1. Synchronization of intercostal motoneurones was studied by the construction of cross‐correlation histograms which related the firing times of paired groups of efferent inspiratory or expiratory discharges recorded from filaments of the external or internal nerves of anaesthetized or decerebrate cats.2. The principal feature of the histograms was always a central peak but the time course of the central peak showed considerable variation. Three forms of synchronization were defined on the basis of the time course of the central peak: (i) short‐term synchronization (Sears & Stagg, 1976), where the peak was narrow, extending over about ±3 ms but sometimes with weak shoulders to about ±5 ms; (ii) broad‐peak synchronization where the peak was wider than this (often ±20 ms or more) but where there were no strong periodicities; (iii) high‐frequency oscillation (h.f.o) synchronization, which was named from the related phenomena in medullary and phrenic recordings (Cohen, 1979), where there were periodic peaks on either side of the central peak with a frequency in the range 60‐120 Hz. Combinations of these forms of synchronization were seen in some histograms.3. When different animals were compared, broad peak synchronization was seen in association with light anaesthesia and with polysynaptic excitation of the motoneurones from muscle spindle afferents.4. In individual animals, additional anaesthesia depressed both broad peak and h.f.o. synchronization.5. Raising PA, CO2, which increased the respiratory drive to the motoneurones, favoured short‐term or h.f.o. synchronization at the expense of broad‐peak synchronization.6. In three decerebrate animals only short‐term or h.f.o. synchronization was seen.7. Spinal cord lesions above or below the segments of interest promoted broad‐peak synchronization, even with high PA, CO2 or deep anaesthesia.8. We conclude: (i) that short‐term synchronization, due mainly to the branching of presynaptic axons, is generated mainly by those axons which transmit the respiratory drive, that drive providing most of the excitation of the motoneurones in moderately deep anaesthesia; (ii) that h.f.o. synchronization arises from the periodic synchronization of the discharges in these same presynaptic axons; (iii) that broad‐peak synchronization is generated by the activity of other presynaptic neurones whose discharges are also synchronized, but aperiodically, these neurones most likely including spinal cord interneurones which are active in light anaesthesia or when released by spinal cord lesions.9. These conclusions are supported by comparisons between intracellular recordings from inspiratory motoneurones in animals showing different forms of motoneurone synchronization, the comparison including the measurements of ‘average common excitation’ (a.c.e.) potentials (Kirkwood & Sears, 1978).

An improved method for the determination of foetal heart rate from Doppler ultrasound signals is described and evaluated. It determines the most probable pulse interval from the recurrence times of multiple echoes generated by each cardiac pulse. The method, when optimised, is shown to offer an improvement over current systems, especially in reduced signal loss.

The Doppler ultrasonic technique for the monitoring of the foetus has been available for well over a decade. It offers a noninvasive assessment of the condition of the foetus and has therefore become an important antenatal technique. However, the foetal heart rate information obtained by this technique is considered to be less precise than that obtained from the foetal e.c.g. The waveforms recorded by the Doppler ultrasonic method are complex and variable and this often leads to difficulty in identifying a reliable and regular measuring point within these waveforms from which to calculate the heart rate. This has resulted in a cautious acceptance of ultrasonically determined foetal heart-rate records. This paper does not dispute the difficulties of interpreting such rate information, but outlines a pair of algorithms for processing and presenting a more accurate and reliable indication of the foetal heart rate determined from the Doppler ultrasonic method. These algorithms calculate a two second average rate from only those intervals which have been determined to be ‘valid’, based on the criterion that they should be within ±10% of the most previous valid interval.

An instrument has been developed to help speed up the clinical assessment of the thyroid-gland condition as indicated by the Achilles-tendon-reflex duration. This is achieved by processing the signal from a commercial photoelectric displacement transducer and immediately providing both an image display and a digital readout of the duration. Incorrect operation is detected and displayed by an appropriate error-indicating lamp, and the readout is inhibited. Although the instrument is designed for this particular clinical situation, it may find other applications where a similar signal is to be timed. Overall, the operation of the instrument has been very successful, allowing rapid analysis while maintaining simplicity of operation.