4 THE TRANSATLANTIC LONGITUDE. 



Observatory, by the same methods which have been employed for all the other 

 measurements from our boundary to New Orleans. 



Using the values above cited, the following are the determinations of the longi- 

 tude of Washington from Greenwich which have appeared best entitled to confi- 

 dence in recent years. 



I. From Edijjses and Occultations. — These furnished the values generally adopted 

 prior to the year 1848, namely, not less than 5"^ 8°" 14^ Thus Gilliss in 1846 used^ 

 Qh gm ^s g £qj. ^q provisional observatory on Capitol Hill which was,^ geodetically, 

 10105 east of the present observatory. And in the volume of observations made 

 in 1845, the first issued by the Washington Observatory, the adopted longitude is 

 given' as 5'' 8" 14'.64. 



Peirce's reductions, in 1845, of occultations observed by Bond at Dorchester from 

 1839 to 1841 gave"* 5'' 8™ 13\9; and Walker, from an elaborate discussion of all 

 available observations between 1769 and 1842, inclusive, obtained' 5*^ 3" ]4M6, a 

 value subsequently'' reduced to 13". 85 by change in the adopted longitude of Phila- 

 delphia, Cambridge, and Washington. 



In 1839 Walker had deduced a new value for the moon's horizontal parallax 

 from a discussion^ of the eclipse of 1836 May 14, according to which the mean 

 value used by Burckhardt, in the lunar tables employed in the computation of the 

 longitude, required an increase of 1".52; and he discovered' that, although the 

 probable accidental error of his former result for the longitude of Philadelphia was 

 but ± 0'.35, (subject, however, to the influence of any error in the adopted parallax 

 and semidiameter of the moon,) yet the employment of his new value of the hori- 

 zontal parallax would diminish the longitudes assigned to all the stations of the 

 Coast Survey by about two seconds of time. Prof Airy, at Greenwich, had, in 

 reducing the Greenwich observations of 1840, already adopted'' Henderson's deter- 

 mination,^" according to which Burckhardt's constant required to be increased by its 

 twenty-six hundredth part. So, too, Olufsen, from discussions," in 1837, of Lacaille's 

 meridian altitudes at the Cape of Good Hope, had inferred the need of an increase 

 of this constant by 2". 24, and Henderson, in the same year, from his own observa- 

 tions with the mural circle at Capetown, deduced^^ 1".3 as the requisite increase. 

 AU these investigations, though greatly varying among themselves, agreed in the 

 results that Burckhardt's value was decidedly too small, and thus corroborated the 

 change which Walker's computation of the eclipse of 1836 showed to be necessary. 

 Relying on these confirmations. Walker adopted^ the correction + 1".5 to Burck- 

 hardt's constant, and found that the trans-Atlantic longitude deduced from eclipses 

 was thus diminished by 2'.67 for the whole coast of the United States. The report 



' Gilliss, Astr. Obs. p. x. 



^ By Ellicott's original survey of Washington City. See Coast Survey Report, 1846, p. 12. 



3 Wash. Obs., 1845, p. 81. ' Coast Survey Report, 1846, p. 11. 



"■ Coast Survey Report, 1848, p. 113. « Ibid. 1851, p. 480. 



7 Transactions Amer. Phil. Soc, VI. .S83. ^ Coast Survey Report, p. 115. 



9 Greenwich Observations, 1840, p. slviii. "> Mem. R. Astr. Soc, X. 283. 



" Astr. Nachr. XIV. 226. " Mem. R. Astr. Soc, X. 284. 



