Ordnance Survey Astronomical Observations. 361 



into the probable errors of the final results. These numbers 

 are given to hundredths of a second, but when we examine the 

 components, and find one star giving a result many whole 

 seconds from another, and all the southern stars having a per- 

 verse discordance to all the northern ones, we evidently 

 cannot depend for certain on the last hundredth. To within 

 how many hundredths of a second, then, can we depend =tz 1 

 What is the real breadth of the foundation on which we 

 can securely build a superstructure of theory and inference, 

 free from the effects of error of observation ? 



This quantity should be obtained with observations for lati- 

 tude and for longitude also, wherever there may exist so per- 

 fect a means as the electric telegraph for communicating 

 time between two stations. The next step will then be to de- 

 termine in each case the configuration of the neighbouring 

 ground, by careful contouring (a superb mode, by the way, 

 of settling the levels of a country for engineering and other 

 operations, though discountenanced by the late Committee 

 of the House of Commons) : and with the size of the hillocks 

 so ascertained, and with the nearest approach that can be 

 made to a knowledge of their specific gravity, their attrac- 

 tion may be computed. 



Then according to the character of the residual quantity, 

 obtained by applying the computed to the observed attrac- 

 tion, deduced by comparison of one with many stations, — 

 further astronomical observations should be instituted at 

 various places, until all the sources of local attraction, and 

 the means of computing their effects for all distances, shall 

 have been discovered. 



So far for the elimination of the disturbing effects of the 

 attraction of the mountainous masses. But that having been 

 ascertained sufficiently well for practical purposes ; it is hoped 

 that then the important scientific and physical result of the 

 weight of the world will occupy the attention of the survey- 

 ing department. So long a time has occurred since Dr 

 Maskelyne tried the great experiment on Mount Schehallion, 

 that much advantage might result from repeating his mea- 

 sures again, both terrestrial and astronomical, with improved 

 means : especially adding observations for longitude as well 

 as for latitude ; and so observing on four sides of the moun- 



