In fact, in 1963 the BIPM reverted to the original definition of the litre from 1795 (law of 18 Germinal an III) that stated that 1 L is equal to 1dm^3. The definition that defined the litre as 1 kg of pure water at 4 degrees Celsius was adopted around 1900 and the error of 28 ppm cropped up. The 10 pound of pure water definition of the Imperial gallon was taken at 62 degrees Fahrenheit = 16.6 degrees Celsius. Having to rely on ambient temperatures of water in such definitions is error prone and probably caused the error of 28 ppm. I an certain that the same problem also existed with the Imperial gallon. In 1814 a temperature of 56.5 degrees Fahrenheit = 13.6 degrees Celsius was proposed as part of the definition of the new British gallon. It would mean that the gallon would be 276.48 cubic inches; going to the higher temperature in 1824 made the Imperial gallon equal to 277.2 cubic inches. When one converts those two F-values to degrees Celsius, the result is a difference of exactly 3 K.
Han ----- Original Message ----- From: "Joseph B. Reid" <[EMAIL PROTECTED]> To: "U.S. Metric Association" <[EMAIL PROTECTED]> Sent: Wednesday, 2003-11-05 3:11 Subject: [USMA:27448] Re: Pint/lb vs Liter/kg John David Galt wrote in USMA 27443: When I first learned the metric system I thought this match-up was a great thing, too. Then I learned it wasn't really true. Pre-1963, the liter was defined as one kg of water (at max density or 4 degrees C), but that was actually 1000.028 cc. In 1963 the French committee in charge of the metric system voted, not to fix this discrepancy by adjusting either the meter or the kilogram, but to sweep it under the rug by redefining the liter as 1000 cc (and no longer related to 1 kg of water). In my opinion, this continuing discrepancy means that the metric system is no more "elegant" or "integrated" than the imperial system. Indeed less so, if an imperial gallon weighs 10 pounds. Now that they are thinking of getting rid of the standard kilogram block of platinum/iridium, I urge the authorities to correct their original mistake by reducing the size of the kilogram so that 1000 cc = 1 kg of water, thus defining the kilogram in terms of the meter and integrating the whole system of measurements better. John David Galt The comparison of the international prototype kilogram with secondary standards has an uncertainty limited to about 1 part in 10 to the tenth. A comparison with water can not be made with anything like that precision. For most practical purposes a cubic decimetre of water has a mass of 1 kilogram. The imperial gallon suffers from the same difficulty of definition in terms of a mass of water. It is in fact defined by a prototype container. (The definition of the imperial gallon involves the only factor of 10 in the entire imperial "system".)
