The young Princetonian geniuses very likely get acquainted with Bohr's assistant Aage Petersen, who pursued an interest in quantum mechanics with religious zeal [19, page 10]. At one party in the Graduate College, after a good bit of sherry, Petersen steered a discussion with Everett and Misner to paradoxes of quantum mechanics. Misner had not yet thought deeply about these paradoxes, but Everett already had. [19, pages 1, 4]. The 24-year-old Everett, obviously, was already a crackerjack thinker, a feature mentioned by all who knew him. He, probably not realizing himself the true scale of his impromptu remark, offered a conceptual scheme in which the inconsistencies (the so-called paradoxes) were removed. That was the idea that the next year would yield a major piece of work, finished even before his dissertation, about the basis of quantum mechanics and would later immortalize his name.
Despite the shift to physics, Everett continued to work in mathematics. In December 1954 (half way through his second graduate year), he delivered a lecture on military applications of game theory in Washington, DC. His Christmas arrival at his parents that year, together with his friend Arnold, was worthy of attention as a local news item  (Evidently "parents" meant his father and stepmother Sara T. ). (Incidentally, Hugh's father, Colonel Everett, was transferred at this time from Alexandria to the Military District of Washington, DC as the head of supply and logistics. Later he rose to become the chief of staff [6, 8]).
For some months after that Christmas vacation, Everett buckled down to get ready for the general exams, which he passed in the spring. Not until the summer of 1955 did he begin to write up his ideas on quantum mechanics. The resulting 137-page manuscript was typed by Nancy Gore (February 13, 1930-November 11, 1998 ), whom Everett married a year later. [19, page 6; 24]. Someone advised Everett that if he wanted to finish his dissertation more quickly, he should transfer to John Wheeler as an advisor. Wheeler, who had been a postdoc with Niels Bohr in the 1930s and had collaborated with Bohr on the 1939 theory of fission, had served as a principal scientist in the Manhattan Project [19, page 2]. Everett probably approached Wheeler around the end of 1954 (the middle of his second year at Princeton). He (Everett) later recalled that before writing the long manuscript he went with the idea to Wheeler and asked, "Hey, how about this, is this the thing to do?" [19, pages 2-3] In the "Calendar of Events", composed by Everett's widow in 1990, the time of writing the dissertation for Wheeler is given as winter 1954-55 . This is undoubtedly in error. The archives show that in both terms of Everett's third year (1955-56), he, under Wheeler's guidance, worked on a dissertation referred to in the fall as Correlation Interpretation of Quantum Mechanics and in the spring simply as Quantum Mechanics) . The formal submission of the dissertation did not occur until spring 1957  and the famous paper based on it was published in July 1957. 
In September, 1955 (the beginning of Everett's third year at Princeton) he presented two small papers to Wheeler. (In Everett's archives, in the same folder with these two, there is stored a third paper, just four pages in length, that may have been written earlier. This third paper deals with objective vs. subjective probabilities. In it, he proves the inconsistency of the concept of objective probability and chooses as the most fruitful way to consistency an acceptance of the concept of hidden variables. The marginal remarks in this paper probably belong to Shoemaker, because the handwriting is different from the usual handwriting of Wheeler ). In one of the September papers submitted to Wheeler, Everett introduces a new concept-the correlation of values X and Y (not to be confused with a coefficient of correlation), based on the expectation of change of the quantity of Shannon information about X depending on information about value Y. The paper concludes with a formula for the correlation of observable values X and Y, described by a wave function .
In the second of these two papers, this one nine pages in length, the concept of "Everettism" appears for the first time. Everett writes about splitting of the observer at each measurement (Wheeler wrote in the margin: "Split? Better words needed."), and about a branching "life tree," and admits that this beautiful physics has philosophical implications that must be addressed. In his summary, Everett illustrates the concept by an image of splitting an "intelligent amoeba with a good memory" (Wheeler wrote in the margin: "This analogy seems to me quite capable of misleading readers in what is a very subtle point. Suggest omission.") .
On September 21, 1955 Wheeler wrote Everett a note, judging both papers as important works. The first one, on correlation, he is ready to send somewhere for publication, but as to the second one, "Probability in Wave Mechanics," he say he is "frankly bashful about showing it to Bohr in its present form" since it can be "subject to mystical misinterpretations by too many unskilled readers" . So, it seems that Everett's theory was too advanced for its time. (Everett received his master's degree that year, probably before submission of these papers to Wheeler. At that time in physics at Princeton, passing the general exams was all that was required for the M. A. degree. Only the Ph.D. required research accomplishment.)
Everett's main 137-page work, "The Theory of the Universal Wave Function" , is dated January 1956. (It was reprinted in a 1973 collection ). Chapter II of this work was taken from his unpublished article on correlation. Everett recalled later that Wheeler hurried him to a dissertation defense before his third year ended in the spring of 1956, although he (Everett) would have preferred delay because leaving the University might have meant being drafted into the military [19, page 6] (the Korean War had recently ended, and being drafted was still a possibility). Everett later thanked Bohr, H. J. Groenewald, Petersen, A. Stern, and L. Rosenfeld for criticism . But something did delay his defense. Perhaps it was Wheeler's leaving to accept a Lorentz Professorship at Leiden University  for the period January to September 1956 [24a, page 248]. In any case, his startlingly original and important work on quantum mechanics caused much less of a stir than it should have, and Everett turned toward a new career full of military secrets.