IIL.—TrRvuE SURFACES AND ACCURATE MEASUREMENTS. By D. 
W. Ross, A. S. M. E., Amherst, N. S. 
(Received 24th November, 1895 ). 
That absolute truth is almost unattainable, becomes apparent 
to the skillful mechanic as well as to the thoughtful scientist, 
and the degree of success of each may be measured by the 
nearness of approach to absolute accuracy, whether it be in the 
result of the scientist’s reasoning, or the more material product 
of the mechanic’s hands. 
For the production of flat surfaces of metal, the mechanic 
uses an instrument called the “surface plate,” which is simply a 
plate of cast iron well stiffened by ribs and resting upon three 
points of support to prevent springing, the upper side of which 
is carefully scraped by hand until its surface is approximately 
true. When one of a pair of such “surface plates” is placed 
with its trued surface above the other it will not immediately 
come in contact with the lower plate, but will for a time float 
upon the air confined between the surfaces, because the air can 
only escape at the edges, and, as the plates come closer together, 
it will do so more slowly so that a noticeable time, depending 
upon the truth of the surfaces, weight and size of plates, will 
elapse before the plates will really touch each other. When the 
air is fully excluded, or as fully as the truth of the surfaces will 
allow, the plates will adhere, or rather the atmospheric pressure 
on the outside will press them together. 
It is evident that, if we interpose between the surface plates 
a fluid, more viscid than air, such as oil, it will require a longer 
time and greater weight to expel it, and we may move the 
upper plate back and forth for a long time before the plates 
come into direct contact, the particles of fluid forming a perfect: 
system of rollers, upon which the metal will roll with very 
slight friction. If we provide means whereby the oil will be 
renewed, the iron will never come into contact, provided the 
(21) 
