MEASURING APPARATUS OF THE 



centre of the axis, which works in the iron bar on the line produced 

 through both axes, we ought to arrive at a point which would in no wise 

 be affected by the forces arising from the jutting outward or shrinking 

 inwards of the ends of brass and iron bars in consequence of increase of 

 temperature. 



Upon the principles of similar triangles, this point would form the 

 vertex of a set of triangles whose bases are the relative increments of 

 the brass and iron bars from heat, and as there would also be some 

 nodal point on a tongue similarly situated at the other extremity of the 

 compound bar, therefore the distance between these two nodal points 

 would be constant, and in no wise liable to be affected by changes of 

 temperature, provided such increase or diminution of temperature were 

 suffered by each metallic bar at the same time and in equal intensity. 



The difficulty then is to mark these neutral points and it is accom- 

 plished as follows. Two miscroscopes, with moveable wires, are fixed on 

 solid stone pillars let into a wall of masonry, and when this has had time 

 to dry and become perfectly consolidated, the moveable wires are adjusted 

 over the two dots of an iron, or other standard often feet in length, at the 

 temperature of 62° of Fahrenheit. 



The bars are then heated in an oven as equably as possible up to the 

 temperature of 180° or upwards, in that state they are taken out and 

 placed under the mural microscopes, and a point on one tongue having been 

 assumed, the corresponding point on the opposite tongue is sought for, which 

 maintains an invariable distance from it during the process of cooling. 



This operation in slow and tedious like all works of approximation, 

 but it is particularly so because it appears from experiment that the brass 



