so that normal atmospheric pressure = 76X 13.5951 X980.665 = 1.013250X dynes per sq.cm. (Atmospheric pressure at latitude 45° differs from the above in that the gravitational acceleration is taken as 980•616 An elaborate in vestigation at the International Bureau has shown the volume of the litre to be 1.000027 cubic decimetres (Travaux et Memoires du Bureau International, 1910, tome XIV). In determinations of volume which do not admit of a high degree of accuracy the cubic deci metre can be taken as equivalent to the litre.
In Great Britain and Northern Ireland the metric standard of capacity is the litre, represented (Order in Council, May 19, 1890) by the capacity of a hollow cylindrical brass measure whose internal diameter is equal to one-half its height, and which at o° C., when filled to the brim, contains one kg. of distilled water of the temperature of 4° C., under an atmospheric pressure equal to 760 millimetres at o° C. at sea-level and latitude ; the weighing being made in air, but reduced by calculation to a vacuum. In such definition an attempt has been made to avoid former confusion of expression as to capacity, cubic measure, and volume; the litre being recognized as a measure of capacity holding a given weight of water.
3. The British System.—The imperial standard yard is defined (Weights and Measures Act, 1878) as the distance, at 62.00° F, between two fine lines engraved on gold studs sunk in a bronze bar. This bar was cast by Troughton & Simms in Recent comparisons by the National Physical Laboratory (N.P.L.) show that this 1878 yard= o.9143987m., the present legal equivalent of the yard in the metric system is m., a value which makes r inch= 2•54 cm. correct to 1 part in a mil lion ,(International Critical Tables (I.C.T.), Vol. 1. p. 7).
The imperial standard pound avoirdupois is a cylinder of pure platinum about 1.35 in. high and 1.15 in. diameter (fig. 3). The grain is one seven thousandth part of this pound and the troy pound is equivalent to 5,76o grains. The standard pound is kg.
The standard gallon is the volume of io lb. avoirdupois of pure water as weighed in air against brass weights, the tempera ture of the air and the water being 62° F., and the barometric pressure 3o in. of mercury. This legal definition is incomplete in that it does not state the density of the brass weights ; in official comparisons the density is taken to be 8•143 g. per cu.cm. Legally the gallon is equivalent to 1 (see para. 5). The fluid ounce (apothecaries' measure) has a volume of pint (th gal.) thus 16 fluid ounces of pure water weigh 1 pound (avoir dupois).
In the measurement of the cubic inch it has been found (Proc. Roy. Soc., 1895, P. 143) that the mass of a cubic inch of distilled water freed from air, and weighed in air against brass weights (density =8.'3), at the temperature of 62° F, and under an atmospheric pressure equal to 3o in. (at 32° F), is equal to 252.297 grains weight of water at its maximum density (4° C). Hence a cubic foot of water would weigh 62.281 lb. avoir., and not 62.321 lb. as at present legally taken.
The imperial standard measure of capacity is a hollow cylinder (fig. 4) made of brass, with a plane base, of equal height and diameter; which when filled to the brim, as determined by a plane glass disc, contains ro lb. weight of water at 62° F weighed
in air against brass weights, when the barometric pressure is 3o in. A secondary standard measure for dry goods is the bushel of 1824, containing 8 imperial gallons, represented by a hollow bronze cylinder having a plane base, its internal diameter being double its depth. It may be noted that the term "Imperial" first occurs in the Weights and Measures Act of 1824.
In the United States the fundamental units of the national system are defined in terms of the metric system, and originated from British standards which, in some cases, have been altered since the original U.S. legislation was passed (Dictionary of Ap plied Physics [D.A.P.], vol. iii., p. 594; vol. i., p. Thus the United States unit of capacity, the gallon, is the old Queen Anne gallon and is equal to 231 cu. in., or 3.785 1. The U.S. pound is the same as the British pound, and the U.S. yard (3 g m.) is about 3 parts in a million greater than the British yard.
4. Materials.—The Matthey alloy iridio-platinum (9o% platinum, 0% iridium) is probably of all substances the least affected by time or circumstance and it is therefore used for the prototype metre and kilogram. It is very costly and though its coefficient of linear expansion (8.7X per deg. C) is small, it is not negligible, hence ordinary length standards are often made of Guillaume's alloy (invar) of nickel (35.7%) and steel (64.3%). This metal can be highly polished and is capable of receiving fine graduations. Its coefficient of expansion is very small—only 0.9X per deg. C. There appears to be some objection to the use of iridio-platinum for weights, as, owing to its great density (21.57 g. per cu.cm.) the slightest abrasion will make an appre ciable difference in weight ; sometimes, therefore, quartz or rock crystal is used; but to this also there is some objection, as, owing to its low density (2.65 g. per cu.cm.) the exposed surface is unduly large and a large buoyancy correction is necessary. For small standard weights platinum (density= 21.45) and aluminium (density =2.67) are used, and also an alloy of palladium (6o%) and silver (40%) (density = ii.00).
Ordinary weights, whether lacquered or plated with gold or platinum frequently gain in weight for years without any visible alteration and lacquered weights are liable to vary with large variations in the humidity of the air (for example by 0.2 mg. per 5. Effects of Temperature and Pressure.—The graduations on the imperial standard yard (fig. 2) are sunk below the surface of the bar, partly to protect them from damage, but chiefly so that they may lie in the plane of the neutral axis of the bar. The Tresca section of the prototype metre (fig. I) gives these ad, vantages and, in addition, (a) renders it possible to graduate the bar along the whole of its length, and (b) gives great strength for the weight of metal used.
The variation in the length of a metal bar with temperature makes it necessary to define the temperature at which standards involving extension are to be used. The choice of o° C by the