SUMMARY1. In order to investigate the characteristics of synapses formed by a foreign nerve in skeletal muscle of the frog, the hypoglossal nerve in Rana pipiens was implanted under the cutaneus pectoris (c.p.) muscle which was denervated 1 month later. .multiphasic e.p.p.s whereas these were only seen on original nerve stimulation during the first 3 months after regeneration. These abnormalities of the foreign innervation persisted for at least 6 months after crushing the original nerve. 4. The geniohyoid muscle (which is normally innervated by the hypoglossal nerve) and c.p. appear to contain similar types of muscle fibre. Furthermore, when the geniohyoid muscle became reinnervated by the hypoglossal nerve after it had been crushed, the synapses formed were more effective than those formed by the hypoglossal nerve in c.p.5. Even if reinnervation of c.p. by its nerve was delayed by cutting the brachial nerve at its exit from the vertebral column, the synapses formed by the hypoglossal nerve in c.p. remained abnormal, suggesting that an incompatability exists between the c.p. muscle and the hypoglossal nerve.6. In spite of the differences found between the synapses formed by the hypoglossal nerve and the original nerve in c.p., there was no evidence of regression of the hypoglossal innervation during the period of observation.

The effect of muscle damage on nerve growth and ectopic synapse formation in the frog was investigated by bisecting the cutaneus pectoris (c.p.) muscle following implantation of the hypoglossal nerve. Axons from the stump of the implanted nerve grew considerable distances towards the ends of the cut muscle fibres where they formed synapses. In some preparations, c.p. was only partially cut and in these muscles axonal growth from the implanted nerve occurred across intact fibres to reach the cut ends in the damaged area, suggesting the local release of a growth‐stimulating factor. Stimulation of the hypoglossal nerve elicited contraction of the end‐plate‐free half of c.p. The characteristics of these ectopic synapses were similar to those found in a previous study when the hypoglossal nerve innervated the original end‐plates after denervation of c.p. The mean quantal content of the end‐plate potentials (e.p.p.s) was abnormally low, with long latencies between nerve stimulation and the onset of the e.p.p.s and frequent occurrence of multiphasic e.p.p.s in response to single nerve stimulation.

SUMMARYInjection of a sublethal dose of botulinum toxin (type D) into the cutaneus pectoris muscle of the frog caused paralysis for about three months, but in contrast to previous studies in the mammal, did not appear to cause axonal sprouting from motor nerve terminals. In frogs in which the cutaneus pectoris had been denervated by crushing its nerve, reinnervation occurred within 2-3 weeks and axonal sprouts beyond the original end-plates were often observed. When the hypoglossal nerve was implanted into the cutaneus pectoris, crushing the original nerve caused profuse axonal growth from the implanted nerve towards the denervated end-plates within one week, whereas injection of botulinum toxin had little effect. Stimulation of the implanted nerve caused contraction of those cutaneus pectoris muscles whose original nerves had been crushed, but no response to stimulation of the implanted nerve was seen in those muscles in which botulinum toxin had been injected. The failure of botulinum toxin to induce nerve sprouting and acceptance of foreign innervation in the frog may be due to the fact that activity may play a less important role in the neural control of the physiological properties of muscle in this species than in the mammal.