machines are composed of parts connected together by rigid and movable joints; rigid joints are necessary because of the expense, and in most cases the impossibility, of constructing framing and other fixed detail in one piece.

all moving parts must of course be independent of fixed parts, the relation between the two being maintained by what has been called running joints.

it is evident that when the parts of a machine are joined together, each piece which has contact on more than one side must have specific dimensions; it is farther evident that as many of the joints in a machine as are to accommodate the exigencies of construction must be without space, that is, they represent continued sections of what should be solid material, if it were possible to construct the parts in that manner. this also demands specific dimensions.

in arranging the details of machines, it is impossible to have a special standard of dimensions for each case, or even for each shop; the dimensions employed are therefore made to conform to some general standard, which by custom becomes known and familiar to workmen and to a country, or as we may now say to all countries.

a standard of lineal measures, however, cannot be taken from one country to another, or even transferred from one shop to another without the risk of variation; and it is therefore necessary that such a standard be based upon something in nature to which reference can be made in cases of doubt.

in ages past, various attempts were made to find some constant in nature on which measures could be based. some of these attempts were ludicrous, and all of them failures, until the vibrations of a pendulum connected length and space with time. the problem was then more easy. the changes of seasons and the movement of heavenly bodies had established measures of time, so that days, hours, and minutes became constants, proved and maintained by the unerring laws of nature.

a pendulum vibrating in uniform time regardless of distance, but always as its length, if arranged to perform one vibration in a given time, gave a constant measure of length. thus lineal measure comes from time; cubic or solid measures from lineal measure, and standards of weight from the same source; because when a certain quantity of a substance of any kind could be determined by lineal measurement, and this quantity was weighed, a standard of weight would be reached, provided there was some substance sufficiently uniform, to which reference could be made in different countries. such a substance is sea or pure water; weighed in vacuo, or with the air at an assumed density, water gives a result constant enough for a standard of [147] weight.

it is a strange thought that with all the order, system, and regularity, existing in nature, there is nothing but the movements of the heavenly bodies constant enough to form a base for gauging tests. the french standard based upon the calculated length of the meridian may be traced to this source.

nothing animate or inanimate in nature is uniform; plants, trees, animals, are all different; even the air we breathe and the temperature around us is constantly changing; only one thing is constant, that is time, and to this must we go for all our standards.

i am not aware that the derivation of our standard measures has been, in an historical way, as the foregoing remarks will indicate, nor is it the purpose here to follow such history. a reader, whose attention is directed to the subject, will find no trouble in tracing the matter from other sources. the present object is to show what a wonderful series of connections can be traced from so simple a tool as a measuring gauge, and how abstruse, in fact, are many apparently simple things, often regarded as not worth a thought beyond their practical application.

(1.) why are machine frames constructed in sections, instead of being in one piece?—(2.) why must parts which have contact on opposite sides have specific dimensions?—(3.) what are standards of measure based upon in england, america, and france?—(4.) how can weight be measured by time?—(5.) has the french metre proved a standard admitting of test reference?