152 
HINTS TO TRAVELLERS, 
The first require, when the time elapsed since the rate of the chrono¬ 
meter was last ascertained is great, a good and carefully-guarded 
timekeeper, and is known by the name of “ meridian distance,” or 
measuring the difference between the meridian of the place and that 
of the place where the chronometer was last rated, whose longitude is 
known. This method, when applicable, is by far the best, but in 
travelling requires that a continuous chain of observations should be 
taken from the time of leaving a place whose position is known; and as 
a watch, carried either by a pedestrian, or on horseback, rarely keeps 
an equable rate, the points where halts must be made for rating should 
not be more than five or six days apart. 
The second method depends, in its various forms, almost entirely upon 
the rapidity of the moon’s motion in the heavens, and, while it gives the 
longitude without reference to any previous observation, the result is 
always more or less rough, unless a great many observations are made on 
different nights, when the mean may approximate to the truth. 
In any of these observations, with the exception of moon culminating 
stars, the true time at the place is required, and the method of finding 
this will first be described. 
To find Error of Watcli by Absolute Altitudes. 
In finding local time by this observation it is not necessary that the 
longitude of the place should be known with any great degree of 
accuracy, as tho Greenwich date, obtained by the longitude in time, 
is only used for correcting the elements taken from the r Nautical 
Almanac/ and a considerable error in longitude would not produce 
any serious error in the declination or equation of time. The body 
should be observed as far from the meridian as possible, because, when 
nearly E. or W., errors, both of latitude and observation, produce the 
least effects on the hour angle. As a general rule, this observation 
should not be taken unless the sun or star is changing its altitude by at 
least 6' in 1 m. of time. The readings of the barometer and thermometer; 
should be noted, but for an approximate result are not necessary, 
