228 METHOD OF DETERMINING THE LATITUDE BY 



honour to detail in a former communication to the Society, 

 " On the Mode of determining Time with the Sextant" I begin 

 nearly 10' from noon to observe the sun's altitude, from an ar- 

 tificial horizon of oil, or quicksilver, and continue making as 

 many observations as I can accomplish until the sun has near- 

 ly the same altitude as when I began, which will be the case 

 about as long past noon, during which an expert observer will 

 easily take 20 altitudes, which, in most cases, will be sufficient 

 to enable him to retain all those that appear to be consistent, 

 and to reject those that differ much from the mean. The er- 

 ror of the instrument is to be carefully determined * ; and the 

 barometer and thermometer must be noted, and their results 

 deduced, agreeably to the two examples that accompany this 

 memoir. 



I have always found it the readiest mode to correct the 

 noon declination, as given at Greenwich, first, for the differ- 

 ence of meridians, and next for the time of the observations, 

 which is most readily done by calculating what the change in 

 one minute of time will produce, and then multiply that by 

 the number of minutes required for the given observation f. 



Having 



a 



V 



* At sea, or where it is not required to proceed with the greatest accuracy, 

 the mean refraction may be used, without having recourse to the correction for 

 atmospherical pressure and temperature. 



•f- If the place of the observer is west of the first meridian, it becomes neces- 

 sary to add the proportional part to the noon declination, for the difference of me- 

 ridians, from the Vernal Equinox until the Summer Solstice, and from the Au- 

 tumnal Equinox to the Winter Solstice ; but for the rest of the year it must be 

 subtracted. If the observer is situated east of the first meridian, it must be sub- 

 tracted during the spring and autumn, but added during summer and winter. 

 Had the sun no motion in declination during the time that the observations conti- 

 nue, it would be unnecessary to apply any correction, as in the case of observing by 

 the stars ; but although the quantity is very small during so short a period of time, 



stili 



