The method of direct electrical stimulation of populations of pyramidal neurones by weak surface-anodal shocks (Hern, Landgren, Phillips & Porter, 1962) creates specially favourable conditions for reinvestigating the question whether pyramidal neurones are connected monosynaptically with alpha motoneurones in the spinal cord.The scope of experiment has been deliberately restricted to the 'arm area ' of the cortex and to the alpha motoneurones of the forearm and hand. The arm area was chosen partly because it is more readily exposed than the leg area, and partly because the range and precision of hand movements, even in the baboon, would lead one to expect a powerful cortical command of these final common paths. This expectation is supported by the fact that the cortical electrical threshold for hand movements is lower than the threshold for other movements (Liddell & Phillips, 1950, 1951. To detect minimal synaptic transmission by pyramidal impulses and to measure its quantity and timing, we have made intracellular records from alpha motoneurones innervating the forearm and hand. By thus restricting stimulus and response to their minima, and by attending only to the earliest responses, which are, in fact, the only responses to be recorded at all under these conditions, it is possible to be reasonably sure that one is dealing with pure pyramidal actions, uncomplicated by the so-called extrapyramidal effects that would inevitably be stirred up by stronger and more prolonged electrical stimulation. Experience of these delicate methods has justified our belief that the extra hazards of surgical isolation of the pyramidal pathway by transecting the hind brain, sparing only the medullary pyramids (Lloyd, 1941;Preston & Whitlock, 1960, 1961, can be avoided in these special circumstances.