Concerning a personal experience with one of the early phases of an attack, Mr. Furlong states as follows:--"I am fully convinced that one night, while camping alone with Onas in the heart of the Fuegian forests, that my head man Aanakin, who had a good many killings to his credit, was brooding as he sat in his wigwam, which opened towards the fire; he watched me for nearly an hour with an attitude and expression which reminded me of the look a dog takes on sometimes before he snaps. Aanakin I knew to be of a very moody nature but this particular mood was so marked and portended evil so noticeably toward me without any apparent cause, that I decided to do something to break its mental trend. So putting fresh wood on the fire, to make a more brilliant blaze, I walked directly into his wigwam and motioned to one of his two wives, who were lying beside him. There was a passing look of half-anger, half-surprise, but I gave no time for his mind to dwell in the same mood, for simultaneously I produced my note book and pencil and began to make drawings of animals and other things they were familiar with. They like to watch one draw and name the thing, and so I kept them busy for perhaps an hour, and finally had them in gales of laughter. I am quite convinced that I forestalled an attack or a condition akin to it." It seems that an attack usually begins suddenly. However, an instance is given where an Ona became moody and realized that one of these attacks was coming on and putting his hands together begged to have his wrists and feet bound in order that he would not do himself or others any harm, or that it would not be thought that he meant to kill and consequently be shot in self defence. This would in a way seem to indicate that there was no amnesia for the attack, as the Indian undoubtedly realized what he had done in previous attacks. The moody state and the realization of what might follow as the attack comes on demonstrates a sense of uneasiness as the premonitory symptom of an attack, which ends in a state of utter exhaustion and sleep. The normal condition is resumed, practically on the awakening from sleep and recovery of strength. From a description of Donald McMillan the explorer, the Eskimo Piblokto strongly resembles these attacks of the Ona and Yahgan Indians with the exception that Piblokto was particularly prevalent among the women. How an attack begins is shown by the case of Aanakin, an Ona of Furlong's expedition. A certain form of melancholia, brooding or moodiness, seems to precede many of these attacks, with a realization sometimes that an attack is coming upon them. The Onas not being naturally a quarrelsome people, it may be that this realization and foreboding of the attack accounts for their tendency to run away from their associates, when they have endured the strain as long as they can, thus placing themselves in a position to avoid deliberate attack or injury to those about them.