OP ARTS AND SCIENCES. 323 



(2.) If only one microscope is used, either the movement of the 

 microscope along the horizontal defining line must be in the line of 

 the stops, or there must be a convenient way of placing the line stan- 

 dard at an equal and constant distance on either side of tlie line of the 

 stops, without the necessity of adjustment for i^arallelism and for 

 focus. If two microscopes are used, means of adjustment in three 

 planes become necessary. 



(3.) The pressure of the stops at each end must be equal and 

 constant. 



These three conditions are met in the following arrangement. 

 Upon a table attached to S, Fig. 1, two slides move freely upon cylin- 

 drical ways of rather small diameter, but they can be firmly secured to 

 the cylinder at any point by clamps. A second slide is mounted upon 

 each of these plates, which has a movement of about four inches upon 

 a cylinder and a flat way. The parallelism of these upper plates with 

 the plane described by the carriage K, is secured by an adjustable ver- 

 tical piece which passes through the jjlate near one edge and rests 

 upon the plane way, which has also a vertical adjustment. There are 

 two of these plane ways, placed equidistant from the cylindrical way 

 on either side. The carriage can therefore be reversed by simply 

 lifting it from its semicircular bearings. These upper slides terminate 

 with tempered steel stojjs, which were set by the maker in a line with 

 each other, and in the vertical plane which passes through the axis of 

 the cylinder. The forward motion is given by means of an adjustable 

 spiral spring, while a lever and stop control the backward motion. 



An adjustable graduated bar is firmly attached to each of these 

 plates, which has its upper surface nearly in a line with the stops. 



With this arrangement of stops it is possible to compare line with 

 end measures by the aid of either one, two, or three microscopes. 

 The method by two microscopes has decided advantages over any 

 other tried by the writer. The explanation, therefore, will be limited 

 to this method. 



(1.) The two double-slide plates are placed in such a position upon 

 the main cylindrical ways that the distance between the spring-stops 

 is e. g. about one meter. 



(2.) The end meter, supported at its neutral points, is placed be- 

 tween the stops, and an equal pressure upon each end is given by 

 means of the adjustable springs. 



(3.) Plate K is then placed nearly opposite one end of the meter, 

 and plate I nearly opposite the other end. 



The micrometer line of M-*^ must now be i>laced exactly in the verti- 



