lOo [ 237 ] 



moon and the sun, we lose, of course, tl.is advantage of observing, in the 

 same day, such coupled series us I have been insisting on. But here ex- 

 perience shows that an observer has greater opportunity in the daytime for 

 .good adjustment, and better observing, than at night. If, then, he avails 

 himself of this, by extreme care he can still obtain results of satisfactory 

 exactness for geography. Further, in this case he has often the occasion 

 of improving the result by star distances, measured either in the evening 

 after sunset, or in the morning before sunrise; and if in such star dis- 

 tances there be found some lying on the side of the moon opposite the 

 sun, these night observations serve effectually to balance against those 

 made in the day. It is true, we cannot say here that observations, al- 

 though so made on both sides of the moon, but yet at great intervals of 

 time, compose what I have called just now coupled series; because it is 

 not exact to assume that the instrument is, during and after these intervals, 

 under the same physical conditions, and in the same degree of adjustment. 

 But, at least, we may be sure that such observations combine in favor of a 

 longitude more advantageously than if they had been all partial, and taken 

 upon one side only of the moon. 



In every case we see that the method is to provide, as much as possible, 

 distances observed to the east and to the west of the moon, and to take 

 the mean between the average of those on one side, and the average of 

 those on the other. In general, five or six couples of series, each seties , 

 containing 10 or 12 individual observations, arc capable of giving a longi- 

 tude, whose error will be within limits from one to three or four minutes 

 in arc. 



The method of lunar distances may be put in practice on land, as is 

 sometimes done for hydrography ; and it becomes, then, of great useful- 

 ness, by the facility it affords, with the aid of well-regulated chronometers, 

 of referring to one principal station all the series of distances taken at 

 different stations that have been occupied in several successive days, so 

 as to introduce them all to a concurrent determination of the absolute 

 longitude of this principal station, just as if they had been all made at 

 that one station. In this way the geographer, who can rely upon his 

 chronometer, with sufficient indications of favorable weather for a few 

 days, and is well informed by his guides of the country he has to pass 

 over, may dispense with a long halt at one of his stations, and then gain 

 time — not unfrcquently precious, from the circumstances in which he finds 

 himself placed. 



It is in accordance with these methods that the observations have been 

 made, from which are deduced the geographical positions presented in the 

 table at the end of this report. 



These positions are, in all, 240 ; of which 39 are due to the labor of 

 others, and have been collected for use, from the sources named in the 

 appropriate cohmm of the table. The remainder, which are my own, 

 are in number 201; resting upon about GOO series of 5 observations each, 

 for the determination o{ time ; and about 500 series of from 10 to 20 ob- 

 servations each, for the determination of latitude and longitude: making, 

 together, 1,100 series, composed of nearly 11,000 astronomical altitudes 

 and distances. 



As to the meteorological observations for the determination of the ele- 

 vation of the stations above the sea, it is not possible, without a tedious 

 and unnecessary research, to state their number accurately. The table 



