354 PHILOSOPHICAL TRANSACTIONS. [aNNO I766. 



mersions of the same year, or at least of the samo month, will recede almost 

 equally from calculation, if the observations are made with equal certainty; and 

 the emersions the same. Therefore the variations, and as it were leaps of errors 

 are to be attributed to the unequal goodness of the observations. If the obser- 

 vations made in the place to be determined differ almost equally from the tables, 

 it will be an indication that the assumed difference of meridians is the true one : 

 but if unequally, then how much it is to be augmented or diminished. If the 

 immersions seem to require the difference of meridians to be greater or less than 

 the emersions, it is to be ascribed to the unequal power of the tubes, and the 

 mean difference of the meridians may be accounted so much the more certain, as 

 it has more observations to confirm it. 



For example, to determine the longitude of a place in the island of Barbadoes, 

 by means of J 7 observations of the first satellite made there, in the months of 

 November and December 1763, and in January and February J764. Now 

 there were 2 1 other observations of the same satellite made in Europe in the 

 same months, in places well determined, of which 2 only are correspondentes to 

 the former, the medium of which gives 5*" 10'" ]4' for the difference of meridians 

 of Stockholm observatory and Barbadoes. To know whether that will agree 

 with the rest of the observations also; to all the observations in these months add 

 the last in October 1763, and the first in March 1 764, arranging them in due 

 order, and compute the true time, according to the tables, when each one 

 should happen in its own meridian. 



This calculation being made, it appears that the European immersions in some 

 places were later, and in others sooner, than by the tables they ought to be. 

 The former seem to be the best, but the latter, especially those which differ the 

 most, are rejected as doubtful. By a medium, all the immersions seem to agree 

 nearly with the calculation. But among 15 emersions observed in Europe, 

 mostly calculated, some at the most are 3P too early, excepting three at Tyma, 

 which are too late, and rejected. The medium of the good emersions were about 

 20^ too early. 



Considering the calculation of the other observations at Barbadoes, it appears 

 that, supposing the difference of meridians S"" 10™ 14% almost all of them are 

 marked much more before the computation than those of Europe ; but that they 

 differ almost equally from the tables, if the difference of the meridians be in- 

 creased 2T. Mr. W. therefore takes it to be very nearly 5*' 10"" 36*; as will 

 more plainly appear by the following correction of all the observations, with the 

 corrected Barbadoes calculation. 



The Paris observations were made by M. Messier, with a Gregorian telescope 

 magnifying 104 times. Those at Vienna by Fa. Hell, with a 4^-foot telescope, 

 the longitude from Paris being 56™ 11'. Those of Tyma by Fa. Weifs, with a 



