80 Intelligence and Miscellaneous Articles. 



Next, suppose the temperature of the standard metre to be increased 

 30° F., and so to be made equal to the temperature of the English 

 standard. 



The coefficient of the linear expansion 1 nnnnnmAnAA 



of brass per degree F. being ! j"" 0-0000104344 



That for 30° will be 0*000313032 



Each length of the standard metre which corresponded with the 

 inch of the English scale will have expanded through a space equal 

 to -00795 millim., and the length which corresponded with 30 inches 

 through -238 millim. 



It is manifest that whatever lengths correspond at the temperature 

 of 62° F. will correspond also at all temperatures common to the two 

 scales. 



We have therefore the following relations : — 

 At the normal temperatures of 



the standards. At all common temperatures. 



1 inch = 25*39954 millims. 1 inch = 25*3916 millims. 



30 inches =761-986 „ 30 inches =761*748 



In a large number of double-scale barometers which have recently 

 been examined by my friends and myself, the metrical reading of 

 762 millims. has been found to coincide exactly with 30 inches, and 

 therefore to be in excess by nearly a quarter of a millimetre. I ac- 

 cordingly requested one of the makers to inform me on what prin- 

 ciple he made the graduation. His answer was as follows : — ■ 



" I graduate the English scale with great care by comparison with 

 an English standard ; I find from Guyot's Conversion-tables that 30 

 inches =762 millims. very nearly; I make a mark on the French 

 side of the scale opposite to 30 inches, and by means of a dividing 

 engine I divide the space between that mark and the zero-point into 

 762 equal parts, which I call millimetres. I believe this method is 

 universally practised in England." 



It is obvious that these millimetres are all too short, and that it is 

 impossible that the reading of a metrical scale so constructed can, 

 when reduced, ever agree with the English one. 



Mr. Tuckett has prepared a new set of conversion-tables, based 

 on the hypothesis of the two scales having the same temperature, 

 and has distributed them to several of the London makers, most of 

 whom immediately recognized the inaccuracy of the method they had 

 been employing, and promised to correct it for the future. It is of 

 course desirable that the production of inaccurate barometers should 

 be stopped as speedily as possible, and I therefore venture to give a 

 word of advice to purchasers. When a barometer is offered to you, 

 set the vernier to 30 inches, and if the metrical reading should be 762 

 millims., reject the instrument at once. I trust also that the matter 

 will not be lost sight of at Kew and Greenwich, and that barometers 

 with incorrect metrical scales will no longer be passed without com- 

 ment when sent to those observatories for examination. 

 I am, Gentlemen, 



Your most obedient Servant, 



51 Carpenter Road, Edgbaston, William Mathews, Jun. 



December 12, 1864. 



