A complete description of a plant community must include the buried viable seeds in the soil. The plants occurring in this form are a part of the flora, which helps to determine the community, even though they are not readily evident. The importance of defining an ecosystem's flora is reviewed. When the soil's buried viable seed population is used to help determine the flora, various problems arise since perennial plants reproduce much less abundantly by seed than do annuals, laboratory germination conditions probably do not suit all species, and seeds are not distributed at random so sampling is made difficult. In an investigation of buried viable seeds in two grazed and ungrazed California bunchgrass sites poor correspondence between vegetation and soil seed populations was found. Numbers were 8000 to 12,000 seeds per m2. Figures of this order or magnitude are general in the periodical literature which describes the buried viable seed populations and the factors which influence them qualitatively and quantitatively in various kinds of arable, mesic pastured or mown, steppe, and forest plant communities. Not only are the plants evident above ground in a stand of vegetation part of the ecosystem which includes that vegetation, so also are the plant disseminules in the soil. This paper records a determination of the kinds and numbers of viable seeds in two stands ofStipa pulchra, bunchgrass vegetation in the eastern foothills of the central Coast Ranges of California. Such vegetation is variable from year to year, over a climatic cycle of years, and successionally (i.e.), it is seral and therefore changes with time, other habitat factors being constant (Heady 1958). By no means all the species which can occur in this vegetation over a period of several years are evident above ground at one time. We have attempted to use the soil content of buried, viable seeds in the two stands as a measure of this potential flora.