302 Proceedings of the Royal Society. 



missioners of weights and measures, who find that the latter con- 

 tains 252.458 gr., each grain being the 5760th part of the stand- 

 ard troy pound. 



The avoirdupois pound is fixed by assigning its proportion to 

 the standard troy pound, so as to contain exactly 7000 such 

 grains. 



The imperial standard gallon is defined by stating its contents 

 under the same circumstances of temperature and pressure, at 

 lOlbs. avoirdupois of distilled water, and the bushel by its con- 

 taining SO such pounds. 



The author, having in compliance with a request of the Lords 

 Commissioners of His Majesty's Treasury, undertaken to superin- 

 tend the construction of, and to adjust, the principal standards to 

 be deposited at the Exchequer, Guildhall, Dublin, and Edinburgh ; 

 Mr. Dollond was directed to prepare those of linear measure, and 

 Mr. Bate those of weight and measure, the proper quality of me- 

 tal for the latter purpose being determined by experiments insti- 

 tuted for the purpose. The experiments for adjusting them are 

 then given in full detail. The troy pounds were first adjusted, 

 and the exactness with which this operation has been performed 

 may be appreciated from this, that the final errors of none of them 

 exceeded 22 ten-thousandths of a grain. When brought so near, 

 it was of course not thought necessary to attempt further correc- 

 tion. 



The avoirdupois pounds and the weights of the gallon of water 

 were then derived from the troy pounds, and finally adjusted, like 

 them, by enclosing within the weight in hollows left for the pur- 

 pose, wires equal to the errors ascertained to exist in them. The 

 weights of these wires in each case is stated, so that should they 

 by any accident be taken out and lost, they may be restored. 



He next describes the method used in adjusting the gallon itself, 

 the method of filling it exactly, and of weighing it when filled, 

 together with the corrections depending on the circumstances of 

 temperature and pressure under which the experiments were made. 

 As a final result, it appears that one only of the gallons was ulti- 

 mately found in error to a greater extent than G tenths of a grain, 



