s 229 exercise. Fatigue reduced the total stiffness at any given level of initial torque in proportion to the degree of muscle gain reduction. This indicated that reflexes did not preserve the mechanical properties at the joint in the face of fatigue. Furthermore, the stiffness was also reduced when compared at the same level of initial EMG before and after fatigue, without obvious changes in the reflex EMG response. Thus, stilfness was not uniquely related to the motor neuron pool output and was relatively insensitive to feedback from muscle force receptors. TIBIAL VIBRATIOS TRANSMISSIOS-THE EFFECTS OF INCREMENTAL OSTEOTOMY W. J. MINTOWT-CZYZ, J. A. FAIRCLOUGH, I. G. MACKIE and L. D. M. NOKES (Cardiff Royal Infirmary, Cardiff, U.K.) The human tibia will transmit vibration along its length in response to an impulse at one end. Accelerometers positioned proximally and distally detect the vibration and can measure its magnitude. The ratio of the distal to the proximal measurements yields a measure of the attenuation of the vibration as it passes along the bone. This ‘attenuation factor’ (AF) has been shown to change predictably during fracture union and has been used as a measure of bone healing. This cadaver study was performed to investigate the changes in vibration attenuation consequent upon the making of an osteotomy in stages through the tibia with the intention of discovering what was the influence of the residual intact bone in allowing vibration transmission to occur. Five tibiae were studied. Incremental osteotomies were made with a gigli saw in five stages through each bone and the attenuation factor measured at each stage. The depth of the cut was used as an assessment of the residual intact bone. Standard transmission theory suggests that energy transmission along an object is proportional to its cross sectional area. The results of this study show that the attenuation factor is proportional to the calculated residual area of intact bone, making the assumption that the tibia at the point ofdivision has a triangular cross-section with a triangular section to its medullary cavity. The results imply that measurement of the attenuation factor might be a useful technique for the assessment of the mechanical integrity of bone and support the finding that in life attenuation factor changes in parallel with fracture healing. VARIABLE FORCE COIL SPRING HAND SPLINT J. DANOFF, W. SCHNEIDERWIND, B. MOY, B. THORNTON, L. GERBER, A. RICH and J. LEDERMANN (Department of Rehabilitation Medicine, National Institutes of Health, Bethesda, MD, U?S.A.) Medical conditions such as rheumatoid arthritis can cause severe disfigurement and pain in the joints of the fingers with significant functional loss. In recent years surgical intervention has been applied to regain function of these joints. Hand splints as components of post-operative management have been used to apply low forces to the affected digits in such a manner to ensure that functional alignment will be retained after healing has been completed. Dynamic hand splints allow for digit movement, and forces are usually generated by appropriately positioned elastic bands. Several problems are associated with these splints including non-linearity of forces, deterioration of elastic elements, and inability to perform functional movements. For these reasons a new hand splint was developed using stainless steel linear coil springs to provide variable, but easily adjustable, forces to each of the fingers. This splint can also be adjusted in length, flexion/extension angle, and radial/ulnar orientation. Three units are currently being used on post-surgical patients as needed. ASSESSMENT OF FOOTPAD DEPENDABILITY J. DANOFF, T. WAGGONER and G. HUNT (Department of Rehabilitation Medicine, National Institutes of Health, Bethesda, MD) Commercially manufactured switched footpads are available for monitoring foot-surface contact patterns. These are used in many clinical settings to provide information on gait analysis. Testing procedures are necessary to assure the consistency of these pads. In this study footpads were evaluated by mounting each in a rig under a Chatillon Pressure Gauge. The rig was manipulated to determine switch closing force at nine marked points on each pad. Ten pads were tested twice. Five of these were retested additionally to reach ten trials. Correlated r-test and linear correlations (r) were used to compare the paired data, and coefficients of variation (CV) were