This paper sets out to investigate the place and status assigned to geometry in various contributions on the topic of homology relationships, starting with the papers written by Poincaré on Analysis situs, and pursuing the issue up to the early 1930s. For that purpose, the concept of « geometrical content » is introduced, and is shown to be indeed present in writings by Poincaré, Veblen and Alexander, while it is absent, on the other hand, from the contributions of other authors (Vietoris, Cech). Further, the analysis of certain distinctions introduced by these writers allows an appreciation to be arrived at, of the extent to which their insistence on a « geometric meaning » acted as a compelling imperative and constraint in their work. The analysis in fact points to the variety of stances adopted by these mathematicians, with regard to geometric meaning, thus complementing the examination of the status assigned to geometry, on the basis of the analysis of geometrical content. Beyond the historicity of mathematics — as a historical process in the making — thus highlighted by these modes of analysis, the semiotic variance and diversity itself of mathematical writings is thus made manifest.