Historians in the twenty-first century are faced with an abundance of digitized material. Large-scale scanning efforts have made millions of textual primary sources available to historians at the click of a mouse. This paper discusses using topic modeling, a probability algorithm, to extract themes and discourses in large corpuses of digitized text. Mind and Body, a monthly publication about physical education, which ran from 1894 until 1936, was often at the very center of developments in the field of physical education. Major developments such as professionalization; the battle between the American, German, and Swedish exercise systems; and shifts in ideas about the role of exercise in American life were discussed and debated within the pages of Mind and Body. Applying distant-reading methodologies to such a central text aids scholars in identifying themes, discourses, and points for further examination.