VARYING A3IPLITUDES OF VIBRATION OF THE FORK ON ITS VIBRATORY 

 PERIOD; AND ON THE EFFECTS OF VARYING PRESSURES OF THE STYLE ON THE PAPER-COV- 

 ERED CYLINDER. 



The experiments on this fork of Kceuig's were made not so much for the determination of its 

 vibratory period at a given temperature, as to discover any effect on the vibratory period caused by 

 ditterence of amplitude of vibration, and by varying jiressures of the tracing style on the smoked 

 paper. This series of measures is given as an average example of series of similar sheets on which 

 we have made measures. It will be observed that the vibration-numbers opposite the successive 

 seconds, given in the first column, are alternately small and large. This is due to the fact that the 

 center of the globule of mercury is not exactly on the vertical of the jjenduluTu, but by taking 



