232 Gates. 



exact agreement with the expectation of equahty. The Fj of grandi- 

 flora X rubricalyx was much more strongly biennial than the reciprocal, 

 for while in the former cross only 58 plants bloomed out of a total of 

 147 rosettes, in the latter 58 bloomed out of 67. The latter also bloomed 

 earlier in the season, having 50 plants (70%) in bloom by August 16, 

 while only 30 (20%) of the former were in bloom by that date. This 

 difference is significant, showing that as regards the quaUty of ,,ear- 

 liness" both reciprocal crosses were decidedly patroclinous. 



The following facts show that the Fj is intermediate between 

 rubricalyx and its relatives on the one hand and grandiflora races on 

 the other hand, as regards "earliness" or the tendency to pass out 

 of the rosette stage and come rapidly into bloom. Under the same 

 conditions as the two F^ cultures above-described, I grew at St. Louis 

 in 1911, a total of 439 plants belonging to rubricalyx, rubrinervi-', and 

 Lamarckiana. Of these only 19, or a little over 4%, came into bloom, 

 mostly at tlie end of the season. On the other hand I grew cultures 

 of ten different grandiflora race.=; numbering hundreds of plants, all 

 of which under the same conditions, omitted their rosette entirely 

 (as in fig. 7) and formed a stem before they could be planted out. 

 The F2 and back-crosses of th.e F, with the original parents have 

 shown correspondingly intermediate conditions as regards time of 

 flowering, when grown in the cooler climate of England. 



Fig. 14 represents the upper part of a plant of rubricalyx : , grandi- 

 flora, in bloom. The plants were, as far as could be judged, almost 

 exactly intermediate in foliage and pubescence. The leaves, as shown 

 in the figure, exhibited about half the amount of crinkling found in 

 rubricalyx; th.ey were darker green than grandiflora, and in shape were 

 more or less intermediate between the two parents. The characters 

 of the foliage in the mature hybrid plants are extremely difficult to 

 appraise, for in rubricalyx especially, and to a less extent in grandi- 

 flora, there are continuous changes in leaf-characters from the bottom 

 towards the top of the stem. The lowermost cauline leaves are petiolate 

 like those of the rosette. There is a gradual transition by shortening 

 of the petiole and change in shape of the leaf-base, to the bracts 

 which are sessile and anrate. The bracts also continuously diminish 

 in size and change in shape in passing up the stem. The amount of 

 crinkling also usually, but not- invariably, diminishes towards the 

 top of the stem. These conditions make it practically impossible to 

 classify the F2 plants as regards their foliage, for every conceivable 

 intergrade between the two parents occurs. 



