RESUMO
The histochemical studies that search for osmophores differ in the protocol they follow for fixation and discoloration of the samples, and also in the type and number of stains utilized. Despite these differences, all the studies have one point in common: the use of fresh material either collected directly in the field or cultivated in botanical gardens and greenhouses. This is an obvious limitation for the studies of osmophores. Flower parts of herbarium specimens of different dates of collection and different plant families were exposed to variable times and percentages of discoloration agents and under different stains for finding if it is possible to positively test osmophores in this type of material. We obtained positive results discoloring the samples with ethyl alcohol 96° and sodium hypochlorite, and staining with Lugol, Oil Red O, and Neutral Red (TIOFH). A protocol (TIOFH3) for osmophore testing that combines these three stains into a single procedure is proposed.