The co-existence of goitre and functional spasm of the neck suggested to Pauly[204] that pressure on the recurrent laryngeal nerve might occasion a reflex spasm via the muscular branch of the spinal accessory. By analogy, in some cases of spasmodic torticollis a point of irritation on one of the sensory nerves of the cervical plexus might generate a reflex motor reaction in the area of the accessory, with possible diffusion to neighbouring trunks.[205] It might then be a good plan to divide the branches of the superficial cervical plexus, just as the trigeminal is divided for tic douloureux of the face. It soon became obvious that resection of the spinal accessory was insufficient. Risien Russell[206] adduced physiological evidence to show that some of the muscular groups involved in the condition are not innervated by the spinal accessory, but by the second, third, and fourth cervical roots, section of which is imperative to obtain positive results. The surgeon had not been behindhand, however. Gardner in 1888 was convinced of the necessity of dealing with the posterior branches of the second and third cervical pairs, a method practised a few months later by Smith and by Keen. One or two cases recorded by Ballance, according to whom division of the posterior roots was performed as far back as 1882 or 1883, are highly instructive: A woman, thirty-two years old, had suffered for seventeen months from convulsive movements inclining the head to the right shoulder and turning the face to the left, the muscles affected being the sternomastoids, right trapezius, and complexus. On May 30, 1887, half an inch of the left spinal accessory was resected before its entry into the muscle, whereupon the spasm diminished in intensity and the sternomastoids ceased to contract. On June 6 two-thirds of an inch of the right accessory was removed, the patient being able four days later to keep her head straight by the application of her hand to the right side; but on July 4 violent spasms of the trapezius recommenced, demanding section of the posterior branch of the second pair. By the 21st there was a little stiffness of the neck on the right which speedily disappeared, and in March, 1891, recovery was still complete.