Anomalous innervation of the hand muscles.

According to the usual text-book descriptions the median nerve supplies the abductor polhicis, opponens pohlicis, flexor pollicis brevis and the two lateral lumbricals, while the ulnar nerve supplies all the other muscles in the hand (Gray 1946, Cunningham 1931, Wood Jones 1941). Several articles published since about 1890, notably in the French and the German literature, have mentioned variations found in anatomical dissections. The articles are remarkably contradictory. For example, Spourgitis (1895) stated that ulnar innervation of flexor pollicis brevis was very unusual-he had found it only once in all his dissections ; whereas Riche (1897) considered it to be constant. Brooks (1886) drew attention to the existence of two heads of flexor pollicis brevis and found that their nerve supply varied. In dissections of thirty cadavers, flexor pollicis brevis was supplied by the ulnar nerve alone in five and by the median nerve alone in five ; in nineteen the superficial head was supplied by the median nerve and the deep head by the ulnar nerve ; in one the ulnar nerve supplied all the thenar muscles. Hovelacque reviewed the literature in 1927. He quoted Froment (1847) as stating that the median nerve constantly gave a branch to adductor polhicis ; but Frohse and Frankel (1908) found this in only 10 per cent., a figure with which Poirier (1901) agreed. The first clinical study of the subject seems to be that by Highet (1943). In twenty patients with median nerve division, flexor pollicis brevis was paralysed in only four; in twenty-five with ulnar division only one showed any wasting of flexor pollicis brevis. The obvious explanation of these figures is dual innervation of this muscle. Highet’s patients were the early ones in the Oxford series and are included in my analysis. Murphey, Kirklin and Finlayson (1946) examined 698 ulnar nerve lesions, and found that the first dorsal interosseous was innervated by the median nerve in four patients, in one of whom abductor digiti minimi was also supplied by the median. In all four patients the lesion was above the elbow; in two, blocking of the ulnar nerve at the wrist produced paralysis, suggesting an anastomosis from the median to the ulnar nerve in the forearm. In .551 median nerve injuries, opponens pollicis was found innervated by the ulnar nerve once only. These workers used exceptionally strict criteria for the activity of a given muscle, and in their large series there may have been other anomalous innervations which they did not accept. Clinical material-I have studied the records of all the median and all the ulnar nerve lesions from Mr Seddon’s clinic at both the Wingfield-Morris and the Royal National Orthopaedic Hospitals. There were 688 cases in the group, but many were unsuitable for inclusion in the analysis. The final number accepted was 226, namely 102 median and 124 ulnar nerve injuries. For a case to be accepted there had to be definite evidence of complete interruption of axonal continuity; either the nerve was seen to be divided at operation or, in certain lesions in continuity, direct electrical stimulation of the nerve exposed at operation failed to produce contraction in any muscle. Percutaneous stimulation was used to obtain corroborative but never absolute evidence of the distribution of a nerve. Incomplete lesions and lesions of both nerves were discarded. Throughout the series there has been a high standard of notation, but the statement “ulnar intrinsics paralysed” without any record of more detailed motor examination made it necessary for a few cases to be excluded. The findings in all the accepted cases may be regarded as accurate, for they were made by workers who were fully conversant with the pitfalls of clinical examination of individual muscles and with the trick movements described by Wood Jones (1919). The power of muscles was