Search

The high concentration in brain of unbound l-glutamic acid (in its anionic form, l-glutamate) fuelled considerable speculation as to its role in central nervous function more than 50 years ago. Claims in the 1940s that it could improve cognitive acuity in patients with mental impairment were particularly intriguing, though later refuted. In the early 1950s Hayashi [(1954) Keio J. Med. 3, 183–192] found that l-glutamate could cause convulsions and proposed that it might be a central synaptic transmitter. Soon thereafter, Curtis and colleagues [Curtis, Phillis and Watkins (1959) Nature (London) 183, 611] showed that l-glutamate depolarized and excited central neurons, as expected for an excitatory transmitter; however, various aspects of the action of l-glutamate seemed to argue strongly against a transmitter function. This negative view prevailed for some 20 years, before compelling evidence for such a role was adduced. Over the last two decades, extensive research has revealed a host of glutamate receptor subtypes, subserving several different functions in excitatory synaptic transmission. This paper gives a very brief and personal overview of the development of the field over the last 50 years from a mainly pharmacological standpoint.

Troponeiron tricarbonyl, 2-methyltroponeiron tricarbonyl, and 4-methyltroponeiron tricarbonyl are converted to fluxional tropyliumiron tricarbonyl salts by treatment with either trimethylsilyl or di-n-butylboron triflate. These pentadienyliron cations react with (η 1 -allyl)-, (η 1 -propargyl)-, and (η 1 -allenyl)Fp complexes [Fp=C 5 H 5 Fe(CO) 2 ] to give ketohydroazulene cycloadducts derived, in each reaction, from a single tautomeric tropyllium cation. The further reaction of two such cycloadducts with lithium dimethyl cuprate gave tricyclic ketones, through a sequence depicted as involving initial formation of an anionic acyliron complex, followed by migratory insertion and an intramolecular aldol condensation

This report describes four sets of tests sponsored by the U.S. Nuclear Regulatory Commission and conducted by the Idaho National Engineering and Environmental Laboratory. The tests support research addressing the need to provide assurance that motor-operated valves are able to perform their intended safety function, usually to open or close against specified (design basis) flow and pressure loads. One of the parameters that affects a gate valve's operability is the friction between the disc seats and the valve body seats. In most gate valves, these surfaces are hardfaced with Stellite 6, a cobalt-based alloy. The tests described in this report investigate the changes that occur in the friction as the Stellite 6 surfaces develop an oxide film as they age. Stellite 6 specimens were aged in a corrosion autoclave, the oxide films were examined and characterized, and the specimens were subjected to friction testing in a friction autoclave. A very thin oxide film formed after only a fe w days of natural aging. Even a very thin oxide film caused an increase in friction. The surface structure of the oxide film was dominated by a hard crystalline structure, such that the friction response was analogous to rubbing two pieces of sandpaper together. In the limited data provided by naturally aged specimens (78 days maximum exposure, very thin oxide films), the friction increased with greater aging time, approaching an as-yet-undetermined plateau. Although the thickness of the oxide film increased with greater aging time, the mechanical properties of the oxide film (larger granules with greater aging time) appeared to play a greater role in the friction response. Friction testing of specimens subjected to simulated in-service testing strokes at intervals during the aging process showed only a slight decrease in friction, compared to other specimens. Results from specimens subjected to accelerated aging were inconclusive, because of differences in the structure and comp osition of the oxide films, compared to naturally aged specimens. For the naturally aged specimens, the highest friction occurred on the first stroke. The first stroke smeared the oxide film and dislodged some of the granules, so that subsequent strokes saw lower friction values and less variation in the friction. This result underscores the importance of planning in-plant tests so that data are collected from the first stroke following a period of inactivity.

For the past several years, researchers at the Idaho National Engineering and Environmental Laboratory, under the sponsorship of the US Nuclear Regulatory Commission, Office of Nuclear Regulatory Research, have been investigating the performance of motor-operated valves subjected to design basis flow and pressure loads. Part of this research addresses the friction that occurs at the interface between the valve disc and the valve body seats during operation of a gate valve. In most gate valves, these surfaces are hardfaced with Stellite 6, a cobalt-based alloy. Analytical methods exist for predicting the thrust needed to operate these valves at specific pressure conditions. To produce accurate valve thrust predictions, the analyst must have a reasonably accurate, though conservative, estimate of the coefficient of friction at the disc-to-seat interface. One of the questions that remains to be answered is whether, and to what extent, aging of the disc and seat surfaces effects the disc-to-seat coefficient of friction. Specifically, does the environment in a nuclear plants piping system cause the accumulation of an oxide film on these surfaces that increases the coefficient of friction; and if so, how great is the increase? This paper presents results of specimen tests addressing this issue, with emphasis on the following: (1) the characteristics and thickness of the oxide film that develops on Stellite 6 as it ages; (2) the change in the friction coefficient of Stellite 6 as it ages, including the question of whether the friction coefficient eventually reaches a plateau; and (3) the effect in-service cycling has on the characteristics and thickness of the oxide film and on the friction coefficient.

The US Nuclear Regulatory Commission (NRC), Office of Nuclear Regulatory Research, is funding the Idaho National Engineering and Environmental Laboratory (INEEL) in performing research investigating the performance of gate valves subjected to pressure locking and thermal binding conditions. Pressure locking and thermal binding are phenomena that make a closed gate valve difficult to open. Pressure locking can occur when operating sequences or temperature changes cause the pressure of the fluid in the bonnet (and, in most gate valves, between the discs) to be higher than the pressure on the upstream and downstream sides of the disc assembly. Thermal binding can occur when thermal expansion/contraction effects cause the disc to be squeezed between the valve body seats. If the loads associated with pressure locking or thermal binding are very high, the actuator might not have the capacity to open the valve. The authors tested a flexible-wedge gate valve and a double-disc gate valve under pressure locking and thermal binding conditions. The results show that these valves are susceptible to pressure locking; however, they are not significantly affected by thermal binding. For the flexible-wedge gate valve, pressure locking loads (in terms of stem thrust) were higher than corresponding hydrostatic opening loads by a factor of 1.1 to 1.5. For the parallel disc gate valve, pressure locking loads were higher by a factor of 2.05 to 2.4. The results also show that seat leakage affects the bonnet pressurization rate when the valve is subjected to thermally induced pressure locking conditions.

We describe experimental results of cw laser operation using a red (670 nm) laser diode pumping two near4nfrared dyes(Rd-800 and Rd.700). Red diode lasers operating with 300 mW cw power were characterized, paying particular attention tobeam quality. Collimating optics and a highly astigmatic focusing system were designed and demonstrated, producing an approximately 50 pun diameter focal spot, resulting in over 15 kW/cm2 pump intensity (jer diode). The resulting diode/ optical system was used to pump a commercial Coherent Model 599 cw dye laser utilizing a flowing dye jet. 1. INTRODUCTION Tunable dye lasers have been one of the most important classes of lasers for more than two decades. They have manyuses that span the range from basic research tools to an emerging form of cancer treatment known as photodynamic thempy.For many applications, continuous wave (cw) lasers have significant advantages. However, the excitation sources for these devices are large and inefficient gas phase ion lasers, for example argon or krypton ion. Typically, these ion lasers are

Important findings in the excitatory amino acid (EAA) field that have stemmed directly from work initiated in chemical laboratories are discussed from a historical point of view, showing how each has contributed to our present knowledge of EAA receptors. The groups of compounds discussed include simple analogues and derivatives of short-chain excitatory amino acids, longer-chain analogues, analogues containing ring structures essential to the acidic nature of the ω-terminal, and conformationally restricted agonists and antagonists. Recent interest has centred on antagonists for the N-methyl-D-aspartate (NMDA) and α-methyl-3-hydroxy-5-methyl-isoxazole-4-propionic acid (AMPA) receptors typified by the long-chain ω-phosphono amino acids and the quinoxalinediones, respectively. Consideration of conformational aspects of both agonists and antagonists is currently providing considerable insight into the disposition of charged sites and general topological features in the various receptors, as well as the likely conformation adopted by L-glutamate in its physiological interaction with these receptors.Key words: excitatory amino acid receptors, historical development, structure–activity relations, NMDA receptors, AMPA receptors, kainate receptors.

Troponeiron tricarbonyl, 2-methyltroponeiron tricarbonyl, and 4-methyltroponeiron tricarbonyl are converted to fluxional tropyliumiron tricarbonyl salts by treatment with either trimethylsilyl or di-n-butylboron triflate. These pentadienyliron cations react with (η 1 -allyl)-, (η 1 -propargyl)-, and (η 1 -allenyl)Fp complexes [Fp=C 5 H 5 Fe(CO) 2 ] to give ketohydroazulene cycloadducts derived, in each reaction, from a single tautomeric tropyllium cation. The further reaction of two such cycloadducts with lithium dimethyl cuprate gave tricyclic ketones, through a sequence depicted as involving initial formation of an anionic acyliron complex, followed by migratory insertion and an intramolecular aldol condensation

1) The KD values for a range of antagonists including DGG, pCB-PzDA, pBB-PzDA, HDC-QXCA, DNQX and CNQX have been determined using a series of non-NMDA receptor agonists in the isolated spinal cord of the neonatal rat. 2) CNQX and DNQX, and, to a lesser extent, DCH-QXCA, were by far the most potent antagonists, although the degree of selectivity did not vary much throughout the whole range of antagonists used. 3) AMPA and domoate were the most and least sensitive agonists, respectively, to the action of all the antagonists. Ionophoretic experiments in the cat spinal cord in vivo confirmed this order of susceptibility in the case of the antagonists CNQX and pCB-PzDA. 4) Acromelic acid A was a more AMPA-like than domoate-like agonist. 5) The results suggest that two receptors contribute to the responses induced by several of the agonists, and that quisqualate and kainate are less selective agonists at these receptors than are AMPA and domoate, respectively.

The differentiation of “kainate” (K) and “quisqualate” (Q) sub-classes of non-NMDA excitatory amino acid (EAA) (Davies et al., 1979; Watkins, 1978) receptors was originally based on the selective depressant action of L-glutamic acid diethyl ester (GDEE) on quisqualate induced responses in ionophoretic experiments in vivo (McLennan and Lodge, 1979; Watkins and Evans, 1981). This concept was supported by the opposite selectivity shown by other antagonists subsequently developed, e.g., γ -D-glutamylglycine (DGG) (Davies and Watkins, 1985) and γ -D-glutamylaminomethyl sulphonate (GAMS)7, and by the marked differences observed in the relative potencies of kainate and quisqualate in different tissues (Shinozaki, 1978; Agrawal and Evans, 1986). The results of binding studies with 3H-kainate and 3H-L-glutamate under various conditions also generally supported the notion of separate kainate and quisqualate receptors, as did those of later experiments with the quisqualate analogue, 3H-AMPA (Foster and Fagg, 1984; Monaghan et al., 1989). In conformity with this concept, autoradiographic patterns obtained in brain sections with 3H-kainate and 3H-AMPA are quite distinct (Cotman et al., 1987).

Cationic organometallic complexes derived by complexation of vinyl ethers with the dicarbonyl cyclopentadienyliron cation (Fp+) are readily prepared, highly reactive reagents. These substances have been shown to function as vinyl cation equivalents for the vinylation of ketones, and for the synthesis of Qmethylene-'-lactones and dihydrofurans. Similar complexation of propiolic esters gives a reactive species, which functions as a n-acrylic ester cation equivalent and yields cyclobutenes, 1,3-dienes and dihydro—a-pyrones with olefins and cinnamic esters with aromatic systems. Although carbonium ions constitute important reaction intermediates, few are sufficiently stable to be isolated. By contrast, isolable cationic organo-transition metal complexes with significant charge localization on the ligands are relatively commonplace species)Their use as reagents in organic synthesis has become a focus of increasing interest, and it is my purpose to review here some of the synthetic options provided by the C5H5Fe(C0)2 cation (F+) and its olef in and acetylene complexes. It is well recognized that olef in complexation by F+ labilizes the ligand toward nucleophilic attack.2 But such activation, especially for carbtn nucleophiles, may be complicated by poor regioselectivity. The use of complexes derived from a,13-unsaturated carbonyl compounds provides a class of electrophile which is both more reactive and highly regioselective in its reactions with a wide range of nucleophiles. The methyl vinyl ether complex 1 reacts readily with both cyclohexanone lithium enolate or the related silyl ether to give only the Michael adducts 2 and 3, and these may in turn be converted to the octalones, with concurrent loss of the F group, by the action of base.3 0Li 27 KOH Ko / 2 + // F 0 I N SiMe Fp C112C12 r 00 -0 0 2 3 K0 3 Even comparatively mild carbon nucleophiles such as (i'-allyl)F 4 condense with 1, and since simple (allyl)F complexes may be functionalized by successive electrophile addition and deprotonation, more complex, polyfunctional systems may be elaborated. o + 0 0 Fp 1 Fp + Fp 4

— A correlation has been attempted between the uptake characteristics of l- and d-homocysteate and the time courses of neuronal excitation by these and other amino acids related to l-glutamate. The uptake of l- and d-homocysteate and of l-[35S]homocysteate was studied in individual slices of rat cerebral cortex at 37°C. Tissue: medium ratios attained over l0 min for the unlabelled enantiomers at 2.5 mM were 3.7 for l-homocysteate but only 0.8 for the d-isomer. The uptake of l-[35S]homocysteate over the concentration range 0.09 μm-2 mm can be attributed mainly to a low-affinity transport process with Km approx 3 mm and Vmax 1.7 μmol/g/min, but a high-affinity process of low Vmax may make a minor contribution at the lower concentrations within this range. In terms of dependence on energy metabolism and [Na+], and on inhibition by p-chloromercuriphenylsulphonate, ouabain and structural analogues of the amino acid, the main uptake system for L-[35S]homocysteate appears to be similar to that mediating low-affinity uptake of l-glutamate and other acidic amino acids. d-Homocysteate was but a weak inhibitor of this uptake system compared with other structural analogues. The time courses of excitation by 6 amino acids were determined by microelectrophoretic application to rat spinal neurones. d-Homocysteate induced responses with recovery times considerably longer than those of the other amino acids; this correlates with the absence of rapid uptake systems demonstrated for this amino acid in cortical tissue. d-Glutamate and l-homocysteate, which are only accumulated by low-affinity transport mechanisms, induced responses with recovery periods similar to those of l-glutamate, l-aspartate and d-aspartate, which are accumulated by both high- and low-affinity uptake systems. Although contributions of other factors to the observed time courses, such as rates of association and dissociation of the amino acid-receptor complexes, cannot be excluded, the present results are consistent with the hypothesis that low-affinity uptake systems of high Vmax play an important role in the rapid termination of the effects of amino acid excitants.

Mice were anaesthetized with nembutal and the effects of intraventricularly injected excitant amino acids on [U-14C]acetate metabolism were investigated. The natural excitant amino acids, l-glutamate and l-aspartate, reduced the incorporation of 14C from [U-14C]acetate into glutamine, GAB A and possibly alanine. The synthetic excitant amino acid, N-methyl-d-aspartate caused a reduction in the incorporation of 14C from intraventricularly injected [U-14C]acetate into all of the brain amino acids labelled by [U-14C]acetate within 5 min. It is suggested that these effects may be due to changes in pool sizes of tricarboxylic cycle intermediates, to inhibition of acetyl-CoA formation, or both. Differences in the metabolic effects of the synthetic and natural excitants are interpreted in terms of the uptake of the natural amino acids into glutamine-forming pool(s) of glutamate metabolism.

AFTER the discovery of the excitatory effects of glutamate and aspartate on single neurones of the mammalian central nervous system1, the possibility that these substances might act as central excitatory transmitter substances has continued to gain support. Compelling evidence either for or against this hypothesis would, however, require the use of specific pharmacological antagonists and unfortunately there has been only limited success in the search for antagonists of either the excitant amino-acid action or of synaptically-evoked excitation in the CNS2.

L-Glutamate and L-aspartate were administered electrophoretically near spinal interneurones and Renshaw cells of pentobarbitone-annaesthetized rats. Other spinal interneurones were consistently more sensitive to L-glutamate than to L-aspartate. Renshaw cells, however, showed no consistent difference in their sensitivity to these two amino acids. The results, which are compared with those reported previously in the cat, support the hypothesis that L-glutamate could be a transmitter at spinal primary afferent terminals.

1. The effects of D-alpha-aminoadipate (DalphaAA), D-alpha-aminosuberate (DalphaAS) and other excitatory amino acid antagonists have been compared on the excitatory responses of neurones of the cat spinal cord to acetylcholine, a range of glutamate-related amino acids and stimulation of appropriate excitatory synaptic pathways. The ionophoretic technique was used for administration of excitants and antagonists. 2. DalphaAA and DalphaAS had little or no effect on acetylcholine-induced excitation of Renshaw cells. Responses of either Renshaw cells or dorsal horn neurones in the spinal cord to excitatory amino acids were depressed in the order: N-methyl-D-aspartate (NMDA), L-homocysteate, D-glutamate, ibotenate greater than D-homocysteate, L-aspartate, D-aspartate greater than L-glutamate, kainate and quisqualate. 3. These effects are consistent with the existence of different excitatory amino acid receptors, one type being sensitive to the actions of the antagonists, and activated predominantly by the NMDA group of excitants, with other receptors being relatively insensitive to DalphaAA and DalphaAS and activated predominantly by quisqualate and kainate. On this hypothesis, many amino acids are assumed to have mixed actions on DalphaAA-sensitive and -insensitive receptors. 4. 2-Amino-4-phosphonobutyrate (2APB) and L-glutamic acid diethyl ester (GDEE) produced different patterns of antagonism of excitatory amino acid-induced responses from those observed with DalphaAA and DalphaAS. Neither substance was as potent as DalphaAA or DalphaAS as an excitatory amino acid antagonist. 5. Both DalphaAA and DalphaAS selectively antagonized synaptic excitation of Renshaw cells evoked by dorsal root stimulation without affecting cholinergic excitation of these cells evoked by ventral root stimulation. These latter responses were selectively antagonized by dihydro-beta-erythroidine (DHbetaE). DalphaAA also antagonized synaptic excitation of unidentified dorsal horn neurones of the spinal cord evoked by dorsal root stimulation. Neither GDEE (particularly) nor 2APB were as effective as DalphaAA or DalphaAS as depressants of synaptic excitation. 6. Taken in conjunction with the results of in vitro studies on the specificity of action of Dalpha? and related substances, these observations suggest that certain synaptic excitations in the spinal cord are mediated by an excitatory amino acid transmitter, and that this transmitter interacts with receptors which are activated selectively by NMDA, less selectively by other amino acids, including L-aspartate, and probably only slightly by quisqualate, kainate and (exogenous) L-glutamate.

At least two different types of excitatory amino acid receptors have been identified in the mammalian and amphibian central nervous systems. One type ('NMDA receptors') appears to be important in amino acid-mediated synaptic excitation, NMDA being the most potent and specific exogenous agonist for this type of receptor. Many antagonists have selective blocking actions at these NMDA receptors, and such substances are also selective antagonists of synaptic excitation in the vertebrate spinal cord. It is proposed that these receptors are transmitter receptors activated by an excitatory amino acid. In addition, extrasynaptic receptors, activated by domoate, kainate, quisqualate and L-glutamate, but not by NMDA, and only weakly by L-aspartate, have been identified on dorsal root fibres of the immature rat.

The applicability and scaling capability of RELAP5/MOD2 when applied to a Babcock and Wilcox (B and W) loss-of-feedwater transient is assessed using a code applicability methodology. A loss-of-feedwater test with a feed-and-bleed recovery was selected from the once-through integral system (OTIS) test data as a reference transient. Nondimensional comparisons are made between code assessment calculations and code applications calculations using computer code models scaled according to scaling criteria derived from the work of Ishii and others. The results indicate that RELAP5/MOD2 can scale the phenomena observed in the experiment and that the code is applicable for transients for which phenomena are within this envelope. The results also demonstrate the usefulness of the code applicability methodology for interpreting and verifying code calculations. 21 refs., 59 figs., 12 tabs.

Three excitatory amino acid antagonists, 2-amino-5-phosphonovalerate (APV), gamma-D-glutamylglycine (gamma DGG) and cis-2,3-piperidine dicarboxylate (PDA) have been compared with respect to their ability to block the action of amino acid excitants and both mono- and polysynaptic excitation in the cat spinal cord evoked by stimulation of primary afferent fibres. Each of the three antagonists depressed polysynaptic excitation of dorsal horn neurones. This action correlated with the ability of all of the substances to antagonize responses evoked by N-methyl-D-aspartate (NMDA) and L-aspartate. The order of potency of the antagonists in producing these effects was APV greater than gamma DGG = PDA. PDA (particularly) and gamma DGG also proved to be effective depressants of monosynaptic excitation. This action correlated with the ability of these substances to antagonize quisqualate- and L-glutamate-induced excitation. Antagonism of kainate-induced excitation was usually (but not always) also associated with depression of monosynaptic excitation. These results suggest that, following impulses in low threshold afferent fibres, a transmitter is released from primary afferent terminals which acts at quisqualate- (and possibly kainate-) type receptors and that excitatory interneurones, activated by similar impulses, release a transmitter which acts at NMDA receptors. L-Glutamate and L-aspartate may be the transmitters involved in these monosynaptic and polysynaptic responses, respectively.

We report here a model of submaximal spinal cord injury, employing direct thoracic cord compression in ferrets. This yields consistent and reproducible long term consequences on local cellular structure and on locomotive function. We show also that both the anatomical and functional losses can be measured objectively using simple techniques, and that useful long-tract mediated locomotion is preserved after severe axonal losses at the site of injury.

— A range of acidic amino acids differing in (i) their potency as neuronal excitants, (ii) their transport properties and (iii) their ability to act as substrates for metabolism have been compared with respect to their effects on energy metabolism of rat cerebral cortex in vitro.l-Glutamate, and d- and l-homocysteate, increased tissue slice NADH levels, and the same three amino acids, together with d-glutamate and kainate, increased oxygen uptake by the slices. It was concluded that these effects were predominantly due to neuronal depolarization and the ensuing activation of ion pump mechanisms. l-Glutamate, d-glutamate and l-homocysteate increased lactate production by the slices, whereas d-homocysteate and kainate did not. Since the two latter amino acids are the strongest neuroexcitants but probably the least rapidly transported, it is suggested that stimulation of lactate production in slices by amino acid excitants is a consequence of the energy requirements of active uptake of the amino acids, and probably occurs mainly in glial cells. Although the metabolism of l-glutamate appeared not to be an essential requirement for the effects observed with this amino acid in the present work, such metabolism may make a proportionately greater contribution under sub-optimal conditions of slice preparation and incubation, where electrical activity of the tissue may be impaired.

A novel 4-substituted derivative of piperazine-2-carboxylic acid, 3-((+/-)-2-carboxypiperazin-4-yl)-propyl-1-phosphonic acid (CPP), with potent N-methyl-D-aspartate (NMDA) antagonist activity was evaluated as a muscle relaxant in genetically spastic rats. CPP, 0.02-0.1 mmol/kg, given intraperitoneally reduced the tonic activity in the electromyogram recorded from the gastrocnemius muscle of genetically spastic rats in a dose- and time-dependent manner. Muscle relaxation was also seen following intrathecal application of CPP, 0.0002-0.002 mumol, in genetically spastic rats. CPP, 0.1 mmol/kg, while not affecting Hoffman (H-) reflexes, depressed flexor reflexes in anesthetized rats following intravenous administration. In mice, CPP, 0.001 mumol, given intracerebroventricularly preferentially antagonized myoclonic seizures induced by NMDA and quinolinate, and had no effect on convulsions elicited by kainate, quisqualate and L-glutamate. These observations identify CPP as the most potent preferential NMDA antagonist so far tested with muscle relaxant and anticonvulsant activity resembling the profile of action of 2-amino-7-phosphonoheptanoate.

Rather than attempt an overview of the whole field (which, in any case, is comprehensively and much more eloquently provided by the rest of this volume) I am taking the opportunity offered by the nature of the occasion to review those aspects of the subject with which my colleagues and I, over a period of twenty seven years, have been most closely associated. In chronicling our endeavours over these years, our failures as well as successes, some trails leading nowhere, others to completely unexpected and important findings, I hope to provide some indication, particularly to the young, of the true nature of scientific research, which rarely flows as smoothly as polished publications might suggest.

A series of substances known to combine with acetylcholine receptors have been administered electrophoretically to spinal interneurones of anaesthetised and decerebrate non-anaesthetised cats. In the majority of preparations but not in all, the sensitivity of these neurones to excitant amino acids was, reduced by acetylcholine, acetyl-β-methylcholine, β-carbomethoxyethyl trimethylammonium and carbamylcholine. Carbamylcholine also depressed the synaptic excitation of interneurones by afferent volleys. The depression by acetylcholine and acetyl-β-methylcholine was followed by an increased sensitivity to excitant amino acids. Excitation was the major effect of gallamine triethiodide and (+)-tubocurarine. Atropine, but not dihydro-β-erythroidine, reduced the action of carbamylcholine and acetylcholine. It is considered unlikely that these effects are physiologically related to synaptic processes, and their relevance to the depression and excitation of other central neurones by electrophoretically administered acetylcholine is discussed.