Publicity regarding the loss of existing genetic resources because of human activity in the centers of plant, origin and diversity has stimulated efforts to collect and maintain germplasm in international genebank networks (Poopathy, 1986). The problem is particularly acute for Citrus and related genera because government-sponsored land development programs are destroying the natural habitat of many Citrus spp., and relatives. The International Board of Plant Genetic Resources (IBPGR)[with liaison to the United Nation’s Food and Agricultural Organization (FAO)] is negotiating with the Univ. of Malaysia to set up a genebank network for Citrus (Poopathy, 1986). Justification and impetus for this program may be provided by the abundance of economically important traits (eg, tolerance to biotic and abiotic stress, horticultural! characteristics and performance, etc.) that can be found among Citrus spp. and related genera. For such programs to be successful, however, technology must be available to maximize the use of the collected germplasm.

The use of somatic hybridization to obtain wild germplasm for woody fruit tree improvement is discussed in this article, using Citrus as a model. Recovery of intergeneric somatic hybrid plants produced between sexually incompatible Citrus sinensis L. Osb. CV. Hamlin (an early sweet orange) and the wild relative Citropsis gilletiana Swing. and Ken.(Gillet’s cherry orange), a genus native to Africa, is described. Methodology used to produce these hybrids is outlined. Hybridity was confirmed by analysis of morphology, somatic cell chromosome number, and phosphohexose isomerase (PHI; EC 5.3. 1.9) isozyme banding patterns obtained by electrophoresis on horizontal starch gels. This hybrid and other intergeneric somatic hybrids that we have produced are being fieldevaluated for citrus rootstock potential. In-