1 10 Professor Renwick on the 



chance of error in the determination of the most usual denomi- 

 nation of dry measure (the bushel,) and it is itself the cus- 

 tomary unit of liquid. 



We shall now proceed to consider what ought to be the size 

 of this unit. The Act of Parliament, of 13th William III. cap. 

 5. prescribes that the standard bushel shall contain 21 50 inches. 

 But among the standards preserved in the British Exchequer, 

 for the bushel and its several parts, considerable discrepancies 

 exist, and it would be difficult to determine from which of these 

 the bushel of the Revolution was derived. Their several magni- 

 tudes are as follows : — 



Bushel. Oallon. 



2215.2 276.9 



2234.4 279.3 



2163.2 274.4 



2128.9 266.1 



Mean 2180.4 272.5 



Still greater discrepancies exist in the usual bushels of dif- 

 ferent parts of the United States, which vary in the two ex- 

 tremes from 1925 to 2358.6 inches, (giving the gallon the two 

 dimensions of 241.25 and 294.8,) as stated in the report of Mr. 

 Adams. 



Instead, however, of prescribing the cubic contents of mea- 

 sures of capacity, it would, as has been already explained, be 

 much better to determine their magnitude by declaring the 

 number of pounds of distilled water at the maximum density 

 and mean pressure of the atmosphere that the unit should con- 

 tain. If this weight be taken at 80 lbs., the bushel will be 

 2214.36 cubic inches, and the gallon 276.8. These dimensions 

 differ less from the standard now adopted in England of 2217.6 

 and 277.2 than either of the above dimensions do from each 

 other. 



This method has a great advantage over that of actual mea- 

 surement, inasmuch as the weight can be determined with suffi- 

 cient accuracy, wherever good balances and just weights are to 

 be had : while, to determine the cubic contents of a vessel is 

 one of the most difficult operations of practical geometry, and 

 is liable to such uncertainty, that in all recent investigations 

 into the magnitude of measures, their weight when filled with 



