51 



a second respectively. These rates would not discredit 

 an astronomical clock. We have now considered the 

 methods of determining exact time, some of the precau- 

 tions necessary to keep it, and our last division of the 

 subject will be how to distribute it without sensible error. 



We have been talking in describing star transits, of 

 Sidereal or Star Time, and since the stars rise four min- 

 utes earlier every day, the sidereal day is four minutes 

 shorter than our common day. Now it is common or 

 mean time which we want to distribute, so first we must 

 convert the sidereal time into mean time. We have here 

 a mean-time clock loaned through the courtesy of Messrs. 

 E. Howard & Co. Within this clock is an arrangement 

 for breaking the electric circuit each alternate second 

 except the fifty-eighth. We have also here a chronometer 

 provided with a similar break circuit arrangement, and 

 we shall cause both of these time-pieces to register their 

 beats upon this telegraphic sounder. Now the sidereal 

 clock beating faster than the mean-time clock very soon 

 catches up with it, and for a few seconds they beat so 

 closely together that we can compare them to within one 

 one-hundredth of a second. And now we are able by a 

 short calculation, and knowing the error of the sidereal 

 clock, to exactly determine the error of the mean-time 

 clock, which latter error we can reduce to nothing by 

 altering the clock. Thus we have a mean-time clock set 

 perfectly to mean time, and by means of an electric cir- 

 cuit ready to automatically distribute its beats over as 

 long a circuit as we choose. 



We have about the hall a miniature telegraph line with 

 telegraphic instruments at two or three points, which if 

 }'ou please we will imagine to be Boston, Springfield, and 

 New York. We have only to switch the clock into this 

 circuit, with some precaution to avoid the strong battery 



