servations, he makes the possible error very small. But, in addition to this, he also usually makes observations on at least four clock stars, which gives him twenty observations to average up and determine by. As he inspects the record of these observations which has gone upon the chronograph sheet along with the clock beats he is able to determine, after proper calculations, how his clock stands.
Such observations are made every three or four evenings, and thus the clocks are not given time to get far out of the way. It is not usual for a good clock to show a variation of more than half a second. If the astronomer finds that his clock which is sending the time is running a fraction of a second slow, he goes to it and lays on the top of the pendulum bob a minute clipping of metal, which is equivalent to shortening the pendulum an infinitesimal amount. When he takes his next observation he discovers how his clock has been affected, and again treats it accordingly. Thus the time that is sent out automatically by the clock is kept always correct within a small fraction of a second. Those who receive the time sometimes arrange electro-magnets near the pendulums of their clocks, which act with the beats of the observatory clock, and their attraction is enough to hold or accelerate the pendulums as needed to make them synchronize with the observatory clock.
It will be seen that the means of obtaining exact time involve a very considerable outlay, and that the services of highly trained men are needed. The public is thus greatly indebted to the railroads, telephone companies, and other corporations which usually bear the expense of securing standard time. It is probable, however, that from motives of scientific pride no observatory would undertake to charge for this anything like what would be exacted for such rare service in any department of the commercial world.
It is worth while to note that even with such perfect clocks and favorable conditions it is still impossible to secure perfect timekeeping. Add to this the fact that it is not usual for those who send out the time, after it has been received from the observatory, to pay much heed to variations, even of several seconds, in their master clocks, and we see why it is a disheartening task to keep the best watch as near the second as the owner would fain have it. In the first place, the watch could hardly be made to keep such time if kept still in an unchanging temperature; secondly, it is still less capable of it when subjected to the jolting and changes of temperature it encounters when carried; and, thirdly, the means of obtaining time with sufficient exactitude are rarely available to the general public.