This thesis traces the various ways in which the aurora was imaged, visualised and understood during the International Polar and Geophysical Years of 1882-1883, 1932-1933 and 1957-1958. I explore the depiction of the phenomenon, from hand-drawings to radio echoes, while paying heed to what was occluded from portrayals, the imaginative and aesthetic considerations involved in rendering the aurora and the epistemological problems of capturing a transient, unpredictable and intangible atmospheric object. Photography, spectroscopy, radio imaging and the introduction of the all-sky perspective were integral technological developments, influencing the ways in which the aurora was presented and viewed. Nevertheless, experiential knowledge of the phenomenon, gained through watching the affective light displays and occasionally listening for its potentially illusive sounds, remained crucial to each of the endeavours. With a focus on the practices of Polar research, I trace the shifting balance between reliance on embodied and instrumental registration of the phenomenon. This perspective reveals the significance of amateur participation to the Polar Years and the centrality of outdoor, situated practices of knowledge creation, complicating our understanding of the spaces of the nineteenth and twentieth century physical and geophysical sciences. The project to perfectly reproduce the aurora, and thus come to know it, was from the outset an impossible task. This thesis is, therefore, a story of incremental learning, of the calibration and standardisation of the phenomenon across vast distances, of bringing together fragments of the aurora’s ontology to create a fuller, more complete picture of the phenomenon, and of both fallibility and success.