BETWEEN SURFACES MOVING AT LOW SPEEDS. 
511 
By means of an apparatus which differs essentially from any previously employed, 
we have been able to make definite measurements of the friction between surfaces whose 
relative velocity varied from about 0-01 foot per second down to about 0 , 0002 foot per 
second, and have found that in those cases in which the static coefficient largely exceeds 
the kinetic, the kinetic friction gradually increases as the velocity diminishes between 
those limits, so that in all probability there is continuity between the two kinds. 
In designing an apparatus for the purpose of carrying out this inquiry the following 
requirements suggested themselves : — That the velocity of the rubbing surfaces (besides 
being exceedingly low) should change very gradually, so that the acceleration might 
be capable of measurement at velocities differing only slightly from one another, and 
differing as little as possible from rest. That the change should be from motion to rest 
rather than from rest to motion, so as to avoid any jerk at the instant of passage from 
one to the other state. That no force whatever except the friction to be measured 
should take part in the action ; especially that the means adopted of registering the 
motion should be such as to cause no retardation. That the surfaces in contact should 
not change progressively during the motion, but should be periodically restored several 
times during the motion, so that any change observed to take place in the friction might 
not be due to a specific change in the surfaces : this last condition might perhaps be 
more generally expressed by saying that the whole apparatus should periodically return 
to exactly the same configuration. It was also desirable that a very small change in 
the velocity should cause a complete change of the rubbing surfaces. 
The present inquiry being entirely limited to the question of what influence velocity 
has on friction at very low speeds, we did not consider it necessary to provide means of 
measuring the friction at speeds exceeding about 0-01 foot per second, and that is the 
highest velocity to which our determinations extend. Further, in order to make the ap- 
paratus as well conditioned as possible for this particular purpose, it was not arranged so 
that the intensity of pressure on the rubbing surfaces could be either measured or varied. 
The apparatus with which the experiments were conducted, and which was designed 
with a view to the fulfilment of the above conditions, is shown in fig. 1, Plate 20. A is 
a disk of cast iron turned on both sides as well as round the circumference, so as to be 
exactly cylindrical, and weighing 86-2 lbs. Its diameter is nearly 2 feet and its thickness 
f inch. The disk is supported by means of a steel spindle (a), the ends of which, resting 
on the bearings b b, are equal cylinders with a diameter of only about 0T inch. The 
small ends of the spindle are shown more clearly in fig. 2, which gives a section of the 
apparatus in a vertical plane through the axis of the disk, and also in figs. 3 & 4, which 
give a full-size side elevation and vertical section through the centre of one bearing. 
The bearings ( b b) which carry the ends of the axle ( a ) were made successively of the 
various materials whose friction against the steel axle it was desired to determine. 
These pieces are fixed by screws to strong iron uprights forming part of the frame B. 
The bearings consist of rectangular notches, the lower surfaces of both being in the 
same horizontal plane and each pair of corresponding sides in the same vertical plane. 
4 c2 
