The experiments were designed to determine the influence of the upper respiratory tract (URT) on liquid flow in the fetal trachea. This flow probably influences pulmonary distension, which is thought to be a major determinant of prenatal lung development. In six fetal sheep the URT could be bypassed by connecting the lower trachea, via an external flowmeter, to a cannula in the amniotic sac. In confirmation of our earlier findings, when the URT was in circuit, the mean rate of tracheal efflux was greater during episodes of fetal breathing movements (FBM) [mean 13.8 +/- 2.6 (SE) ml/h] than during apneic periods (mean 3.2 +/- 1.0 ml/h). When the URT was bypassed there was a reversal of net tracheal flow during FBM episodes (mean 19.6 +/- 5.6 ml/h toward the lungs); during apnea there was a much greater rate of efflux (mean 33.1 +/- 10.2 ml/h) than when the URT was in circuit. Nonlabor uterine contractions were associated with an increased rate of efflux during apnea only when the URT was bypassed. We conclude that during fetal life the URT imposes an essentially unidirectional flow of pulmonary liquid away from the lungs, preventing ingress of amniotic fluid and maintaining constancy of composition of liquid in the developing airways. By retarding outward flow during periods of apnea and thoracic compression and by preventing net influx during episodes of FBM, the URT has the probable effect of maintaining the volume and composition of liquid in the fetal airways within narrow limits.

Fetal breathing movements (FBM) and lung liquid volume are known to affect lung development, but little is known about mechanisms controlling movement of liquid through the upper respiratory tract (URT). Therefore we measured resistances of the URT in 8 unanesthetized fetal sheep during late gestation while FBM were monitored from pressures in the lower trachea or from electromyogram of respiratory muscles. URT resistance to liquid flow toward the amniotic sac increased from 3.5 +/- 1.9 Torr X ml-1 X min during episodes of FBM to 21.1 +/- 5.7 Torr X ml-1 X min during apnea. Laryngeal resistance during apnea was greater (P less than 0.001) than supralaryngeal resistance in each of six fetuses in which URT resistance was partitioned. Fetal paralysis abolished the increase in laryngeal resistance to efflux that was previously related to the high-voltage electrocortical state and apnea. We were unable to quantify URT resistance to fluid movement toward the lungs because the larynx acted as a valve, permitting flow toward the lungs only in the presence of FBM. The supralaryngeal portion of the URT also apparently acts as a valve, normally preventing the entry of amniotic fluid into the pharynx. These findings help to explain our earlier observations that efflux of liquid from the fetal lungs is greater during episodes of FBM than during apnea.

SUMMARYSwallowing was recorded electromyographically in unanaesthetized fetal sheep in utero during the last third of gestation. In control fetuses bouts of swallowing were, on average, 2 0 min in duration and occurred at intervals of 2-31 h. Swallowing bouts were associated with a low-voltage electrocorticogram, eye movements and, in 82% of instances, with phasic activity in the dorsal neck muscles. These data indicate that fetuses are not in quiet (non-rapid eye movement; n.r.e.m.) sleep when swallowing and also suggest that they are not in r.e.m. sleep as conventionally defined; we consider that fetuses may be in a state of heightened activity resembling wakefulness. Swallowing bouts were always accompanied by fetal breathing movements. When a swallow coincided with an inspiratory effort the diaphragm was briefly inhibited. Evidence was obtained that bouts of fetal ingestive activity were not dependent upon outpourings of fluid from the lungs and were not significantly influenced by volume information from the gastric compartments.

Summary. The co-ordination of spontaneous and evoked electromyographic (EMG) activity was examined in the uterus of ewes from Day 100 of pregnancy and during labour. During pregnancy the onset of bursts of EMG activity was synchronous to within 2\p=n-\3min when recorded by electrodes at any recording site. Individual EMG spikes, whether occurring spontaneously or evoked by electrical stimulation of the myometrium, appeared to propagate between recording electrodes that were 1\p=n-\3cm apart, in the longitudinal or circular orientation of the uterus, but not when the electrodes were separated by more than 3 cm. Apparent conduction velocity in the longitudinal orientation was 7 cm sec\m=-\1 during pregnancy and increased to 13 cm sec\m=-\1 during labour. No consistent pattern could be distinguished in the sites at which EMG activity was first observed at any stage of pregnancy studied. The nature of propagation of EMG activity remains unresolved but it is clear that it is not initiated at a single site in the myometrium.

SUMMARYOesophageal cannulations were performed in seven fetal sheep between 109 and 128 d of gestation. Fluid swallowed by the fetuses was collected externally and periodically returned to the fetal stomach by a peristaltic pump. The volume of fluid swallowed daily was calculated from the movement of water in a parallel pump circuit. Samples of oesophageal fluid were removed from the collection bag at the same time as tracheal and amniotic fluid samples.