Embry o culture. Whole embryo culture was performed as previously described (5). Wistar rat embryos were explanted on gestational d 9.5, d 0 being the day of positive vaginal smear.Embryos were explanted in Tyrode's solution, the decidual mass was removed, and Reichert's membrane was opened. Embryos were then placed in roller bottles containing heat-inactivated rat serum and cultured under an atmosphere of 90% N z , 5% Oz, and 5% COz as described by New (6). At explanation, embryos contain up to three somites and the primitive heart tube has not yet formed. Embryos were studied either at the five-to sevensomite stage (24 h in culture), at which stage the heart tube has formed but has not begun to loop, or at the 13-to IS-somite stage (36 h in culture), after cardiac looping is complete but before development of the atrial or ventricular septa.Wall motion recording. Embryos with prelooped or looped hearts were examined in Tyrode's solution using a Wild M5 microscope. Embryos were positioned and immobilized as shown in Figure I in a depression made in 9% agar. Embryos were maintained at 37.5 ± OSC by a thermostatically-controlled stage warming/cooling device (Sensortek, Clifton, NJ). Video-tape recordings of the embryonic hearts (at lOOx and 200 x magnification) were made using high-resolution closed circuit video and electronic processing to produce monochrome images of maximum contrast. For wall motion recording, these videotape recordings were played back on a 19-inch monitor. Fiber optic cables were placed over the video images of the proximal atrium, distal atrium, proximal ventricle, and proximal bulbus cordis as indicated in Figure I. The fiber optic cables were connected to photocell amplifiers that converted the changes in light intensity of the video image caused by wall motion to voltage signals, which were then recorded by a multichannel recorder at a paper speed of 50 mm /s. This apparatus has been described previously (7). Because of the nature of the opt ical system , the moment of peak wall motion is the most reliably detected point. Peak wall motion is also less affected by motion of adjacent structures than earliest motion would be, so position of the fiber optic cables was adjusted to optimize recording of peak wall motion. Cycle length was measured from the most proximal site recorded. Intervals between peak wall motion in the proximal and distal atrium (intra-atrial delay), distal atrium and proximal ventricle (atrioventricular delay), and proximal ventricle and proximal bulbus cordis (intraventricular delay) were measured to the nearest 5 ms as shown in Figure 2. Each measurement reported was the mean of five sequential contractions. Linear distance between recording sites was measured on the video screen and converted to JIm by reference to micrometer calibration recorded on each videotape at the same magnifications.Stat istical analysis. Intervals between maximum contraction within the atrium and ventricle and atrioventricular delay in prelooped and looped hearts were analyzed by analysis o...