35 



derived measures, viz., the one cubit foot, the five cubic foot, 

 and ten cubic foot holders at the Standards Department, the 

 greatest precautions were taken to maintain the temperature 

 of the water in the tanks of the holders and in the cistern of 

 the cubic foot apparatus, the same as that of the air of the 

 room by means of hot and cold water properly mixed. But 

 with all these precautions, and the use of screens to keep the 

 heat of the body from affecting the contents of the bells, there 

 was frequently a difference of 1° Fahr. between the two 

 measures. 



The great difficulties experienced in maintaining a body of 

 water at a uniform temperature led the Board of Trade to 

 instruct the Cambridge Scientific Instrument Company to con- 

 duct experiments with a view to solve this question. The 

 report of their inquiries, dated December, 1885, shows that the 

 maximum variation of temperature in fourteen days was '04»° 

 centigrade, while the variation in the room itself was about 

 4° centigrade. This was effected automatically, and the 

 results are satisfactory so far. 



In the first case water is used as a means of comparing the 

 measures, as in the bushel measure, but in the latter saturated 

 air has to be used, hence the difficulty is augmented. 



The limit of error for official standards proposed by the 

 Astronomer Royal and approved by the Treasury is '005 for the 

 cubic foot, "025 for the five cubic foot, and '05 for the ten 

 cubic foot. As the errors in the above standards were proved 

 to be, for the one cubic foot, • — '00053 cubic foot ; for the five 

 cubic foot, — *00667 cubic foot ; and for the ten cubic foot, 

 — *02495 cubic foot, they were passed as correct, but these 

 errors are allowed for and taken into consideration in using 

 these standards. 



Although these, viz., the cubic foot bottle and gasholders, 

 are the only forms specially legalised, Clause III. of the Act 

 gives power to adopt other forms, and under this clause 

 "secondary derived standards" have been constructed and 

 adopted, with the object of obtaining a portable standard. 

 They take the form of wet meters, and are capable of almost 

 absolute accuracy, as I have found myself. At our Standards 

 room we have a 50 light and a 100 light meter as models under 

 the Act of 1881, which were made by Alex. Wright & Co., and 

 are very satisfactory. 



COINAGE. 



Amongst other standards of weight are coin weights, corres- 

 ponding with the standard weight of each coin of the realm, 

 and with these the coin weights used by Banks throughout 

 England and by the Mint are verified. 



The duty of verifying coin weights used by the public was 



