A scheme for ob initio simulations of extended systems, which involves the use of the elecmn density as the basic variable, is discussed The form of the kinetic energy functional is chosen U, incorporate several exact limits (uniform system linear response and rapidly varying density) while the r e ~t of the energy functional is exactly the same as in a K o h S h a m density functional calculation with the local density approximation. We show that for sodium the present scheme yields highquality results in a fraction of the time required with an orbital-based functional. The elecmnic part of the algorithm scales linearly with system size. An analysis of the stability of the method is made, and leads U, criteria for selecting the non-physical parameiers in the calculation so as to maximize the computational efficiency.

475 patients with suspected uncomplicated myocardial infarction (MI) were divided into 3 groups based on their entry ECG: group 1--significant ST elevation; group 2a--ST depression or T inversion; group 2b--normal ECG. Infarction was confirmed in 99.7% of group 1, 68.5% of group 2a and 39.7% of group 2b patients. Despite similar clinical, haemodynamic and historical variables at presentation, group 1 patients had significantly larger MI, more in-hospital complications and a higher short-term and long-term mortality (P less than 0.005) than group 2 patients. The entry ECG of patients with suspected MI is an excellent and simple predictor of those who are most likely to have an MI confirmed and identifies a group of patients at high risk of death or developing complications.

SUMMARY1. Stimulation of left atrial receptors by distension of a balloon in the lumen of the left atrium of anaesthetized dogs was shown to result in an increase in urine flow. Plasma samples obtained from these dogs during control periods and during periods of stimulation were applied to the Malpighian tubules of Rhodnius prolixus.2. It was found that the tubules suspended in test plasma secreted at a significantly lower rate than those suspended in control plasma.3. These differences were also evident in extracts of plasma prepared using the solvent n-butanol.4. Cutting or cooling the cervical vagi abolished these differences along with the increase in urine flow. It is argued that this preparation of the Malpighian tubule of Rhodnius prolixus could be used as a means of detecting the diuretic agent responsible for the increase in urine flow.

SUMMARY1. To determine whether stimulation of atrial receptors reflexly results in a consistent change in plasma renin activity experiments were performed in dogs anaesthetized with a-chloralose. Either a large balloon in the lumen of the left atrium or small balloons positioned at the upper and middle pulmonary vein-atrial junctions and atrial appendage were distended. Plasma samples obtained from these dogs during control periods and during periods of stimulation were assayed for plasma renin activity.2. In the first series of experiments, stimulation of left atrial receptors by either large or small balloon distension was shown to result in an increase in urine flow and a decrease in plasma renin activity.3. In a second series of experiments, cooling of the cervical vagosympathetic nerve trunks abolished the response of an increase to balloon distension in urine flow and decrease in plasma renin activity.4. It was shown that a decrease in plasma renin activity resulted in response to stimulation of atrial receptors, and that this decrease is reflex in nature and is likely to involve only atrial receptors discharging into myelinated vagal fibres.