SUMMARY1. The membrane potential of isolated muscle fibres in solutions containing tetrodotoxin (TTX) was controlled with a two-electrode voltage clamp. The striation pattern in the region of the electrodes was observed microscopically.2. With square steps of depolarization of increasing magnitude, contraction occurs first in the myofibrils just beneath the surface membrane, and then spreads inwards towards the axis ofthe fibre asthe depolarization is increased.3. From the depolarizations which make the superficial and axial myofibrils contract it is possible to estimate a space constant (AT) for electrotonic spread in a transverse tubular network. 6. Action potentials, recorded from a sartorius fibre, were used as the command signal for the voltage-clamped fibre in tetrodotoxin. The central myofibrils of this fibre did not appear to contract unless the imposed 'action potentials' were of normal size.7. The passive electrical characteristics of the transverse tubular system will just allow an action potential, at room temperature, to activate the myofibrils at the centre of a frog muscle fibre. An active potential change would be required to achieve a safety factor appreciably greater than one for this process.

SUMMARY1. The effect of extracellular calcium and magnesium on the contraction threshold and on the thresholds for an increase in sodium and potassium conductance with depolarization was studied in voltage-clamped frog muscle fibres.2. A larger depolarization was required to reach each of the three thresholds when the concentration of divalent cation was increased.3. The contraction and potassium conductance thresholds appeared to shift in parallel with alterations in calcium over the concentration range 0*2-10O0 mm and in magnesium over the concentration range 5 4-900 mM. The shift amounted to about 4 mV for a threefold change in concentration of divalent cation.4. The sodium conductance threshold was much more sensitive to alterations in divalent cation concentration than was either the contraction or the potassium conductance threshold.

When frog muscle fibers from which the sarcolemma had been dissected away were perfused with a calcium solution and then treated with oxalate, electron-opaque material, probably calcium oxalate, accumulated in the terminal sacs of the sarcoplasmic reticulum. These regions of calcium accumulation were identified with the intracellular calcium sink that controls the relaxation phase of the contraction-relaxation cycle; their proximity to tubules implicated in intracellular stimulus conduction suggests that they might also be regions from which calcium is released to trigger contraction.

The membrane potential of isolated muscle fibers was controlled with a two-electrode voltage clamp, and the radial extent of contraction elicited by depolarizing pulses of increasing magnitude was observed microscopically. Depolarizations of the fiber surface only 1-2 mv greater than the contraction threshold produced shortening throughout the entire crosssection of the muscle fiber. The radial spread of contraction was less effective in fibers exposed to tetrodotoxin or to a bathing medium with a greatly reduced sodium concentration. The results provide evidence that depolarization of a muscle fiber produces an increase in sodium conductance in the T tubule membrane and that the resultant sodium current contributes to the spread of depolarization along the T system.Although it is generally accepted that the T system of twitch muscle fibers transmits the influence of surface depolarization radially, the mode of transmission within the T system has not been established. Huxley and Taylor (1958) found only a graded inward spread of contraction with increasing depolarization of local sites on the surface membrane, while the results of Gonzales-Serratos (1966) on the temperature dependence of the radial spread of activation were more compatible with active propagation within the T system. Recently Adrian, Costantin, and Peachey ' (1969) reported that the radial spread of contraction with controlled surface depolarizations was not entirely consistent with a passive electrotonic spread of depolarization along the T system; they suggested that delayed rectification, an increase in the potassium conductance with depolarization, might be present in the T tubules. The experiments of ACP were performed on fibers exposed to tetrodotoxin (TTX), so that the presence of a regenerative increase in sodium conductance within the T tubules of a normal muscle fiber could not be ruled out.

"Skinned" muscle fibers, single fibers from the frog semitendinosus muscle in which the sarcolemma had been removed, could be reversibly activated by electrical stimulation. Electrical responsiveness was abolished when the skinned fiber was prepared from a muscle exposed to a cardiac glycoside, and the development of responsiveness was delayed when the muscle was bathed in high potassium solution. The findings were taken as evidence that active sodium-potassium exchange across the internal membranes restored electrical excitability, after the sarcolemma had been removed, by establishing a potential gradient across the internal membranes. In general, the contractions were graded with the strength of the applied current. On occasion, however, "all-or-none" type responses were seen, raising the possibility that the internal membranes were capable of an electrically regenerative response. Activation could also be produced by an elevation of the intraceUular chloride ion concentration or a decrease in the intracelhilar potassium ion concentration, suggesting that depolarization of some element of the internal membrane system, that is, a decrease in the potential of the lumen of the internal membrane system relative to the potential of the myofibrillar space, was responsible for activation in these experiments. The distribution of both the electrically induced contractions and those produced by changes in the intracellular ion concentrations indicated that the responsive element of the internal membrane system was electrically continuous over many sarcomeres. Csapo (1959) and Natori and Isojima (1962) have reported that single fibers from a m p h i b i a n muscle could still be activated by electrical stimulation following removal of the surface membrane. It seemed possible that current flow across some component of the internal m e m b r a n e system could account for this p h e n o m e n o n and the experiments described in this paper are an attempt to examine this question. Preliminary accounts of these results have been published (Costantin and Podolsky, 1965, 1966).