According to Oswald (6), the use of a mother's response to sound from her child as an example of the ability to discriminate auditory stimuli during sleep dates back to 1830. Recent references to this ability in mothers can be found in the papers of Evans et al. (3) and Wilson and Zung (11, 12). Therefore, it is surprising that this response has not been studied in the labora tory, although other forms of auditory dis crimination have been. Oswald found that his subjects were more likely to awaken to the sound of their own name than to the sound of the name of another person. Wilson and Zung (11) reported that there is a difference in the general arousal system between men and women for insignificant stimuli but they found none for significant ones. These authors hypothesized that facili tation of arousal in women is related to the feeding of newborn infants and leads to the capacity of the mother to be alerted to the cry of the infant while she is asleep. Indeed it is known that the sleep architecture is modified during late pregnancy and the early postpartum period (4).The aim of the present study was to deter mine, under laboratory conditions, the dif ferences, in awakening threshold in pregnant women to a baby's cry, to a voice saying "baby is crying" and to a speech noise, both before delivery (phase I of the experiment) and after (phase II). MethodsThe sample consisted of six women, aged 21 to 32 years. They were about 8V2 months pregnant during phase I. Sleep EEG record-

The effect of chlorpromazine on the REM deprivation rebound was studied. Five normal subjects participated in the experiment which consisted of three periods of nine nights each. The first period was without medication, the second with medication and the third again without medication. Four dependent variables — Total Sleep Time, REM latency, REM time and REM percentage — were measured in each period as follows: baseline — three nights, REM deprivation — two nights and recovery — four nights. Analysis of variance (three-way classification) was computed in order to evaluate both the main effects as well as the interaction between medication and sleep conditions. The results showed significant difference over the three conditions of sleep — baseline, REM deprivation and recovery. While there were no significant differences due to medication the replication of the measurements in the same subjects proved to be a slightly significant source of the variance of the scores in all variables except REM latency.