TRIGONOMETRICAL SURVEY OF INDIA. VJQ 



We have still therefore much to abolish before we can arrive at any 

 uniformity of measures, but in the present case it is an object to avoid 

 confounding the subject of measurements made with a view to determine 

 the figure of the earth, with those whose object is the establishing the 

 length of the metre. The latter seems to be now generally allowed to be 

 a fanciful attempt, for what can it matter, whether we get our standard 

 unit from nature or from art? Whether we first fix our foot or yard, and 

 deduce the quadrant of the meridian in terms of it ; or derive the length 

 of our standard from the length of the second's pendulum, or from an 

 aliquot part of the computed quadrant ? 



But to the determination of the figure of the earth, the relative values 

 of the different national units which were employed, must be accurately 

 known ; moreover, as all operations undertaken for the purpose must 

 depend on some base, which has been measured under different states of 

 the thermometer, and perhaps under great alternations of heat and cold, 

 it follows of course that the exact mean temperature must be known and 

 corrected, unless some method can be devised of introducing into our 

 apparatus the principle of its own compensation. 



It is not an easy task, however, to ascertain the exact tempcraturitof 

 a rod of metal ; thermometers may be placed near it, but they will only 

 give the temperature of the surrounding air, which may differ very sen- 

 sibly from the metallic bar. 



The mode of measurement with a steel chain has hitherto prevailed 

 in England and in India, but it is very objectionable principally on the 

 following accounts. 



1st. The impossibility of knouinu,- its exact tcni[)craliuo and the 

 consequent allow ance to be made for expan-^ion. 



