culling, and when the hot moist weather of May and June is very favorable to development of 

 mildew, the plants are again critically examined, every feeble and. mildewed plant pulled out. 

 Plants which resist mildew at this stage will rarely, if ever, be injuriously attacked when older. 



When the plants are five or six inches high, they are staked, as the gardener stakes peas which 

 climb, and none allowed to sprawl on the ground. By thus carefully cultivating and training 

 the first year, they generally fruit a year sooner than if not staked and well cultivated. 



Testing, Noting and Selecting Varieties for Introduction 



By using good fresh land, well prepared, and planting the selected one-year seedlings in 

 November, cut back to three or four buds, four feet apart between plants and rows eight feet 

 apart, and trellising at once, or before growth starts in spring, so as to train onto the trellis the 

 first season after setting, and allowing only a single shoot to grow up, the writer has succeeded 

 in getting about three-fourths of the vines to fruit the second season after transplanting, or third 

 season after sowing the seeds. The remainder fruit the next year. As ordinarily grown, it re- 

 quires one to two years longer than this. This is a big clear gam to one doing much experimental 

 work, and in the course of fifteen or twenty years of extensive experimentation makes a large 

 saving in time and labor. 



While the vines are fruiting the first time, they are very carefully examined and noted in 

 every particular of growth, foliage, season of leafing, flowering, ripening, quality, diseases 

 affecting, leaf- fall; the character of flower, whether staminate, pistillate or hermaphrodite; 

 degree of perfection in setting fruit, persistence to pedicel, etc., are recorded. 



After all the vines that can bear (some will be staminate) have fruited, and those decidedly 

 not worthy further trial are grafted to the most promising (it will be found the great majority 

 are not worth trying further), then comes the long, careful test of three to five years, to determine 

 the complete character of the few that turn out worthy of recommendation for general trial over 

 the country. Their record in all points should;be high, approaching the "ideal vine" in character. 

 Several vines of each of those of high merit atre planted in various different soils and situations 

 and fruited several years before they are disseminated. 



Time Required for the Full Development of a Variety 



On an average, it requires about eight years from germination, under good treatment, for 

 a variety to fully display its true permanent character, hence should not be disseminated until 

 that age, unless its parentage is exceptionally good, and itself every way satisfactory. 



Oftener there is improvement rather than deterioration in a young variety, but sometimes 

 a variety changes from good promise to worthlessness. 



Percentage of Meritorious Varieties 



Out of over 75,000 grape seedlings, chiefly hybrids, having been subject to culling and 

 selecting by the writer, not over 100 have passed meritoriously through the entire period of 

 testing and been considered worthy of recommendation for general trial by planters. 



Hence it may be safely estimated that one really good variety to every 1,000 carefully selected 

 and hybridized grape seeds will be the originator's reward. 



However, this is a ratio obtained by starting with much new material — untried combinations 

 of new species in a large degree. By now using only the best of the varieties tested and new ones 

 produced, a much greater percentage of worthy varieties should be obtained. 



Some lots of hybrids, especially in my later work, have yielded as high as one or two per 

 cent of good varieties, and as the work progresses farther and farther toward thorobreds, the 

 greater and greater will be the percentage of valuable varieties; providing all the laws of adapta- 

 tion, congeniality, health, etc., and fine judgment in selecting and making cornbinations, are 

 employed. 



—136— 



