OF ARTS AND SCIENCES. 321 



securely locked after the contact has been made by the rack and pin- 

 ion movement. By means of a screw, shown at the left of H, the 

 core of the magnet is adjusted at the proper distance with respect to 

 the armature shown at the left of K, after contact has been made be- 

 tween the stops. A battery is employed of sufficient force to move 

 the carriage K, when the magnet is one eighth inch distant from the 

 armature. 



(e.) Comparison by means of tico Microscopes placed in a Horizontal 



Position. 



(1.) By means of two angle-plates, microscopes M and M^ are made 

 to occupy a position at right angles to the position shown in Fig. 1. 



(2.) The standards to be compared are placed upon the table S, 

 supported at their neutral points, and with their graduated surfaces in 

 a vertical plane. 



(3.) The comparisons are then made in the manner described 

 under (a). 



By this arrangement, the apparatus becomes a vertical comparator. 

 The intersection of the transverse line with the defining lines being 

 in the plane of the neutral axis, the deflection will be inappreciable, 

 even if the bar is not supported at its neutral points. The first sug- 

 gestion of this form of apparatus is generally credited to Neuman, 

 and the first construction to Wilde, who has given a full description 

 of the vertical comparator, constructed under his supervision, in the 

 Repertorium fiir Experimental-Physik, Vol. XIII. p. 567. But a 

 comparator of this form, invented by Lane, had been in use for 

 several years previous to this time in the office of the U. S. Coast 

 Survey at Washington. 



Under the fifth requirement the following methods are employed. 

 The frame of the stop V, Fig. 4, is securely fastened to the bed-plate 

 of the comparator. It has a vertical adjustment by means of gibbed 

 ways. The stop U is firmly attached to the plate K. This stop has 

 a vertical adjustment also ; but the line V U was made parallel with 

 the cylindrical ways by the makers. 



For observations of this class, the plate N, Fig. 4, which carries 

 the microscope M i, occupies a position the reverse of that shown in 

 the cut. 



The line bar to be compared is first placed upon the table at a 

 known distance outside of the line U V. The adjustments are then 

 made as follows : — 



VOL. XVIII. (X. 8. X.) 21 



