But he successfully created interspecific hybrids by crossing Asimina obovata, A. longifolia, A. incana, and A. reticulata with A. triloba. The A. triloba × A. obovata hybrids appeared fertile. Unfortunately, Zimmerman died in 1941 before his other crosses matured. Only a tiny portion of his varietal collection was donated by his widow to the Blandy Experimental Farm: eight interspecific Asimina hybrids and four controlled crosses of ‘Ketter’, ‘Buckman’ and ‘Taylor’ (Flory, 1958).

The Blandy Experimental Farm collection, Boyce, Va., also originated with the contest of 1916 (Fig. 1). Orland E. White, the director from 1927 to 1955, was interested in pawpaws and conducted a collection effort that lasted his whole tenure. In 1926, he grew a seedling of ‘Fairchild’ (named ‘Fairchild No. 2’) and seedlings from other superior trees identified in the 1916 contest. The exact number of pawpaw accessions at Blandy is uncertain, due to problems of record maintenance. For nearly 30 years the staff at Blandy accessed pawpaw seed from diverse areas of Alabama, Indiana, Kentucky, Maryland, New York, and Tennessee (Flory, 1958). Flory lists more than 39 accessions at Blandy. In 1980, five trees which were the remnant of a once much longer row still grew near the headquarters; and a major planting existed in the back woods consisting of two parallel rows of evenly spaced trees that must have originally contained more than 60 trees (Peterson, 1986).

John W. Hershey, a nurseryman living in Downingtown, Pa., bears mention although no known varieties resulted from his work (Fig. 1). Hershey sold seedlings of ‘Fairchild’ crossed with various trees from Zimmerman’s collection (whether seedlings or grafts is not known) via a mail-order catalog. Hershey worked with pawpaw roughly from 1925 to 1958. Material from Hershey was later incorporated into Peterson’s breeding program.

During this period, several selections were made that bear no relationship to the 1916 contest. Ernest J. Downing and his father were great lovers of pawpaws and spent many Sundays on horseback searching for superior pawpaw trees (E. Downing, personal communication) (Fig. 1). Their Sunday travels eventually resulted in the discovery of ‘Middletown’ (1915) and ‘Mason-WLW’ (1938) which they propagated onto their fruit farm near New Madison, Ohio. They regarded ‘Mason-WLW’ as the better of the two. In 1934, Homer Jacobs of the Holden Arboretum, Kirtland, Ohio, brought back seed from a West Virginia tree (Fig. 1). From these seeds Jacobs selected a superior seedling, circa 1945, which he named ‘Sweet Alice’ (C. Tubesing, personal communication). Finally, H. A. Allard of Arlington, Va., who was an entomologist with the USDA-Agricultural Research Service, collected superior clones from the highlands of Virginia, circa 1930, and engaged in breeding on the Arlington Farm of ARS and then on his own farm in the 1940s and early 1950s (Allard, 1955) (Fig. 1). No named varieties came from Allard’s work, but germplasm from his work became part of Peterson’s collection.

New Madison, Ohio. They regarded ‘Mason-WLW’ as the better of the two. In 1934, Homer Jacobs of the Holden Arboretum, Kirtland, Ohio, brought back seed from a West Virginia tree (Fig. 1). From these seeds Jacobs selected a superior seedling, circa 1945, which he named ‘Sweet Alice’ (C. Tubesing, personal communication). Finally, H. A. Allard of Arlington, Va., who was an entomologist with the USDA-Agricultural Research Service, collected superior clones from the highlands of Virginia, circa 1930, and engaged in breeding on the Arlington Farm of ARS and then on his own farm in the 1940s and early 1950s (Allard, 1955) (Fig. 1). No named varieties came from Allard’s work, but germplasm from his work became part of Peterson’s collection.

In summary, the course of pawpaw variety development between 1900 and 1950 included widespread selection from the wild and resulted in numerous varieties. Except for Zimmerman, few breeders were active, and no recorded varieties were developed through breeding except ‘Fairchild’ and ‘Fairchild No. 2’. Zimmerman apparently compared and evaluated his collection but published only short descriptions of limited utility. The preservation of germplasm was not a priority and ultimately all varieties but ‘Middletown’ and ‘Sweet Alice’ were lost.

1950 to 1985: A BREAK IN CONTINUITY, FRESH SELECTIONS FROM THE WILD, AND NEW BREEDING. Interest in the pawpaw grew in the years between 1950 and 1985, nurtured by the enthusiasm of a few key individuals in the Northern Nut Growers Association (NNGA; Townsend, Del.). Although the primary mission of the NNGA was to advance the cause of temperate nut trees, the organization sustained a strong interest in minor native fruit. The great majority of breeders and explorers of pawpaws have been NNGA members, including Zimmerman, Gibson, Davis, Gordon, Campbell, Peterson, Callaway, and Pomper.

1985, nurtured by the enthusiasm of a few key individuals in the Northern Nut Growers Association (NNGA; Townsend, Del.). Although the primary mission of the NNGA was to advance the cause of temperate nut trees, the organization sustained a strong interest in minor native fruit. The great majority of breeders and explorers of pawpaws have been NNGA members, including Zimmerman, Gibson, Davis, Gordon, Campbell, Peterson, Callaway, and Pomper.

Unfortunately for future pawpaw breeding, the collections of Zimmerman and Blandy were abandoned and neglected after Zimmerman’s death and White’s retirement. Of the many articles published during this second period, only a few authors (McKay, McDaniel, and Slate) made reference to Zimmerman or Blandy. The major articles about variety development during this time by Ward, Glaser, Gibson, and Davis do not mention either Zimmerman or Blandy.

The challenge of assembling superior germplasm and conducting breeding began afresh. In the next 35 years, an influx of superior material from the wild was collected. In 1950, Ward selected ‘Overleese’ from the Overleese property in Rushville, Ind. (Pape, 1965) (Fig. 2). This variety is now a standard for pawpaw varieties because of its fruit size, flavor, fleshiness, vintage and availability. Although Pape’s account supposes that ‘Overleese’ is new genetic material from the wild, this assumption may not be