There is now much evidence that the mechanical excitation of certain sensory receptors sets up a potential change in the receptor and that this potential change initiates impulses. The microphonic potential of the ear has been known for a long time, but recently its relation to impulse initiation has been clarified by Davis, Tasaki & Goldstein (1952). This situation, however, differs from that at touch and stretch receptors, because the microphonic potential is generated outside the sensory axon, while it is probable that the potential changes that follow mechanical stimulation in peripheral receptors are generated across the axonal membrane. Gray & Malcolm (1950) showed that during the interval between a mechanical pulse and the initiation of an impulse at a Pacinian corpuscle the excitability of the receptor rose steadily, and that if no impulse materialized the excitability then fell again; it was concluded that the time course of this excitability change represented the time course of a potential change at the axon terminal. Katz (1950) demonstrated a potential change occurring in muscle spindles during stretch, which he showed was responsible for the initiation of impulses and called the 'spindle potential'. Recently, Alvarez-Buylla & Ramirez de Arellano (1953) observed a potential in Pacinian corpuscles, which is clearly the immediate precursor of the impulse.These potential changes in receptors are the earliest known signs of activity in the process of impulse initiation and further knowledge of the fundamental properties of receptor mechanisms demands a greater knowledge of their properties. These experiments were carried out on the Pacinian corpuscle for a number of reasons; the most important being that it seemed practicable to prepare a single receptor in such a way that records of the current flow along the last internode could be obtained. Such a technique can clearly distinguish potentials generated in the short (c. 0 5 mm) non-myelinated terminal from those in the rest of the axon. It is also an advantage to have a stable preparation. Pacinian corpuscles, although extremely sensitive, are not spontaneously * British Council Scholar.

No abstract

No abstract

Peripheral effects of substances acting like nicotine have already been demonstrated with certainty at two sites, the receptive area of the ganglion cell and the motor end-plate. Evidence suggesting that nicotine-like substances act at sensory nerve endings is presented in this paper. Sensory endings resemble the other sites at which nicotine is known to act in that they are places at which propagated impulses are initiated. The possibility of such nicotine-like action was suggested by Coon & Rothman's (1940) description of an axon reflex initiated by acetylcholine. They showed that intradermal-injections of acetylcholine caused an erection of hairs which was not confined to the site ofinjection and was not abolished by atropine. It disappeared after degenerative section of the post-ganglionic sympathetic supply to the area examined.Our experiments have shown that the close arterial injection of acetylcholine or nicotine into skin and mesentery causes a discharge of impulses in the sensory nerves, and evidence is given that this may be a direct effect upon the endings exposed to the drugs.A preliminary note on the experiments described here has already been published (Gray, 1947 a). METHODSThe experiments have been carried out on dogs and cats anaesthetized with nembutal or chloralose and also on cats decerebrated under ether. Preparations of skin and mesentery were used. Skin preparation. Either the skin over the external aspect of the lower part of the thigh or that over the internal aspect of the upper part of the thigh was used. The main cutaneous vessels of the region were dissected and a convenient small branch, supplying a circumscribed area, was chosen. The nerve associated with the smaller vessel was freed from connective tissue and divided for subsequent recording from its peripheral end. A fine hypodermic needle without bevel was then inserted, in a central direction, and tied into the main artery peripheral to the origin of the side branch. The main artery, central to the side branch, was cleaned to receive a bull-dog clamp, and all other branches between the cannula and the eventual site of the bull-dog clamp were double-tied and divided. The nerve alongside the branch artery was then placed on fine brush electrodes, which were moistened with Locke's solution and bound lightly, almost to the tips, with a fine silver wire. In some preparations the whole area of skin supplied by the small artery and nerve was dissected free, except for a pedicle consisting of the artery, vein and nerve (Fig. 1).

Vital staining with Janus Green, phase contrast and scanning electron microscopy were used to map the distribution of free neuromast organs from first hatching, 10 mm long larvae to 100 mm long juveniles of herring (Clupea harengus L.), with some further observations on juvenile sprat (Sprattus sprattus (L.)). Neuromasts are sparsely distributed on the head and trunk at hatching but soon proliferate on the trunk where, by a length of 13–15 mm, they occur one to every segment. Near metamorphosis there are at least three rows of neuromasts on the anterior trunk region, 6–9 single neuromasts on the caudal fin and as many as 50 on the head. The scales develop at about 40–50 mm and the neuromasts are then found singly or in groups of 2 or 3 on the surface of the scales of the anterior trunk.The lateral line develops at 22–24 mm and appears to incorporate existing free neuromasts on the side of the head. Unlike the cupulae of the free neuromasts, which are cylindrical, the lateral-line cupulae are thin erect plates lying along the axis of the canals. They are probably continually growing and being shed, followed by renewed growth.All neuromasts contain hair cells of opposing polarities; most free neuromasts are arranged with these polarities arranged fore-and-aft, but a few are dorsoventral.