292 sun's distance. 



whether the longitudes of any of the stations named, except those in 

 Europe, are yet known with sufficient accuracy. 



Third. On the application of the method of difference of duration 

 of transits to the transit of 1882. 



It has been already pointed out that there are two tracts, each 

 sufficient to contain a number of observing stations, which are par- 

 ticularly well adapted to these observations. And it is specially to 

 be remarked that the command of a number of stations, sufficiently 

 near together to see the astronomical phenomenon in nearly the same 

 way, but sufficiently separated to take the chances of different states 

 of the sky, is very important. On occasion of the eclipse of 1842 the 

 astronomers at Turin saw nothing, in consequence of the cloudy state 

 of their sky, while the Astronomer Royal on the Superga, not five miles 

 distant in a straight line, saw all the phenomena of the eclipse. 

 Bearing this caution in mind, we Avill consider the circumstances of 

 the two tracts in question. 



The northern tract includes the whole of the United States of North 

 America. The observatories are numerous, and they possess an ad- 

 vantage which even yet is little known in Europe, namely, that from 

 the extent of galvanic telegraph, the habit of using it in the United 

 States, the public spirit of the nation and of the telegraph companies, 

 which would assuredly induce them to devote that wonderful auxil- 

 iary to the exclusive use of astronomy on an occasion so important, 

 and the absence of political suspicions, all the observing-stations 

 would for an observation like this be connected by the galvanic tele- 

 graph. (The Astronomer Royal adverts to the political suspicions, 

 not without some bitterness, for he has been prevented by them from 

 using a European telegraph for a single hour to determine the longi- 

 tude of an important continental point.) The peculiar advantage of 

 connecting, at least of comparing, all the observers' clocks, would be 

 of this kind. Suppose that there were ten observing-stations, and 

 that, in consequence of the changeable weather, the ingress only was 

 observed at five of these stations, and the egress only at the other 

 five*: If the clocks of the observatories were not connected or com- 

 pared, these observations would be totally lost. But if they are con- 

 nected, then every observation is referred to the absolute time of one 

 clock, say the Washington clock; and from a knowledge of the geo- 

 graphical position, a correction of the absolute time may be computed, 

 so as to deduce, from every observation of absolute time of ingress at 

 any station, what would have been the absolute time of ingress had 

 it been observed at Washington, and from every observation of the 

 absolute time of egress at any station, what would have been the ab- 

 solute time of egress had it been observed at Washington; and thus 

 we shall have five observations of ingress and five observations of 

 egress, all as if they had been observed at Washington and noted by 

 the Washington clock. Humanly speaking, therefore, we may say 

 that the probabilities for the accurate and efficient observation of 

 these phenomena in the United States are vastly superior to any that 

 could have been reckoned on in any former time, or to any that could 

 now be reckoned on in any other region. 



The southern tract is a part of the Antarctic land discovered by 



