All in all some serious but limited harm has been done. It was a close shave. To see how close, one only needs to glance at the map below, which shows the roentgen dosage for 48 hours of exposure. At the southern tip of Rongelap, where the inhabitants lived, the dosage was 175 roentgens. But at the northern tip, less than thirty miles away, the dosage was more than a thousand roentgens. If the wind had veered just a little bit farther to the south, probably all of the people on Ailinginae, Rongelap, and Rongerik would have been killed. [Illustration: Dosage in First 48 Hours After Fallout Began] This shot proved what had been argued for many years: that radioactivity is not just an incidental part of an atomic explosion. The people on Rongelap were far outside the area of danger from blast and thermal effects. But they received a sizeable dose of radiation. In fact, a person could have stood unprotected at a distance of thirty miles from the explosion and been perfectly safe from the blast and thermal radiation. But at that same distance in a downwind direction he would have accumulated a lethal dose of radiation within a matter of minutes after the fallout began. Because of the radioactive fallout, the test sites must be located in remote parts of the world. It would be desirable if sites could be found which are so remote from populous areas that the tests could be conducted without regard to the direction of the winds. Unfortunately the bombs are too big and the planet is too small. As a result the winds must be watched before every test; and the tests must be delayed until the winds are favorable. What happened to the Marshallese was an accident which might have been avoided if the winds had been blowing more directly toward the north at shot time. Since this accident the wind requirements for the tests have become far more stringent, our knowledge of the danger has increased, and the rules of safety have in all respects improved. Many large yield weapons have been tested since March 1, 1954, but no other accidents have occurred. We can be confident that accidents of this kind are now very improbable. At the U. S. test site in Nevada there has been no instance of a major fallout on a populated area. Probably the most worrisome situation which has occurred there was in the spring of 1953 during the Upshot-Knothole test series. After the ninth shot of the series the cloud drifted eastward over St. George, Utah, a town of about 5000 people. Some fallout occurred shortly before nine o’clock in the morning. About nine-thirty AEC officials issued a warning advising the residents to stay indoors. By noon the warning was withdrawn and people were allowed to continue with their normal affairs. The incident left everyone a little bit scared, but no one had received a radiation dose greater than two or three roentgens.