LXXXIV. HERBERT E. ROSS. 
The dynamometer used was specially designed by the author to 
carry out the following conditions :— 
1. Capability of taking various lengths of belt. 
2. Capability of adjusting the tension on the belt at rest to 
known amounts. 
3. Of gradually and evenly adding load to the brake to a definite 
slipping point of belt. 
4. Of recording the relative speed of the driver and driven 
pulleys. 
5, Of recording the load, speed, and duration of experiment. 
6. Of automatic tension and release of brake.’ 
The dynamometer brakes hitherto in use would not meet all the 
desired conditions, more especially (3) and (6), and the common 
difficulty of absorption brakes of the heating and consequent 
expansion of the brake pulley, and the variation of the friction 
on the same, due to imperfect lubrication would have been a factor 
in unsatisfactory working. 
The machine as designed consisted of two 12” x 6” steel beams, 
true on bed and edge, laid lengthwise on two 12” x 6” oregon 
beams and bolted thereto. The pair of girders were distanced 
parallel apart 3’ by cast iron distance brackets. The first receiv- 
ing pulley (A) and the driver pulley (B) were mounted on a 3” 
shaft at one end of the bed guides. Upon the steel beams, and 
traversing the greater part of their length upon friction rollers, 
was the trussed wooden frame (c). This frame carried the brake 
shaft and brake pulley (D), and driven pulley (E), and all the 
recording gear of the machine. 
The brake wheel was 36” diam and 7” face and curled the brake 
of tallow-wood blocks secured by a link belt, one end of which 
was attached to the trussed lever (F), and the other was coiled — 
round in the direction of rotation and brought up to a yoke or — 
bridle which was lifted by a small ram acting upward in the 
eylinder (G), also attached to the trussed lever(F). The cylinder 
_ @ was connected by a small and flexible copper pipe to the pres 
_ sure cylinder of the plungers (J) under the end of the lever. 
