Much of the periodical literature within scientific disciplines is characterized by overlap in content among journals and by scatter of content areas across journals. This obscures the inherent structure that may exist within the literature of discipline, creating problems for authors seeking suitable publication outlets, for editors wishing to establish coverage policy, and for libraries wishing to be selective but representative in acquiring journals for a discipline. Four factor analytic studies were done on samples of the periodical literature of physics, taken from Physics Abstracts (1961), in an attempt to define its inherent structure. The first study resulted in ten factors interpreted as journal groups within physics. The remaining three studies resulted in factors that define relatively homogenous topical or content areas. The consistency of results across the analyses justified the conclusion that factor analysis methods are applicable to defining the structure of the literature of science. (Author)