446 Agricultural Experiment Station, Ithaca, N. Y. 



purple apex became grsbx, then yeIloA\ish. like the staminate parent, 

 while the dark purple bodj' of the finiit became dull green. The 

 full extent of variation in this series is showTi in the illustratioiL 

 Eight of these fruits grown in 1890 were selected or again crossed 

 for planting in 1891. These B lots originated as follows: 



Bl. Pollinated by Giant Purple (original pistillate parent). 



B2. Pollinated with another flower on the same plant. 



B3. Pollinated by Griant Puiple, as in Bl. 



B4. Pollinated with another flower on same plant, as B2. 



B5. Pollinated by Giant Purple, as in Bl and B3. 



B6. Selections, not artificially pollinated. These were attract- 

 ive fruits of a purple color and lighter apex, tending to be striped. 



B7. Selections, not artificially jwllinated. These were attract- 

 ive fruits of a pui*ple color and lighter apex, tending to be striped. 



B8. Selections, not artificially pollinated. These were attract- 

 ive fruits of a purple color and lightxT apex, tending to be striped. 



The offspring of these fruits — 479 |»lants — .shoAved a. wide vari- 

 ation in color of herbage, many of theui being green, although the 

 greater part of them were pui-ple. As a whole, however, these 

 plants were comparatively unifonn in size and habit, and should be 

 distinguished from the A and C series at a considerable distance. 

 The plants were low and bushy, but erect, mostly with a grayish- 

 purple tinge when seen in mass. B2, whicn was pollinated from 

 another flower on the same plant, gave a:^ various fruits as the 

 others. Of eighteen large fruits of it, one was very black-purr J e, 

 six were light purple, and one was purple-striped, nine were gi-etn- 

 and white striped, one was pure white, [t is strange, too, that 

 one of the plants of this lot had the peculiar habit of Early Dwarf 

 i'urple, a variation also found in one of tlie A series (A5) which 

 had been similarly pollinated. The lots into which tlu^ original 

 pistillate parent, Giant Pui*ple, was agam impressed — Bl, B3, 

 Bfi — showed very variable offspring, although there were large 

 numbers of purple and purple-striped fruits in the progeny. B3 

 and B5 were very much alike in habit, color and iiuit. Tuere 

 was comparatively little variation in any of the B lots. 



Series C originatcMi from a cross of I-ong AVhite by Black Pekin. 

 This series, then, is much like A in parentage, except that the 



