One difficulty arose in interpreting the early absorption; there was a wealmess or complete lack of echo traces over the entire frequency range during early morning hours even on days when no shot tock place. This has been attributed to a combination of low receiver-gain settings and antennas with poor low-frequency sensitivity. Despite the above-mentioned difficulty, it seems possible to state that, on shot days, there appear to have been three periods of absorption which more or less overlapped. A separate cause may prob-ably be attached to each. First, due largely to the equipment in- adequacies discussed above, the normal pre-sumrise decrease caused the signal intensity to fall below the threshold of detectability. Since this coincided with shot time, it accoumts for the beginning of what was called complete "absorption" coming just before or after each shot. Second, an effect connected with the overhead arrival of the shock wave in the lower ionospheric layers was expected to start at about plus 15 min at Rongerik and about plus 20 min at Elmer for shots. at Bikini. This effect is the increased absorption of radio signals Passing through the highly disturbed D region. From Tables 3.1 and 3.2, it appears that this effect lasted wmtil about plus 35 min at Rongerik and somewhat longer at Elner (which is about 50 miles more distant from Bikini). The third period of absorption, which seems to have occurred only at Elmer, began about 1 1/2 to 3 hr after Shots 1, 4, and 5, and an indeterminate time after Shot 2. about plus 10 hr (plus 12 1/2 hr for Shot 2). It lasted until A conjectural explans- tion of this protracted effect is given in Section 4.1. The lengthy absorption after Shot 6, detonated across the Eniwetok Lagoon from Site Elmer, may be attributable to a similer cause. 302 SPORADIC-E EFFECTS In common with results of previous ionospheric experiments connected with nuclear tests, a tremendous amowmt of sporadic-E activity occurred following all shots, as evidenced by the data taken at Rongerik, At Elmer, the data were very scanty due to absorption (see Section 3.1) and equipment inadequacy, but occasiona}. glimpses ofunusual sporadic-5 returns indicated that E-region conditions there were similar to those above Rongerik. In order to obtain a quantitive comparison between post-shot and normal spovadic-E: conditions, the graph of Fig. 3.3 was plotted. The normal curve (deshed) depicts hourly median values of highest sporadic-E frequency (fEs) at Rongerik for three weeks prior to Shot 1; it is placed with reference to.the time axis so that zero time for it corresponds to the average zero time fc: Shots 1 and 2. Curves for Shots 1 through 5 (solid) connect points each representing the average, over a 15-minute period centered thereon, of the highest sporadic-E frequency found each minute (each five minutes, for Shot 3). 26 SECRET