which are the "common paths"1 of cortical integration, are the key to the organization of the motor cortex (Msl of Woolsey2). They form the output channel of a complex input-output system, whose input side, as yet, has received less investigation than the output (for a review of recent work, see reference 3). This paper is concerned with one only of the outputs: the corticospinal neurons projecting monosynaptically to motoneurons innervating the baboon's arm, forearm, and hand. These constitute the simplest and most direct of the known corticofugal pathways. In the cat, which has no monosynaptic corticospinal projection,4 there are corticospinal projections controlling the interneurons of reflex arcs5 and the neurons and presynaptic arborizations of ascending spinal pathways,6 and it would be surprising if comparable projections did not exist in the primates. A projection to fusimotor neurons has been found in monkeys by Mortimer and Akert.7 Finally, the corticofugal ax¬ ons from area Msl do not all pass through the medullary pyramid into the spinal cord. Corticorubral and other corticofugal neu¬ rons are intermingled with the corticospinal cells, and the characteristic somatotopic or¬ ganization of the baboon's area Msl is found, by electrical mapping, to survive cut¬ ting of the medullary pyramids.8 The cat has a somatotopically organized corticorubrospinal system9·10; again, it would be sur¬ prising if the primates had lost this, and the motor responses that survive pyramidal section may well depend on it.From this complex of projections from the motor area (Msl of Woolsey2) one can read-ily abstract the monosynaptic corticomotoneuronal system of Bernhard et al11 for ex¬ perimental analysis. Such an abstraction is not merely technically feasible, it is also physiologically significant. Since the ele¬ mentary neural basis of muscular control is the motor unit,12 the spatial organization of the corticospinal cells whose axons converge upon the membranes of single motoneurons is of the highest interest.Electrical mapping of the corticomotoneuronal system is difficult if one uses stimuli which are adequate, in virtue of strength, pulse duration, repetition frequency, or train duration, to evoke actual firing of spi¬ nal motoneurons. The difficulty arises from the fact that arbitrary manipulation of these parameters of stimulation allows one to draw different motor maps on the same brain.13 Lilly et al14 proposed that these maps are compounded, in different propor¬ tions, of direct electrical stimulation of corti¬ cofugal neurons and of stimulation of intracortical afferent systems which converge upon the corticofugal neurons and excite them synaptically. It is therefore necessary to choose a stimulus that will bypass the synaptic complexities of the intact cortex and will selectively excite the corticofugal neurons. One can then send single, small, well-synchronized pyramidal volleys from specific parts of area Msl into the spinal cord, and measure the quantity and timing of monosynaptic excitatory action upon mo...