TIME-RECKONING. 13 
The sidereal day fulfils these conditions, and therefore suggests 
itself as being suited for the standard required. 
The sidereal day is not, however, sufficiently marked for the ordi- 
nary purposes of life. The generality of mankind could not easily 
note the culmination of a star. On the other hand, the diurnal 
return of the sun in the heavens is a more striking and easier 
observed phenomenon. Accordingly, there is everything to suggest 
the adoption for the unit measure, not the solar day on account of 
its variable length, but the mean period occupied by a revolution of 
the earth on its axis, in relation to the sun. 
That pericd would be precisely equal in length to the artificial 
day, known as the mean solar day. The unit measure proposed 
should not, however, be considered in the light of an ordinary day, 
but rather as a known period of abstract time—‘‘day” being the 
name given to denote certain local phenomena successively and 
continuously occurring at the earth’s surface. 
It is proposed to divide the unit measure into twenty-four equal 
parts, and these again into minutes and seconds, by a standard time- 
keeper or chronometer, hypothetically stationed at the centre of the 
globe. 
It is proposed that, in relation to the whole globe, the dial plate 
of the central chronometer shall be a fixture, as in Fig. 1; that each 
8 
