Nucellar embryos are produced asexually, from somatic cells of the nucellus. Usually, therefore, nucellar seedlings reproduce the genotype of the seed parent. Nongenetic differences from the seed parent occur as transient juvenile characteristics, and also sometimes result from elimination of disease viruses. Budded descendants of nucellar seedlings of one seed-parent Satsuma tree have been studied in two propagated generations, in eighteen crop years. Trees of the first generation, budded directly from different nucellar seedlings, commonly differed so consistently in fruit shape as to indicate considerable probability of genetic diversity. From two of these budded trees, which differed in fruit shape and also in leaf size, trees were budded for small plots. These plots differed significantly in tree form and leaf size, and in yield, shape, and time of coloring of fruit. The tree types of these two nucellar-line strains were identified repeatedly among sister nucellar seedlings. Tree characters and fruit shape have indicated that at least one or two other genetic types probably were present among nucellar seedlings of the same seed-parent tree. That tree probably either included several genetic types, or produced genetic variations during formation of nucellar embryos. One seedling type, probably of nucellar origin, which occurred only once among several hundred seedlings, differed conspicuously in tree and fruit. A plot of old-line trees, budded directly from the seed parent of the nucellar seedlings, has been inferior to both nucellar-line plots in size of trees and leaves, in yield and time of coloring of fruit, and in percentage of soluble solids and solids/acid ratio of juice. The greater yield of the nucellar lines has been due mainly to their greater tree size. Earliness of fruit coloring has shown marked positive correlation with yield. The better of the two nucellar lines of the plots (strain A) is recommended for trial in climatic conditions favorable to the Satsuma, especially in some navel orange districts. Observation and tests have indicated absence of the psorosis and tristeza viruses from both seed-parent and nucellar lines.