1864.] Report of the Great Trigonometrical Survey. 399 



ations are, consequently, not as valuable as the earlier ones; they 

 consist of 9 culminations, 64 lunar zenith distances, and 36 clock 

 stars. The whole of the Latitude observations have been reduced, 

 and found exceedingly satisfactory. There has not yet been leisure 

 to reduce more than a few of the observations for Longitude, but the 

 results obtained hitherto are satisfactory. The final resulting Longi- 

 tude will be communicated for publication in the Calcutta Gazette 

 as soon as ascertained. It should serve as an excellent datum for the 

 proposed Maritime Surveys, and save the expense of a series of voyages 

 between Madras and Port Blair, which would otherwise have to be 

 incurred to obtain a good chronometric determination of the Longitude 

 of Port Blair. 



[A tabular abstract statement of the field-work executed by each 

 party during the official year 1862-3 is given on the next page.] 



The Computing Officer has been employed in a variety of prelimi- 

 nary operations, which are necessary to form the basis of a general 

 reduction of the whole of the principal triangulation of this Survey, 

 which will shortly become necessary, now that almost the whole of 

 the triangulation of the tracts of country comprised in the great 

 quadrilateral figure connecting Calcutta, Karachi, Attok, and Purnea, 

 is completed. Though the triangulation has been executed with the 

 very best instruments, and though the system of observation which 

 was introduced into this Department by Colonel Everest, is more 

 rigorous and accurate than that of any European Survey, it is evident 

 that, in consequence of the vast length of each Series, and the 

 imperfections which necessarily attend whatever is the work of human 

 hands, each Series generates a certain amount of error, which becomes 

 apparent as linear error, on the termination of the Series on a 

 measured base line, while on the close of a circuit formed by two 

 Meridional Series, and the portions of the connecting Longitudinal 

 Series at their extremities, it produces errors of Latitude, Longitude, 

 and Azimuth. The dispersion of these errors in such a manner as to 

 obtain the most probable results of the whole, giving its due weight to 

 each fact of observation, and taking into consideration, the bearing of 

 every such fact on all the rest, is a matter of great intricacy and difficulty, 

 on which it will be necessary for me to consult with the ablest 

 mathematicians of the present day in Europe, before deciding on the 

 system to be finally adopted. Meanwhile, the necessary preliminaries 



