THE REFORMATION IN TIME-KEEPING. 147 



cal standards soon became decidedly intricate. Travelers were greatly 

 inconvenienced by the lack of knowledge of the standard upon which 

 the time of trains as advertised was based, and to such the situation 

 was full of difficulties. Some of these difficulties were stated in an 

 " open letter " published in " The Century " for September, 1883. The 

 subject in its practical aspect also attracted the attention of scientists 

 and scientific societies. It became a prominent topic of discussion at 

 meetings of the American Metrological Society, the Association for 

 the Advancement of Science, and the Society of Civil Engineers. Al- 

 though astronomers use sidereal time, based upon the position of the 

 stars, and not of the sun, in common with many other scientists they 

 were genei*ally warmly interested in the subject. 



The local time kept by clocks is an average of solar time, and is 

 properly designated " mean time," as distinguished from the variable 

 time shown by the sun-dial. No clock or watch can be made to keep 

 the time as shown by the sun-dial, and this new system of time-keep- 

 ing, therefore, became necessary when clocks and watches were in- 

 vented. The relation between mean and apparent time, and what 

 is meant by " the equation of time," may be seen at a glance by ref- 

 erence to the accompanying diagram. Mean time being represented 

 by the right line graduated for the several months of the year, the 

 variation of apparent time is shown by the curved line entwined 

 around it. In other words, a line drawn through the several posi- 

 tions of the sun at mean noon will describe the curves as indicated. 

 For reasons which need not here be stated, the diagram will be found 

 generally correct for one year only out of four ; but, upon the scale 

 by which the diagram is drawn, this error is infinitesimal. It is 

 hardly necessary to state that the principal cause of the variation be- 

 tween mean and apparent time is " the obliquity of the ecliptic to 

 the equinoctial." 



Apparent time is about fifteen minutes slower than mean time 

 about February 10th, and about sixteen minutes faster on October 

 27th. They agree about April 15th, June 15th, August 31st, and De- 

 cember 24th. If a well-regulated clock were set by apparent time on 

 October 2Tth, it would be about thirty-one minutes faster than appar- 

 ent time on the following February 10th. It will be seen that, under 

 such circumstances, clock-time would vary as much from true sun- 

 time as any clock set by the present system of standard time varies 

 from mean time at the most extreme point. 



The safe operation of a railway requires that the watches of all 

 its employes upon, or who have occasion to refer to, the same trains 

 should always indicate the same moment of time. Railway-time 

 upon lines running east and west can of course never coincide with 

 mean local time except at a single point, and the longer the line of the 

 road the greater will be the variation. Before the recent change to 

 standard time there were several cases where the railroad-time in use 



