﻿140 cox. 



This instrument is shown in Plate I, fig. 1. It is not sufficiently ac- 

 curate for the comparison of measures of length, because the adjustment 

 of the level when the telescope is moved from one end of the upright bar 

 to the other is necessary. 4 The range of the telescope is from three to 

 five .meters and with a level, the fineness of which is twenty seconds, 

 makes possible an error of nearly a millimeter. This is the best level 

 usually obtainable for this class of instrument, although sometimes one 

 sensitive to fifteen seconds is used. 5 Even this would not be sufficiently 

 accurate. 



As a rule the superiority of an apparatus is proportional to its cost 

 and in this case it was discovered that the only instruments made at 

 present which could entirely satisfy our requirements were very expensive 

 and much more complicated than needed. One was therefore constructed 

 after the following plan and I think for a simple apparatus, it has 

 features worthy of note. 6 In it the general method of comparison now 

 employed in nearly all of the most accurate linear measurements is fol- 

 lowed; that is, the length to be measured is placed parallel with the 

 standard bar and the images of the linear lines of the two brought suc- 

 cessively into the field of the observing microscope and their relative 

 positions determined. In the forms of the cathetometers previously 

 referred to, this has been done by observing the length to be measured in 

 the telescope and at the same time the scale on the long vertical axis. 



The new arrangement is shown in Plate I, fig. 2. In this apparatus 

 the short leverage of the sliding parts and the angular displacement of 

 the telescope are obviated. The base and upright travel columns for 

 the crosshead are the base and standards of a Doolittle torsion vis- 

 cosimeter. 7 The crosshead is supported by wires, passing over two "wheels 

 of the same size with a common axis mounted on a frictionless bearing, 

 and exactly counterbalanced by leaded weights. The bars to be compared 

 are suspended from the arch by vertical adjustible screws, in turn hori- 

 zontally controlled by thumbscrews from the front and back. The lower 

 ends of the bars are held in place by appropriate clamps. The microscope 

 is mounted on a travel on the crosshead. With the horizontal motion 

 of this travel and the vertical movement of the crosshead, freedom is 

 obtained for the microscope over the entire plane bounded by the vertical 

 columns of the apparatus. The microscope itself has a magnification 



4 This may have been partially due to the heat radiated from the operator 

 expanding the side of the bar nearest him. 



5 Ztschr. f. Instrumentenkuiide (1905), 25, 16. 



' Constructed in the workshop of the Bureau of Science, Manila, P. I., by 

 Mr. J. A. Gilkerson, Chief Engineer, Bureau of Science. 



7 Doolittle, O. S. : J. Am. Chern. Soc. (1893), 15, 173; J. Soc. Chem. Ind. 

 (1893), 12, 709. Wiley, H. W. : Principles and Practice of Agricultural Analysis 

 (1897), III, 343, Easton, Pa. 



