OP ARTS AND SCIENCES. 299 



(/) 1. Report of Professor Hilgard upon the comparison of R.^* with 

 ''Bronze 11," made by Assistant Edwin Smith. 



2. Comparison of R,^^ with "Bronze 11," at Washington, with 



Comparator No. 1. 



3. Comparison of T""" with Meter " No. 49," at Washington. 



(^f) Preliminary comparisons of standards at Cambridge with Com- 

 parator No. 1, and with the Universal Comparator. 



(Ji) Investigation of coefficients of expansion. 



{%) Final comparisons of prototype standards. 



(y) Probable error of observations. 



(Jc) Tabular values of errors of subdivision. 



(l) Investigation of the equation between the Imperial Yard and the 

 Metre des Archives. 



Comparator No. 1. 



This comparator is of the most simple form. It consists of a bed- 

 plate 60 inches in length, 14 inches in breadth, and is ribbed to the 

 depth of 2 1 inches. V-shaped ways extend the entire length. A 

 carriage 8X6 inches is carried along these ways, either by hand move- 

 ment or by a rack and free pinion. The rack, which is placed midway 

 between the ways, extends the whole length of the bed-plate, and the 

 shaft which carries the pinion is placed at the centre of the carriage. 

 On each side of the centre, two independent platforms are mounted 

 with vertical adjustments, to which the microscopes are attached. 

 There is, besides, sufficient room at one end of the carriage for the 

 tracing apparatus, which may thus be made to maintain a fixed relation 

 to either microscope. 



Gibbed grooves extend the whole length of the bed-plate, to which 

 are fitted two stop-plates, 4x5 inches, which can be clamped firmly 

 to the bed at any point. Oval-shaped tempered steel stops are in- 

 serted in the end foces of these plates, and corresponding flat surfaces 

 of steel are inserted at each end of the microscope carriage. 



The flexure of the bed-plate is taken up by means of two screws, 

 which are attached to an independent plate beneath and near the 

 middle point. 



The V-shaped ways have a decided horizontal curvature, which 

 gave great trouble until the form of the curve had been investigated. 

 Several independent mvestigations of the effect of this curvature have 

 been made, both for the yard and the meter, with the following 

 results. 



