HE SEARCH FOR the hematopoietic stem cell (HSC) T moved from theory and argument to practice beginning with the studies of Till and McCulloch in 1961.'As outlined in that now classic study, a population of clonogenic hematopoietic cells capable of landing in and responding to the hematopoietic microenvironment of the irradiated mouse spleen can be defined quantitatively. The outcome of the interaction between the clonogenic cell and the microenvironment appears to be the generation of several million cells in the myelomonocytic, megakaryocytic, and erythroid lineages, as well as the renewal of more clonogenic spleen colony-forming cells.'It was later demonstrated that lymphoid cells could be included within the progeny from a single clonogenic prec~ rsor.~ The collection of cells that formed in a spleen colony was also shown to be sufficient to restore lethally irradiated animals, thereby preventing hematopoietic death.** These characteristics form the basis of a definition for HSCs that has guided many attempts to isolate them4-'That is, pluripotent HSCs ought to be capable of self-renewal and of multilineage differentiation, and be able to save lethally irradiated animals. Of course, this definition defines whole bone marrow as well, and therefore the caveat that the degree to which that cell is represented in the bone marrow is the degree to which its activities in assays for stem cells should be enriched must be added to the definition. development of an in vitro colony assay followed a few years later. 8 While early efforts to enrich primitive hematopoietic cells relied on these assays, the limitations of the assays in reflecting the self-renewal potential of HSC were recognized early on. Velocity and equilibrium centrifugation studies demonstrated that bone marrow cells with an inherent capacity to produce splenic colonies could be separated from those capable of in vitro colony formation?" and that self-renewal potential, as indicated by the content of splenic colony-forming units (CFU-S) within individual spleen colonies, could be separated from the bulk of the CFU-S in normal bone marrow." These observations indicated that CFU-S activity might not directly correlate with self-renewal potential, a problem that has not been conclusively resolved to this date. However, the early separation studies demonstrated that cell separation methods could indeed be applied to dissect the hierarchy of primitive hematopoietic cells. Specific antisera raised against cell surface molecules added a much more discriminating level of identification of hematopoietic cells at all stages of development, and the advent of the fluorescence-activated cell sorter" allowed cell populations identified by fluorescein-conjugated antisera to be separated from other cells and collected for functional testing in hematopoietic assay systems. Our attempts to study early events in hematopoiesis used this technology in combination with multiple assays for clonogenic cells for each of the lineage outcomes that are measurable: CFU-S for myeloerythroid progenitors, thymic colony generation (CFU-T) for thymic progenitors, and clonogenic cells in Whitlock-Witte cultures for B-lineage