Detection thresholds for movements imposed on the relaxed joints in upper limbs, when expressed in terms of angular or linear displacement, differ from joint to joint. However, when they are expressed in terms of proportional changes in the lengths of fascicles of the muscles serving the joints, they are found to be similar. When the execution of finely graded voluntary movements is analysed, performances of similar accuracy occur when the movements at different joints require alterations of the lengths of active muscle fascicles by similar proportions. These findings suggest that muscle length is a variable of importance to the CNS in both the detection and execution of movements. For faster contractions, another category of movement must be considered. This is the triggered response, which can be voluntarily pre-formulated and stored in the brain, to be released subsequently by some sensory input. Such triggered responses can be demonstrated in experiments in which subjects respond to masked stimuli--low-intensity sensory stimuli which, while readily detected when presented alone, are not detected when followed very soon afterwards by a high-intensity stimulus. Subjects are able to react with simple and more complex movements to low-intensity stimuli whether these are detected (delivered alone) or undetected.