354 



and after these observations the sidereal and mean solar chronometers 

 were compared. This gave the error of the first on sidereal, and of 

 the second on mean solar time for the meridian of the Chicago obser- 

 ving station. Immediately on returning to Chicago the same thing 

 was again done. This gave a new determination of the errors for the 

 second Chicago period, and also the rates of both chronometers during 

 the elapsed time. 



Between these two periods the journey forth and back was made, 

 and also the observations at the place whose position was sought, and 

 the telegraphic signals exchanged, in the manner hereinafter reported. 



The mean solar chronometer was always left at Chicago, and the 

 signals sent from that place, as well as those sent to Chicago, were 

 noted by it there. 



The observations at the distant station, whose position was sought, 

 and the telegraphic signals received at and sent from that station, were 

 all noted there by the sidereal chronometer. These signals were al- 

 ways made by myself by pressing the telegraph key with the fingers, 

 so as to produce the click of that key as nearly as possible in coinci- 

 dence with a given beat of the chronometer. 



The signals at Chicago were always made by an experienced tele- 

 graph operator, who was, in the beginning, carried through a course 

 of practice in making dots as nearly as possible in coincidence with 

 the beats of the mean solar chronometer at every ten seconds of inter- 

 val, for seven to ten and sometimes thirteen minutes as a series. It 

 is remarkable how soon a person, with a good ear for cadence, or 

 time, will acquire an accuracy in making these time-signals approach- 

 ing almost to exactness. The results wliich will presently be pre- 

 sented will serve as evidence on this point. 



After the observations for the time, at the place visited, were com- 

 pleted, the sidereal chronometer was carried to the telegraph oflice at 

 that place. I then began by calling for a certain number of signals 

 from Chicago at intervals of ten seconds apart, sufficient to insure two 

 or three periods of coincidence in the beats of the two chronometers. 



This period of coincidence of beats having been thus ascertained, 

 signals were sent from the distant station back to Chicago, at inter- 

 vals that would ensure a coincidence with the beats of both chronome- 

 ters. These intervals were easily ascertained by making allowance 

 for the difference of the ratcsof the two chronometers atiected by their 

 proper algebraic signs of -f when gaining, and — when losing, as 

 well as for the gain of sidereal on mean solar time. 



Sometimes additional signals were called for from Chicago at stated 



