Romanes published papers demonstrating that the large neurons in the spinal cord that supply individual muscles are clustered together in discrete, nuclei-termed pools that are arranged according to the position of the limb muscles they are programmed to innervate.

In a paper published in 1951, Romanes showed that the pools of motor neurons innervating the muscles that act together to control a limb joint are themselves grouped into larger clusters – thus uncovering a positional registration between a motor neuron and its target muscle.

Romanes reasoned that “all the higher parts of the central nervous system would be organized in a similar, basic way," a premise that is now gathering experimental support.

[4] His contemporaries used his maps to identify the pathognomonic lesions of motoneurones seen in the polio epidemic that was sweeping the United States of America.

Modern workers using transgenic mice have confirmed the genetically-determined location of the motor neuron pools and shown the rest of the spinal cord network, including the connection of the sensory nerves, to be dependent on the positional template described first by Romanes.