I have attempted two major tasks in this thesis. First, I argue that Deborah Mayo’s Error Statistical epistemology makes an excellent tool for historical research into experimental episodes. This is because it focuses the historian’s eye on the nitty gritty details of experimental arguments, particularly on the generation and manipulation of data. Moreover, her hierarchy of models provides an excellent organizing tool for disentangling complex experimental narratives. I illustrate the fruitfulness of this method by contrasting John Dorling’s and Ronald Laymon’s summaries of Cavendish’s Great Globe experiment with my own account. Second, and perhaps less successfully, I have used her concept of "arguing from error" along with her attendant hierarchy of models and severity criterion to make claims for the procedural objectivity of Cavendish’s experimental tests of an inverse square force law for describing electrical attraction and repulsion. Simultaneously, I confronted Harry Collins’ experimenters’ regresses and Pickering’s view of experimenters’ "tinkering" (in his mangle of practice) and show that neither is either a necessary part of experimental practice nor holds for Cavendish’s experiments. / Master of Science