02 Proceedings of the Roydl Society of Victoria. 



The mean of the whole 42 observations is 9 h. 39 m. 59 s., 

 which only exceeds the adopted longitude, as determined by- 

 telegraph, by 4-8 s. The mean of the 21 with the moon east 

 is 9 h. 39 m. 59 s., of the remaining 21, with the moon west, 

 is 9 h. 40 m. The probable error of a single observation is 

 found from the residual errors to be J: 21 s., or a little more 

 than 5 miles of longitude. 



The result is, in my opinion, very satisfactory, for much 

 larger errors than these have been found in the longitudes 

 of places which have been determined with the large 

 instruments of astronomical establishments. In one instance, 

 that of the Lisbon Observatory, an error of nearly nine 

 seconds of time was discovered a few years ago, when the 

 longitude was measured by means of the electric telegraph. 

 As observations can be taken at sea with nearly the same 

 ease as on shore, except perhaps in the case of steamers, 

 where the smoke and vibrations are very annoying, the 

 results should be equally good, if the same care is taken in 

 the reduction ; but where Thomson's and similar tables are 

 used, the probable error would be rather larger. 



In conclusion, I may express the hope that this paper 

 may contribute towards rescuing this once favourite method 

 of tindino- the lono;itude from the nes^lect into which it has 



