The application in this hospital of alternating current for electro-convulsive therapy (E.C.T.) and electronarcbsis in certain mental disturhances afforded the opportunity of investigating the effects of this form of therapy on some of the components of the hlood. Schiitz (1942) ohserved in vitro a delayed coagulation of blood following application of a weak direct current.Brecht and Kummer (1943) noted a leucocytosis immediately after electrical shock; there was no alteration of hlood sugar or protein.Spiegel-Adolf, Spiegel et al. (1945) reported that following the application of a 60 cycle current at 30 volts for 1-2 seconds there were changes in cerehrospinal fluid suggesting the hreakdown of nuclear substances. This statement, however, has since heen contradicted by M. Ii. Hack (1947). Huddleson (1946) observed haemorrhages in the central nervous system in rats following electrical shock. Trojaniello (1947) indicated that dogs subjected to electrical shock by alternating current at 120 volts for a fraction of a second, showed no changes in blood lipase and diastase, but that phosphatase aetivity decreased for about 12 hours. Delay and Soulairac (1943) observed, following E.C.T. in human subjects, a transient inerease in blood albumin, and oeeasionally a slight increase in the globulin fraetion. There was a concomitant increase of blood calcium and inorganic phosphate. Purther, hyperglycemia, leucocytosis, and a decrease in the alkali reserve were demonstrable. EXPERIMENTAL.Electrical convulsions were induced in suitable patients by alternating current (50 cycles) at 120 volts. Three to eight shocks, each of a duration of 0-1 seconds, were given. Blood samples were obtained by venepuncture before and after application of the current and mixed with one-ninth volume of 0-1 M sodium, oxalate solution. The ratio of plasma: cell volume was determined in tubes of uniform size and shape and centrifugation was carried out under identical conditions, which should ensure compai'able results, if not absolute values.

A survey has been made of the amount of muscle water available to inulin, sucrose, and radioiodinated human serum albumin (RISA). The percentage spaces available to the three molecules are of the same order of magnitude, but the sucrose space > inulin space > albumin space. The kinetics of influx and effiux of RISA have been studied, and it appears that a small part of the albumin may be adsorbed in the extracellular phase. Nevertheless the albumin space would appear to give the best index of the extracellular volume.The scatter in values found for the extracellular space by all methods is very great, ranging from 8 to 40 per cent and renders invalid the use of a mean value for the calculation of intracellular concentrations. The variation within paired muscles is less than between pairs, provided the tissue has undergone no volume change. Increase in total muscle volume when the muscle is placed in a hypotonic solution leads to a decrease in the size of the extracellular space.All work on the distribution of electrolytes between the external medium and the sartorius muscle depends for its interpretation on an estimate of the extraceUular volume of the muscle. This also applies to any attempt to relate the ionic gradients across the muscle membrane with the bioelectric potentials. It is consequently of prime importance to know the dimensions of the extracellular space, and the degree of variation in the dimensions of the space within a pair of muscles. Estimates of the extracellular volume of frog muscle in the literature range from 10 to 12 per cent (1, 2) to 35 per cent of muscle weight (3, 5, 7), depending apparently on the method of estimation, and the species of frog used.In this study we have attempted to clarify the situation in the following ways. First we have undertaken a comparison of the accuracy of the various methods of estimating the space. Second we have evaluated the scatter in values found in a population of animals, in order to ascertain whether it is legitimate to use a mean value when calculating intracellular concentrations.

Discrepancies in the membrane theory of K accumulation in muscle and nerve cells led us to re-examine the behaviour of these tissues when soaked in Ringer's solution under various conditions. This work gained further impetus from the recent paper of Ling (19o'2), who has postulated a selective ionic adsorption to account for the high K content of these tissues.The use of radioactive isotopes has disproved any theory involving specific ionic impermeabilities. Subsequent to this, the principles underlying the accumulation of K have centred around the Donnan principle, supplemented by the concept of a Na pump (Dean, 1940; Eccles, 1953).The main consequences of this hypothesis (apart from electrical phenomena, which will be considered in a later paper) are: (i) the ratio of internal to external K should vary with the external concentration according to a Donnan equilibrium, (ii) K movements into or out of the cell should be balanced by Na movements in the opposite direction, or anion movements in the same direction, (iii) when the cell is dead (and there is no energy to maintain the Na pumj)) the cell should eome to ionic equilibrium with its surrounds, i.e., Na in/Na out = Kin/Kout and is greater than 1.That these three criteria are not fulfilled will be evident from the work reported in this paper. This has also been shown by Solomon (1952), who stated on thermodynamic grounds that K movements could not passively result from active Na movements. Lastly, Harris and Maizels (1951) liave demonstrated that none of the above criteria apply in the ease of the human erythrocyte; i.e., the external concentration of K, within wide limits, has little influence on the concentration of eell K attained after 20 hours; the rate of change of Na content is greater than that of K content, and finally, when the cell is dead, the ratio of external K to internal K only rises to about 0-5 after 4 weeks' standing.To aeeount for their findings, these authors (and others) postulate the existence of one or more carrier mechanisms for the transport of these ions, and in most cases an attempt is made to link the Na and K shifts.In this paper we have tested the consequences of the carrier theory either by itself or in conjunction with the Donnan postulates. Our results have led us to discard both these hypotheses and have indicated that a more plausible AoBtraL J. exp. BioL (105S), 33, pp. 153-178.

No abstract