A good seed set was obtained from crosses of Medicago glomerata Balb. with M. sativa L. There was no evidence of a preferential chromosome pairing effect on segregation for floral anthocyanin. From an analysis of different M. glomerata × M. sativa crosses four factors were found to be involved in anthocyanin production. About 90% of the cells showed normal chromosome behavior during meiosis. Taxonomically, M. glomerata stands somewhat apart from the M. sativa — falcata complex, as indicated by a lower seed set from intercrossed F1's, and by somewhat higher numbers of meiotic irregularities in the F1 plants. With these and other morphological differences M. glomerata is considered as a separate species. When hybridized with M. prostrata Jacq., M. glomerata responded in a similar manner to that of M. sativa. The weight of seeds produced by M. glomerata plants was approximately the same, whether fertilized by M. glomerata or by M. sativa pollen. There is a somewhat greater affinity between M. glomerata and M. prostrata than between M. sativa and M. prostrate; M. prostrata plants fertilized by M. glomerata pollen had better seed development than those fertilized by M. sativa.