ON THE ABSOLUTE MEASUREMENT OF HARDNESS. 219 
which s is alone visible in the figure. The plate 0 0/ is about 1.6 thick, 
and its lower face is flush with the corresponding face of p. The whole 
“ase can be moved in the gutter of the bedplate, and clamped in any 
position by aid of the strong screw 8). It is, therefore, easily possible 
to place a part of the glass plate p opposite the vertex of the test lens 
i. 'The microscope M is similarly movable, and the clamp screw S, 
admits of an adjustinent relative to the point of contact to be observed. 
It is expedient to fasten the lens 1 to the top face of the plug 7 with 
cement. The plate p, however, exactly fits the hole in 0 0’, and adjusta- 
bly hinged stops prevent it from falling out. The microscope J con- 
tains an ocular micrometer m, and since illumination from beiow is 
clearly impossible, the light of a lateral gas flame I is retlected down- 
ward by the prism 2, small enough to only half fill the right section of 
the tube. After impinging on the lens and plate, the rays are reflected 
upward through the open half of the tube and the micrometer m, finally 
reaching the eye of the observer. The long arm of the lever H’ abuts 
against the screw A, and its play may, therefore, be stopped short 
high enough to prevent all premature contact between the test lens 
and plate. When A is screwed down, moreover, the long arm would 
much overbalance the weight of the short arm H. To obviate this a 
duplicate wrought-iron arm W has been added, along the free end 
of which a weight w’ may be adjusted to counterpoise the long arm. 
The form given to W is such that the center of gravity of the 
lever as a whole is not seriously depressed, and a balance sufficiently 
sensitive for the present purposes is thus secured. The counterpoise 
w’ is to be fixed so that the position of equilibrium may leave a little 
space between the test lens and the plate. Little rings 7 surrounding 
the pin g are then added until an almost pressureless contact is initially 
obtained. However slight, a true contact is always easily recognized 
by the passage of the colored interference rings into a black spot. In 
order that this initial contact may easily be reproduced, and the prog- 
ress of an experiment may at any time be checked, a second lever 
K K’, supported by the pillar U near the end of the bedplate, is at 
hand. <A stirrup R&, from the hook of which scale pans of different 
sizes may be suspended, suitably connects both levers (knife edges ¢ 
and c’, as shown in figure). By a play of the screw B the K end of 
the lever may be depressed and the A’ arm raised. In this way the 
stirrup is lifted off of the knife edge ¢, and the lever H’ is therefore 
unloaded. Conversely an opposite play of the thumbscrew B depresses 
K’, and the load is therefore transferred from the knife edge c’ to the 
knife edge c, whereby the lever is loaded, gradually or expeditiously, 
at pleasure. These adjustments enable the observer to earry out the 
necessary operations smoothly, and without any danger of jarring or 
striking the parts to be tested together. The elasticity of the long 
lever, moreover, is in favor of a perfectly uniform and slow intensifica- 
tion of the pressure to be applied. It was my plan to return to the 
