and the Methods of calculating their Results. 4 1 9 



[1], as the time of conjunction was there determined without 

 being affected by an error in longitude (or right ascension), 

 whereas such an error here affects the difference of meridians, 

 and consequently likewise the time of conjunction. But it is to 

 be observed that this contradiction is only apparent, and may be 

 considered as arising from the omission of that term of the cor- 

 rection which is to be added to the time of the conjunction cal- 

 culated by that method, and which depends on the error (x) of 

 the assumed difference of meridians as explained in [2]; if 

 this correction is added and x eliminated by the comparison of 

 the time of conjunction found with the one resulting from the 

 tables and A a, the dependence on this quantity will likewise 

 be perceived. Both methods only differ by being made to 

 depend on different unknown quantities. The determination 

 of all the 5 quantities s, £... by observations of an occultation 

 of a star, is mathematically possible; but it is easily perceived 

 that if these quantities are thus kept separate, small errors of 

 observation will greatly affect their determination, unless the 

 observations were made at the most proper spots of the globe, 

 and not confined to the small part of its surface contained 

 between the observatories of Europe. The advantageous se- 

 paration of the two last quantities from the others, requires, 

 for example, that at two of the places of observation the times 

 of the phenomenon should be very different, which will be the 

 case if at the one it takes place a little after the rising of the 

 moon, and at the other a little before her setting; the last is 

 separated from the rest only by the difference of the value of 

 /3 for the different places of observation. 



The difficulty of producing a concurrence of favourable 

 circumstances induces the belief that the determination of the 

 excentricity of the meridians of the earth for which observa- 

 tions of occultations of stars have been proposed (without, 

 however, sufficiently developing the greatest possible advan- 

 tages to be obtained by them), may always be founded on more 

 successful methods. Besides this difficulty, the mountains pro- 

 jecting on the limb of the moon, and other probable deviations 

 from the globular form, may and will spoil observations good 

 in themselves ; the immersion and emersion can rarely both 

 be observed with accuracy ; and lastly, the advantageous se- 

 lection of the places of observation is much restricted by the 

 presence of the sun above the horizon. I believe, therefore, 

 that the calculation of the influence of all the five unknown 

 quantities will only have an interest for the purpose of judg- 

 ing how far they may affect the results of the calculation, but 

 not for their determination. 



3 H2 [11.] In 



