Electrophysiological Studies of the Antrum Muscle Fibers of the Guinea Pig Stomach

The membrane potentials of single smooth muscle fibers of various regions of the stomach were measured, and do not differ from those measured in intestinal muscle. Spontaneous slow waves with superimposed spikes could be recorded from the longitudinal and circular muscle of the antrum. The development of tension was preceded by spikes but often tension appeared only when the slow waves were generated. Contracture in high K solution developed at a critical membrane potential of -42 mv. MnCl₂ blocked the spike generation, then lowered the amplitude of the slow wave. On the other hand, withdrawal of Na⁺, or addition of atropine and tetrodotoxin inhibited the generation of most of the slow waves but a spike could still be elicited by electrical stimulation. Prostigmine enhanced and prolonged the slow wave; acetylcholine depolarized the membrane without change in the frequency of the slow waves. Chronaxie for the spike generation in the longitudinal muscle of the antrum was 30 msec and conduction velocity was 1.2 cm/sec. The time constant of the foot of the propagated spike was 28 msec. The space constants measured from the longitudinal and circular muscles of the antrum were 1.1 mm and 1.4 mm, respectively.     

The genesis and transmission of epidermal potentials in an amphibian embryo

The genesis and transmission of action potentials in epidermal cells of the newt (Cynops pyrrhogaster) embryo were investigated with special reference to cellular differentiation during development. Typical action potentials can be recorded from any of the epidermal cells at Stage 31. These potentials consist of a fast spike (18 msec) followed by a slow component (164 msec). The potential is graded with current intensity, and only the slow component initiates action potentials in adjacent cells and induces a transmission to other cells. The fast spike was found in all epidermal cells throughout the embryonic stages examined (Stages 26–47). The slow potential, however, appears at Stage 28, persists until Stage $\text{36}\text{37}$ just before hatching and then disappears at Stage $\text{38}\text{42}$. Electrical recordings from traumatic embryos (embryos without neural crest cells) or from cultured epidermal cell masses isolated from the pregastrula or the ventral region of the neurula, were compared with the intact embryo. No differences were observed in either the form of the action potential or its transmission. Thus these action potentials appear to be derived from epidermal cells, and are not of nervous origin. Evidence suggests that the transient establishment of excitable membranes in epidermal cells during differentiation is closely related to neural cell differentiation.     

FUNCTIONAL ANALYSIS OF SWIM-BLADDER MUSCLES ENGAGED IN SOUND PRODUCTION OF THE TOADFISH

A functional analysis of the striated swim-bladder muscles engaged in the sound production of the toadfish has been performed by simultaneous recording of muscle action potentials, mechanical effects, and sound. Experiments with electrical nerve stimulation were made on excised bladder, while decerebrate preparations were used for studies of reflex activation of bladders in situ. The muscle twitch in response to a single maximal nerve volley was found to be very fast. The average contraction time was 5 msec. with a range from 3 to 8 msec., the relaxation being somewhat slower. The analysis of muscle action potentials with surface electrodes showed that the activity of the muscle fibers running transversely to the long axis of the muscle was well synchronized both during artificial and reflex activation. With inserted metal microelectrodes monophasic potentials of 0.4 msec. rise time and 1.2 to 1.5 msec. total duration were recorded. The interval between peak of action potential and onset of contraction was only 0.5 msec. Microphonic recordings of the characteristic sound effect accompanying each contraction showed a high amplitude diphasic deflection during the early part of the contraction. During relaxation a similar but smaller deflection of opposite phase could sometimes be distinguished above the noise level. The output from the microphone was interpreted as a higher order derivative function of the muscle displacement. This interpretation was supported by complementary experiments on muscle sound in mammalian muscle. The dependence of the sound effects on the rate of muscle contraction was demonstrated by changing the temperature of the preparation and, in addition, by a special series of experiments with repeated stimulation at short intervals. Results obtained by varying the pressure within the bladder provided further evidence for the view that the sound initiated in the muscle is reinforced by bladder resonance. Analysis of spontaneous grunts confirmed the finding of a predominant sound frequency of about 100 per second, which was also found in reflexly evoked grunts. During these, muscle action potentials of the same rate as the dominant sound frequency were recorded, the activity being synchronous in the muscles on both sides. Some factors possibly contributing to rapid contraction are discussed.     

Comparison of the duration and power spectral changes of monopolar and bipolar M waves caused by alterations in muscle fibre conduction velocity

The muscle compound action potential (M wave) recorded under monopolar configuration reflects both the propagation of the action potentials along the muscle fibres and their extinction at the tendon. M waves recorded under a bipolar configuration contain less cross talk and noise than monopolar M waves, but they do not contain the entire informative content of the propagating potential. The objective of this study was to compare the effect of changes in muscle fibre conduction velocity (MFCV) on monopolar and bipolar M waves and how this effect depends on the distance between the recording electrodes and tendon. The study was based on a simulation approach and on an experimental investigation of the characteristics of surface M waves evoked in the vastus lateralis during 4-s step-wise isometric contractions in knee extension at 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, and 90% MVC. The peak-to-peak duration (Durpp) and median frequency (Fmedian) of the M waves were calculated. For monopolar M waves, changes in Durpp and Fmedian produced by MFCV depended on the distance from the electrode to the tendon, whereas, for bipolar M waves, changes in Durpp and Fmedian were largely independent of the electrode-to-tendon distance. When the distance between the detection point and tendon lay between approximately 15 and 40 mm, changes in Durpp of bipolar M waves were more pronounced than those of distal monopolar M waves but less marked than those of proximal monopolar M waves, and the opposite occurred for Fmedian. Since, for bipolar M waves, changes in duration and power spectral features produced by alterations in MFCV are not influenced by the electrode-to-tendon distance, the bipolar electrode configuration is a preferable choice over monopolar arrangements to estimate changes in conduction velocity.     

Inhibition in branched afferent neurons of the bullfrog tongue

Summary Single fibers of the bullfrog glossopharyngeal nerve give rise to several peripheral branches, each innervating separate fungiform papillae on the dorsal surface of the tongue. Extracellular electrodes were used to stimulate and record simultaneously from several papillae and from the central branch.Minor changes in centrally recorded neural output were caused by collision of action potentials originating in separate branches of a common fiber.Following an antidromic or orthodromic action potential in any branch, a series of excitability changes occured in that branch. Normal excitability was regained within 5 msec of an action potential, but was followed by a secondary decrease in excitability, which reached a minimum approximately 50 msec after the spike, and returned to normal within 200–400 msec after the spike. Subthreshold stimuli caused no depression, while doubling the stimulus strength above threshold did not enhance depression. After several spikes, both amplitude and duration of depression increased. Depression could be evoked even after the gustatory receptors were surgically removed.Post-stimulus depression in fiber branches is suggested as one source of gustatory adaptation, and may also contribute to interference between stimulating substances.The authors are particularly grateful for assistance and advice from Dr. Douglas Junge, of the School of Dentistry and Department of Physiology at the University of California, Los Angeles. The reported work was supported by NIDR Contract No. 69-2227 to Dr. Junge, and was carried out while one of the authors (JAM) held a PHS postdoctoral traineeship with the Department of Zoology, U.C.L.A., and the other (MSB) held a NIH predoctoral traineeship with the Department of Anatomy, U.C.L.A. Draughts of the paper have been read and criticized by Dr. Junge and Dr. J. P. Leader, of Auckland University.     

Auditory brain-stem responses evoked by electrical stimulation of the cochlear nucleus in human subjects

When auditory nerve function is lost due to surgical removal of bilateral acoustic tumors, a sense of hearing may be restored by means of an auditory brain-stem implant (ABI), which electrically stimulates the auditory pathway at the level of the cochlear nucleus. Placement of the stimulating electrodes during surgical implantation may be aided by electrically evoked auditory brain-stem responses (EABRs) recorded intra-operatively. To establish preliminary standards for human EABRs evoked by electrical stimulation of the cochlear nucleus, short-latency evoked potentials were recorded from 6 ABI patients who were either already implanted or undergoing implantation surgery. Neural responses were distinguished from stimulus artifact and equipment artifact by their properties during stimulus polarity reversal and amplitude variation. Other properties contributed to further identification of the evoked potentials as auditory responses (EABRs). The response waveforms generally had 2 or 3 waves. The peak latencies of these waves (approximately 0.3, 1.3, and 2.2 msec) and the brain-stem localization of the region from which they could be elicited are consistent with auditory brain-stem origin.     

Electrophysiology of the isolated central nervous system of the northern octopus Eledone cirrhosa

A successful preparation has been devised for maintaining the octopus brain in a viable condition to allow microelectrode studies of individual nerve cells. Impalements of cells within the sub‐oesophageal mass reveal that three populations of neurones are present These have different resting potentials, ranging from approximately 60 mV down to under 30 mV. Spontaneous activity is recorded from many neurones but some are silent and others exhibit only synaptic noise. Electrical stimulation of silent cells may lead to no response (large resting potential cells) or provoke trains of impulses (30–45 mV cells). Typical action potentials have durations of 20 msec. IPSP and EPSP activity may be observed. Burster cells or oscillators are located in one specific region, and a variety of activity may be recorded. These periodic bursts may be modified by hyperpolarisation so that spiking ceases but the underlying oscillatory potential remains. Some units exhibit two spike sizes, often uncorrelated in discharge.     

Spinal tracts producing slow components of spinal cord potentials evoked by descending volleys in man

Slow negative (N) and slow positive (P) waves are frequently produced in the posterior epidural space at the lumbosacral enlargement by epidural stimulation of the rostral part of human spinal cord. The production of these slow potentials are thought to be responsible for analgesia at the stimulated segment as well as below that level. In order to define the spinal tract which mediates these slow potentials, we stimulated directly or from the epidural space the dorsal, dorsolateral, lateral and ventral columns at the cervical or thoracic level, and epidurally recorded spinal cord potentials (des.SCPs) at the lumbosacral enlargement in 7 patients who underwent spine or spinal cord surgery. The des.SCPs recorded in the lumbosacral enlargement consisted of polyphasic spike potentials followed by slow N and P waves. At a near threshold level of stimulus intensity the slow N and P potentials were consistently elicited only by stimulation of the dorsal column. The slow waves were also produced by intense stimulation of other tracts, but remained significantly (P < 0.05−P <0.01) smaller than those evoked by dorsal column stimulation when compared at the same stimulus intensity. Moreover, the slow P wave could not be elicited even by intense stimulation (10 times the threshold strength for the initial spike potentials) of the ventral column. Thus, the results suggest that the slow N and P waves are mostly mediated by the antidromic impulses descending through the dorsal column.     

Diameter changes of muscle fiber types of the inferior oblique muscle of the rabbit following denervation]

Changes in fibre diameters of extraocular muscles of the rabbit were studied at different times after denervation. The whole inferior oblique muscle hypertrophied, while some of the muscle fibres hypertrophied and others showed atrophy, depending on the fibre type. Fibre types have been determined by their histochemical enzyme profile. In the central layer of the muscle the phasic muscle fibres, which are rich in mitochondria, exhibited a transient hypertrophy being maximal 4-5 weeks after denervation and afterwards they atrophied; other phasic muscle fibres, which are poor in mitochondria, atrophied without having shown any sign of hypertrophy. Special, putatively slow tonic muscle fibres, which have low enzyme activities, underwent small long-lasting increases of their diameters. In the superficial layer of extraocular muscle there are two types of extremely thin muscle fibres rich in mitochondira. Both these fibre types hypertrophied to the greatest degree and for a very long time. Comparable changes in fibre diameters as described here for the muscle fibre types of an extraocular muscle are known from special muscle fibres in other vertebrate     

The properties of the extraocular muscles of the frog. II. Pharmacological properties of the isolated superior oblique and superior rectus muscles.

The pharmacological properties of the superior oblique and the superior rectus muscles of the frog's eye were investigated in comparison with those of a skeletal muscle (iliofibularis muscle) of the same animal. Acetylcholine causes sustained contractures of the extraocular muscles; this effect is increased by physostigmine and decreased or abolished by d-tubocurarine. Also the applications of succinylcholine, choline or caffeine are able to evoke contractures. There are no striking differences in pharmacological properties between extraocular and skeletal muscles of the frog. The time-course of the contractures and the sensitivity of the muscle preparations to the drugs which evoke contractures are identical in extraocular and iliofibularis muscles. In comparison with skeletal muscles there is no higher sensitivity of the extraocular muscles against curare-like drugs. The existence of adrenergic receptors could not be found neither in extraocular nor in skeletal muscles of the frog. It is concluded that in frogs no pharmacological differences exist between the muscle fibre types which compose the extraocular and the skeletal muscles.     

Participation of nonsegmentary neurons of the spinal cord in execution of presynaptic control

Experiments were carried out on cats six days after complete transection of the spinal cord. Cord dorsum potentials (CDP) were recorded in the vicinity of the third lumbar segment during stimulation of the isolated dorsolateral funiculus (DLF). The CDP consist of a rapid monophasic potential (which apparently reflects antidromic excitation of the cells of Clarke's column) and two subsequent slow negative waves, which are replaced by a long positive oscillation. In form, time characteristics, and behavior during thythmic stimulation, this potential differs considerably from the CDP recorded during stimulation of the afferent nerves. The presence of a positive phase of the CDP indicates that stimulation of the DLF evokes primary afferent depolarization (PAD). Stimulation of the DLF causes inhibition of the CDP evoked by stimulation of the afferent nerve. The time course of this inhibition corresponds to the time course of presynaptic inhibition. It is demonstrated that stimulation of the afferent nerve (n. femoralis) inhibits slow components of the CDP evoked by stimulation of the DLF. This inhibition reaches a maximum at the 16th millisecond; its duration exceeds 300 msec. Stronger and more prolonged inhibition of the same components is observed when both the conditioning and the testing stimuli are administered to the DLF. Since primary afferents do not take part in CDP emergence during stimulation of the DLF, it may be hypothesized that the observed inhibition develops as a result of depolarization of interneuron axon terminals.Dnepropetrovsk State University. Translated from Neirofiziologiya, Vol. 2, No. 5, pp. 520–527, September–October, 1970.     

Transmural recording of monophasic action potentials

To investigate the possibility of transmural recording of repolarization through the ventricular wall, KCl monophasic action potential (MAP) electrodes positioned along plunge needles were developed and tested. The MAP electrode consists of a silver wire surrounded by agarose gel containing KCl, which slowly eluted into the adjacent tissue to depolarize it. In six dogs, a plunge needle containing three KCl MAP electrodes was inserted into the left ventricle to simultaneously record from the subepicardium, midwall, and subendocardium. In six pigs, eight plunge needles containing three KCl MAP electrodes and two plunge needles containing similar electrodes except for the absence of KCl were inserted into the ventricles. In three guinea pig papillary muscles, a KCl electrode was used to record MAPs along with two microelectrodes for recording transmembrane potentials. Transmural MAP recordings could be made for >1 h in dogs and >2 h in pigs with a significant decrease in MAP amplitude over time but without a significant change in MAP duration. With the electrodes without KCl in pigs, the injury potentials subsided in <30 min. When the pacing rate was changed to alter the action potential duration and refractory period in dogs, the MAP duration correlated with the local effective refractory period (r = 0.94). The time course of the MAP duration recorded with a KCl MAP electrode in guinea pig papillary muscles corresponded well with that of the transmembrane potential recorded with an adjacent microelectrode. It is possible to record transmural repolarization of the ventricles with KCl MAP electrodes on plunge needles. The MAP is caused by the KCl rather than being a nonspecific injury potential.     

The properties of the extraocular muscles of the frog. I. Mechanical properties of the isolated superior oblique and superior rectus muscles.

The mechanical properties of two extraocular muscles (superior oblique and superior rectus muscles) of the frog were studied and compared with those of a frog's skeletal muscle (iliofibularis muscle) which contains the same types of muscle fibres as the oculorotatory muscles. The extraocular muscles are very fast twitching muscles. They exhibit a smaller contraction time, a smaller half-relaxation time, a higher fusion frequency, and a lower twitch-tetanus ratio than the skeletal muscles. The maximum isometric tetanic tension produced per unit cross-sectional area is lower in the extraocular muscles than in skeletal muscles. However, the extraocular muscles show a higher fatigue resistance than the skeletal muscles. With respect to the dynamic properties there are some differences between the various oculorotatory muscles of the frog. The superior rectus muscle exhibits a faster time-course of the contraction, a higher fusion frequency, and a higher fatigability than the superior oblique muscle. An increase of the extracellular K+-concentration evokes sustained contractures not only in the extraocular muscles but also in the iliofibularis muscle; between these muscles there are no striking differences in the mechanical threshold of the whole muscle preparation. The mechanical threshold depends on the Ca++-concentration of the bathing solution and it is found in a range between 12.5 and 17.5 mM K+ in a normal Ringer solution containing 1.8 mM Ca++. The static-mechanical properties of the extraocular muscles of the frog and the dependence of the active developed tension on the muscle extension are very similar to those which are known to exist in the extraocular muscles of other vertebrates. In tetanic activated frog's oculorotatory muscles a linear relationship exists between length and tension. A variation of the stimulation frequency does not change the slope of this curve but causes parallel shifts of the curve. The peculiar properties of the extraocular muscles of the frog are discussed with respect to the muscle fibre types in these muscles and to the diameter of the muscle fibres.     

Focal evoked potentials in the rabbit visual cortex: Current source density analysis

Focal evoked potentials arising in the rabbit visual cortex in response to photic stimulation from a point source were analyzed by determination of the current source density. The response to a point stimulus arises in a circumscribed area of cortex, corresponding retinotopically to the stimulated point of the visual field and it consists of two components. The first component is created by a local current sink at a depth of 0.6 to 1.0 mm (the level of layer IV) and has a latent period of 30 msec and a peak time of 50 msec. The second component is created by a more diffuse current sink at a depth of 0.2–0.3 to 1.3–1.5 mm (levels between layers III and VI); the time to the maximum was 90–100 msec. These local sinks are regarded as active, created by depolarizing synapses. Passive current sources are concentrated around zones of active sinks. The two components of the response may reflect two consecutive waves of activation of cortical neurons.A. N. Severtsov Institute of Evolutionary Morphology and Ecology of Animals, Moscow. Translated from Neirofiziologiya, Vol. 13, No. 5, pp. 474–481, September–October, 1981.     

The N-Shaped Current-Potential Characteristic in Frog Skin: II. Kinetic Behavior during Ramp Voltage Clamp

Previous step voltage-clamp measurements on frog skin showed the presence of an N-shaped current-potential (I-V) relation in excitable skin. However, the collection and reconstruction of I-V data using discrete step changes of skin potential was tedious because of the long refractory period (up to 1 min) in frog skin. A direct and rapid (5 msec) method for recording the N-shaped I-V characteristic in real time is presented. Ramp functions are used as the command to the clamp system instead of a step function. Consequently the skin potential is forced to change in a linear manner (as commanded) and the skin current can be recorded as a continuous function of the controlled change of skin potential. With the ramp clamp, a low-resistance membrane state ( 10 Omega . cm(2)) resembling a breakdown phenomenon was observed at high skin potential ( 300 mv). Entry into the low resistance state resulted in a collapse of the N-shaped I-V relation to a nearly linear function. The utility of the ramp measurement is demonstrated by predicting (1) that the maximum rate of rise of the spike occurs at a voltage corresponding to the valley (local minimum) in the N-shaped I-V curve, (2) that the rate of rise of the spike increases with increasing clamp currents, (3) the voltage peak of the spike, and (4) the time course of the rising phase of the spike.     

Auditory evoked potentials from the primary auditory cortex of the cat: topographic and pharmacological studies

Wave VI (8.4 msec) of the brain-stem auditory evoked potential (BAEP) was maximal in a discrete region of primary auditory cortex (AI) of the anesthetized cat. Wave VI underwent rapid amplitude decreas over millimeter distances in the AI region and followed high stimulation rates. Wave VI did not show intracortical polarity inversion nor was it abolished by epicortical or intracortical GABA administration. The data are compatible with a wave VI source in the terminal axons of the thalamo-cortical radiations.Middle latency auditory responses (MAEPs) generated 10–40 msec after auditory stimulation were also recorded in a circumscribed area of AI. In contrast to wave VI, these primary auditory cortex potentials (Pa 18.3 msec; Nb 31.9 msec) underwent transcortical polarity inversion, correlated with intracortical multi-unit activity in the AI region and were reversibly altered or abolished by epicortical or intracortical GABA adminstration to the AI region. The data suggest that the Pa and Nb components of the cat MAEP are intracortically generated by neuronal elements in the AI region.     

Analysis of the formation of evoked potentials in neurons of the rat olfactory cortex

Evoked focal potentials which were induced in vitro in a slice of olfactory tract by stimulation of the lateral olfactory tract (LOT) have been studied. The potential consisted of an initial biphasic wave, the compound action potential of LOT, population synaptic responses, and population spike. Functional significance and possible mechanisms of changes of different focal potential waves have been discussed.     

Electrical properties of developing rat heart. Effects of dexamethasone

Action potentials recorded from perinatal rat ventricles exhibited a plateau (phase 2), followed by a rapid repolarization characteristics of all mammalian ventricular cells. Within the second postnatal week, a number of distinct changes occurred in the contour of action potentials. An early slow depolarization, at the foot of the action potential, preceded the beginning of phase zero. The early slow depolarization was observed until day 12 and disappeared by day 13. A second slow depolarization occurred during the terminal phase of the rapid upstroke of the action potential, persisted through day 13 and disappeared by day 14. On day 12, what had been a homogeneous contour of action potentials seen during the first week converted into a heterogeneous contour. Occasionally, action potentials similar to those recorded from Purkinje fibres in adult heart were recorded from hearts as young as 12 days. By day 14, signs of a spike (the hallmark of action potentials from adult heart) were apparent in some fibres. Treatment of newborn rats with dexamethasone on the second day after birth prevented the disappearance of the second slow depolarization. In adult and aged rat hearts, dexamethasone treatment induced a slow depolarization and a plateau in the region of overshoot. In view of the time-dependent change of the second slow depolarization it is suggested that this phase of the action potential is influenced by the levels of circulating glucocorticoid in developing heart and by changes in calcium sensitivity observed in this species. Heterogeneity of action potentials observed on day 12 postnatal may precede structural differentiation of myofilaments.     

Extracellular potential field of excited isolated frog muscle fibres immersed in a volume conductor

The extracellular potential field of isolated frog muscle fibres immersed in a volume conductor was studied at radial distances up to 3 mm during excitation. The shape of the field distant from both the point of the origin of the excitation and the end of the fibre as well as changes in the field when depolarization wave approached the fibre end were described. Different amplitude decrease rates in individual phases of the extracellular potential and the peak-to-peak amplitude at different temperatures were found. Extracellular potentials at long radial distances were recorded using an averaging technique. The shape of the extracellular potentials at long radial distances over the fibre and beyond its end were very similar to the shape of extraterritorial potentials of a single motor unit.     

The role of delayed excitation in the co-ordination of some metathoracic flight motoneurons of a locust

Summary Intracellular recordings have been made from the somata of two metathoracic flight motoneurons, one innervating an elevator muscle of the hindwing, the tergosternal muscle 113 and the other a depressor, the first basalar muscle 127. The locust,Ghortoicetes terminifera was mounted ventral side uppermost with the thorax restrained and opened for access to the thoracic ganglia. Patterns of electrical activity recorded from the thoracic muscles were similar to those shown by a locust during flight when tethered in a more normal posture. In flight the left and right 113 motoneurons each receive a single impulse together at every stroke of the wing, with the 127 muscles active in approximate antiphase. A spike in a 113 motoneuron causes a delayed wave of excitation simultaneously upon itself and its contralateral partner (Fig. 2). The epsp's which form these waves summate and may cause a spike which follows the original one with a delay equal to the wingbeat period. The delayed excitation of the contralateral motoneuron is of larger amplitude than the ipsilateral one so that spikes in either motoneuron must activate separate but symmetrical pathways. A single spike may cause multiple waves in either motoneuron, each separated by intervals equal to the wingbeat period (Fig. 3). In the pathway must be neurons capable of reverberation.A spike in a 113 motoneuron causes a delayed excitation of the ipsilateral 127 motoneuron so that its membrane potential is lowered antiphasically to that of 113 (Fig. 17). A spike in a 127 motoneuron has no effect on the 113 motoneurons. In flight these pathways causing delayed excitation may co-ordinate the motoneurons.The left and right 113 motoneurons receive common synaptic inputs from at least two sources (Fig. 8). These occur as bursts of epsp's at intervals approximately equal to or multiples of the wingbeat period and in the absence of flight. Epsp's of sufficient amplitude cause a spike in the motoneuron which is in the correct phase in the flight pattern relative to any other active motoneurons (Fig. 9). During sustained flight epsp's contribute to the wave of depolarization that the motoneuron undergoes at each wingbeat (Fig. 11). In the absence of the epsp's the motoneuron does not oscillate on its own. At the end of flight bursts of epsp's may continue at the flight frequency long after all activity in the muscles has ceased.Beit Memorial Research Fellow.