Pre-Lavoisian chemistry, maintaining interfaces with physiology, pharmacy, mineralogy, heat and electrical science, and manufacturing, provides analytical access to a broad, qualitatively diverse array of observational practices and their development. These are studied in representative samples from the work of Hales, Cullen, Black, Macquer, Cavendish, and Rush. They included ancient and canonical practices, in some sense normative for the science (e.g., chemical tests), with their more recent additions, techniques for observing new and difficult materials (e.g., gases), and for "unobservables" (e.g., latent heat), even for "nonesuches" (e.g., phlogiston). Much of this project was motivated by the notion of "sensory labour," thus, whilst attention is given to increasingly instrumentalized and quantified observation (volumetrics, gravimetrics, thermometrics), stress was also laid upon the stable survival of the sensory apparatus of chemical practice (colour, turbidity, odour, taste, texture, sound), and the taciturn skill and knowledge embedded therein. The emergent picture was at some variance with theoretical histories of chemistry in the period, which tend to assimilate the science to matter-theoretic natural philosophy. By contrast, the focus upon observation renders chemistry as akin rather to the observational and taxonomic sciences of natural history, here used for positive comparative purposes. As research proceeded, attention was also given to the relations of chemistry to the broader epistemic/epistemological communities within which practitioners were often embedded (philosophers, economists, traders, manufacturers, travelers), to gauge (for instance) senses in which observation may have been sensitive to divisions or transitions between academic chemistry and the practical arts. Finally, and more philosophically, observation was considered logically (i.e., semantically rather than epistemologically), posing an informal model of objects of observation as "intentional objects," which can generate generate a staged account inclusive of marking, noticing etc. as part of the constitution of such an object, but which may have crucial and informative limitations with respect to e.g., accidental discovery, observational surprise, and the like.