so successfully that a dozen or more of his varieties are still cultivated. All are characterized by great vigor and productiveness, and, though they lack the qualities which make good table grapes, they are among the best for wine-making. Rommel has had many followers in hybridizing the native species, chief of whom is Mr. T. V. Munson, Denison, Texas, who has literally made every combination of grapes possible, grown thousands of seedlings, and produced many valuable varieties.
The aim of hybridization in breeding plants is to combine the desirable and eliminate the undesirable characters of varieties or species in a new race. A plant, however, is such a complex sum-total of characters that no one can pi edict with any certainty the result of mingling the characters of two more or less distinct plants. Speculation thus quickens the charm of hybridization. The progeny of crossed grapes is always chaotic and must be passed through the sieve of selection, the meshes of which have grown larger and larger with use until now out of thousands of new forms a grape-breeder will retain few indeed.
Within the last decade, hybridizing has received a great impetus through the publication of Mendel's experiments. In the past hybridization has been a maze in which breeders lost themselves. Mendel's discovery in heredity assures a regularity of averages and gives a definiteness and constancy of action hereto wholly unknown in hybridization. It now appears that many of the characters of grapes follow
