NATURAL PHILOSOPHY. 161 



that uniformity may be as well understood as the uniformity of lan- 

 guage or numeration. Suppose two metallic surfaces, by friction, or 

 otherwise, are made so plain and smooth, that when laid upon one 

 another every part of the surfaces is in equal contact ; and then sup- 

 pose a stratum, inserted between them, composed of particles of air, 

 which act like perfectly smooth rollers, the surfaces will then move in 

 contact with each other, in the most easy manner, owing to the lubric- 

 ity of the air. If the air, however, is excluded by pressure, the con- 

 tact becomes so complete that it is difficult to overcome. These 

 siii-faces are used as tests to other plane surfaces, and with these are 

 tested the ends of a standard measure of metal, which is placed in a 

 horizontal metallic bed, one end bearing against a metallic pin, while 

 against the other end another metallic pin is urged by a screw ; and if 

 this metallic bar suffer a change in its length, amounting to even the 

 millionth part of an inch, by temperature or otherwise, that change is 

 instantly perceptible. And thus the pin which bears against its 

 extremity is moved by a screw, having ten threads to the inch. On 

 the head of this screw is a wheel, consisting of four hundred teeth, 

 worked in a worm by another wheel, the rim of which is divided into 

 two hundred and fifty visible parts. As each thread of the screw cor- 

 responds to one tenth part of an inch, each tooth of the wheel upon its 

 head will correspond to the four thousandth part of an inch, and each 

 division of the wheel connected with the worm will correspond to the 

 millionth part of an inch. A change in the position of the wheel 

 attached to the worm, through one of the two hundred and fifty divis- 

 ions, is, therefore, rendered sensible, at the point of the screw which 

 bears against the standard bar, to the millionth part of an inch. The 

 accuracy of this micronietric apparatus was proved by placing in it a 

 standard yard measure, made of a bar of steel, about three quarters of 

 an inch square, having both ends perfectly true. One end of the bar 

 was placed in contact with the face of the machine ; at the other end 

 a small piece of flat steel was interposed between it and the machine, 

 whose sides were also made perfectly parallel. This was termed the 

 contact-piece. Each division on the micrometer represented the one 

 millionth part of an inch, and each time the micrometer was moved 

 only one division forward the experimenter raised the contact-piece, 

 allowing it to descend across the end of the bar simply by its own 

 gravity. This was repeated until the closer approximation of the sur- 

 faces prevented the contact-piece from descending, when the measure 

 was completed, and the number in the micrometer represented the dead 

 length of the standard bar to the one millionth part of an inch. Eight 

 repetitions in a quarter of an hour produced the same results, there 

 not being a variation of one millionth of an inch. This method was 

 termed the " system of proof by the contact of perfectly true surfaces 

 and gravity." By the application of this instrument standard gauges 

 for axles, taps, and other parts of machinery, which it is desirable to 

 maintain uniform, are constructed ; and so minute is its operation, that 

 magnitudes can be estimated that do not exceed the one millionth part 

 of an inch. 



14* 



