LONGITUDE. 



■ Let us examine one comparifon, for the fake of example, 

 in the above table ; fur inftance, Marcli 5th. It appears, 

 that on that day the right afcenfion deduced from the 

 quadrant obfervation differed + 8 '.6 from that obferved at 

 the tranfit. Now it is prcfumed, that a part of this error 

 mav be in the diviiions of the quadrant, or in the affumed 

 latitude, or in the obhquity of the ecliptic : \vc, therefore, 

 compare this refult with its correfponding one, Ocl. nth, 

 when we find the error to be — 5'. 4; hence we infer, that 

 :j."2 only is to be attributed to the error of the tranfit ob- 

 fervations, and that i."6, or the half, is the real error of the 

 catalogue common to every ftar ; fince, had that quantity 

 been added, the pofitive and negative error would have been 

 equal, and would have been therefore afligned altogether to 

 the folar obfervations. 



The beauty of Dr. Mafkelyne's method, which we have 

 thus endeavoured minutely to defcribe, confil^s in this, that it 

 is not only extremely independent of thofe errors that are 

 moil likely to occur in a feries of folar obfei vations, but that 

 it is capable of furniihing a clue to afcertain both the amount 

 and caufe of thofe errors. As this would lead us to an invelli- 

 gation rather foreign to the prefent fubjecl, we fhall not at 

 prefent enter into thefe coniiderations. 



Secular •variatior. in the longitude ofthejjxedjlars. 



When the longitudes ar.d latitudes of a number of ftars 

 are determined for a given period, thefe are computed for 

 any diftant period, by applying the preceffion of the equi- 

 nodial points, and hkewife the fecular variation for each 

 particular liar, and for which purpofe a very accurate table 

 has already been given under Latitude. This fecular varia- 

 tion arifes from the real change of pofition in the ecliptic 

 itfelf ; inafmuch as this affefits the fituation ef the equi- 

 nodlial point. The effett is common to all liars ; and, there- 

 fore, this part of it only influences the quantity of the ge- 

 neral preceffion ; but becaufe the pofition of the ecliptic is 

 really changed among the conftellations, each flar becomes 

 Teferred to a new point. 



Though the trigonometrical inveftigation of the exacl 

 quantity of thefe changes is extremely complicated ; yet the 

 principle may be rendered fufficienily intelligible, by recol- 

 lecting that a change in the pofition of the equator diilurb- 

 ing the equinoctial points, produces a change in the longi- 

 tudes, right afccniions, and declinations, the latitudes only 

 remaining the fame. But a fimilar change in the ecliptic 

 produces a change in the longitudes, latitudes, and right 

 afcenfions, whiill the dechnations remain unaltered. In 

 other words, the difplacement of the equator affetls every 

 thing but the latitudes, and a difplacement of the ecliptic 

 every thing but the declinations. 



On the methods of deter mlnir:g the pojitions of places on the fur- 

 face of the earth, or their longitudes and latitudes. 



The general nature of the problem having been already 

 explained under Latitude; and feveral practical methods of 

 determining the longitude having been defcribed at great 

 length under Chkokometer and Degree ; we have now 

 to explain a variety of allronomical procefies which have 

 been devifed and brought to a great ftate of perfection 

 within thefe lall fifty years. Longitude, being only a relative 

 term, to find the longitude of a place, is, in faCt, to determine 

 the difference of the longitude of two given places. And 

 here we may obferve rather a curious circumllance, which is, 

 that though the problem is in its ftatement purely geogra- 

 phical, yet it can only be fo'ved by the aid of allronomy, 

 except upon the hypothetical fuppolition of a trigonometrical 

 meafurement extended over the whole furface of tlie earth, 

 or at lead over a great circle of its circumference. Tliis 

 being impoflible, we mull have recourfe to the general 

 principle we have fo often had occafion to refer to in former 

 aftronomical articles. We fuppofe, at any given moment. 



every point of the convex furface of the earth eor- 

 refponds with fome point in the concave furface of the 

 heavens, called its zenith ; and as the angular diilance is the 

 fame on each, by meafuring the angular diilance of the celef- 

 tial arc, which is always acceffible, we obtain the correfpond- 

 ing and equal angular diilance of the terreftria! arc, which 

 otherwife would be practically im.poffible. Thus, for in- 

 ftance, one perfon at London, and ar.other at Jama ca, have 

 no means of knowing the exa£t proportion of the earth's 

 circumference intercepted between them, except, indeed, 

 by the inaccurate eflimate of the length of a fhip's track in 

 failing from one place to another ; but if, by fome 

 artifice, each could afcertain, at any given moment, his 

 zenith point in the heavens ; then, as the angular diilance 

 of thefe zenith points could eafily be meafured, the cor- 

 refponding terreitrial arc would immediately be determined. 

 Now to this, or fome very fimilar principle, may every pro- 

 cefs fur finding the longitude be referrer'. 



The inveftigation of the fubjedt will be much fimplified, if 

 we fuppofe the equator, inltead of being divided into 360 de- 

 grees, to be divided into 24 parts, and each part into Co, and 

 iubdivided again into 60. As each of thefe larger divifions 

 palles under a celeftial meridian in one hour of fidereal time, 

 they are called hours to avoid circumlocution, though it is 

 evident that a portion of a line cannot be an hour, or any 

 part of time. But as the difference of meafure will be 

 expreffed in the fame terms as the difference of time, this 

 mode of divifion is extremely ufeful, and fhews us at once, 

 that to determine the difference of longitude between two 

 places, is equivalent to determining the difference of ap- 

 parent time that exills between two places at any one given 

 inftant. The moil obvious way of accomplifhing this, 

 is for two obfervers to watch fome inflantaneous pheno.menon, 

 and to mark tlie inftant of apparent time at which each 

 obferved it. The inltantaneous explofion of a mafs of gun- 

 powder is extremely well adapted for this purpofe when the 

 diftance is not great, and has been fuccefsiully employed in 

 the iouth of France, and in the north of Europe. It is 

 evident that this method can only be employed for very 

 limited diilances : for places more remote, we are obhged 

 to recur to the celeftial phenomena, and we felect thofe 

 which have the greateft refembkince to the above, that is, 

 which are the moil inftantaiieous, and which appear the 

 moll nearly alike to two obfervers at the fame aftual inftant 

 of time. Unfortunately, there are none which unite thefe de- 

 firable combinations of circumftances. Eclipfes of .Jupiter's 

 fotellites, and of the raoon, unite them in a very confiderable 

 degree, and accordingly have been employed to great advan- 

 tage, particularly in the early ftate of geography, and in 

 cafes where the fituation of the place was previoufly un- 

 known. 



An occultation of a fixed ftar is a very inftantaneous phe- , 

 nomenon, but it is not feen at the identical inftant of aftual 

 time by each obferver ; for, from the vicinity of the moon and 

 its conlequent parallax, it may to one obferver appear to pafs 

 over a fixed ftar, when to another it may appear to pafs 

 entirely over or under it : hence, even in tlie caf\; where 

 an occultation is obferved by two pcrfons, the difference o£ 

 longitude cannot be inferred by finiply noting the difference 

 of time at which the phenomenon happened to each obferver. 

 This defeft, however, may be completely fupplied by calcu- 

 lation, and therefore it is juftly con!idcr<d as one of the 

 moil accurate methods that can poftibly be devifed. We 

 fliall refer our readers to aftronomical v/riters for examples 

 of the various methods : our object at prefent is only to give 

 a flcetch of the dift'erent principles on which they are 

 founded. 



The longitude of a place on land may likewife be found 



wi:b confiderable exaclneis, by obferving the paffage of the 



~ -j-j mooa 



