28 



If this method for the longitude is applied to determine 

 the relative positions of meridians, considerably distant frora 

 each other, it will be necessary to attend the more particu- 

 larly to all the corrections of the apparent difference of the 

 differences of AR.; and also to the obtaining of a correct 

 lunar rate for the mid-interval of the observations. To ac- 

 complish this part of the process with due precision, it were 

 much to be wished, that the AR/s of the moon, for noon and 

 midnight, were to be strictly calculated to the nearest se- 

 cond ; and in the nautical almanac columns of these AR/s, 

 given for portions of sydereal time. Till this be done, it 

 will be desirable for those, Avho wish for particular accuracy 

 in the results of their observations, to turn the i's longitudes, 

 as given to the nearest second, in p. v. of each month in the 

 nautical almanac, into AR.'s; which, though perhaps too te- 

 dious and laborious, to be done in general for several ob- 

 servations, might, however, be readily undertaken for one 

 valuable and important one. 



As the correct knowledge of the moon's rate at the time of 

 her mid-interval, between the different places of observation 

 is essential, where the estimate longitude is considerably un- 

 certain, then, the operation for getting the rate at the mid- 

 interval, must be repeated with the first obtained longitude, 

 and the process renewed with the more correct rate, thus eli- 

 cited. 



the terrestrial sequator have revolved on the earth's axis. \°. for each 8 seconds 

 of AR. so gained by the j . A degree in 54i. of lat. is = to 32,100 toises 

 nearly ; so that in this lat. a possible error of observation of 7e of a second is equal 

 to an error of 400 toises, or less than half an Irish mile in the longitude. 



