256 



Gates. 



of crinkling in the upper bracts. Fig. 21 is from (ruhricalyx x grandi- 



flora) ;■: ruhricalyx, (No. 70. III. i), a cross which yields offspring much 



nearer ruhricalyx than grandiflora, though never identical with the 



former except perhaps in a few cases. The plant in fig. 21 was recorded 



be 

 as — • — , but its whole appearance is very similar to ruhricalyx (cf. 

 24 



fig. 10). Figs. 22 and 23 are from (ruhricalyx x grandiflora ) x grandi- 

 flora, giving offspring which are midway between these parents. In 

 fig. 22 (plant No. 71. IV. 3) photographed July 23, buds had not yet 

 begun to appear, but the foliage closely resembles grandiflora (cf. 

 fig. 4) and was recorded (Aug. 19) as fs. Fig. 23 (plant No. 71. V. i), 

 taken Aug. 27, shows similar foliage (recorded as ps Aug. 19), but the 

 later leaves have indications of crinkling. The great majority from 

 this cross are midway between ruhricalyx x grandiflora and pure 

 grandiflora. 



When the Fi from grandiflora x ruhricalyx or its reciprocal is 

 crossed back with grandiflora (Cults. 67, 68, 69, 71, 73, 74, 78) a new 

 blended type of foliage appears, (figs. 22, 23) which is about three- 

 quarters grandiflora and one-quarter ruhricalyx, although these plants 

 again are by no means uniform. Moreover, in ps plants the upper 

 bracts usually begin to be broader at base and somewhat crinkled, 

 showing the ruhricalyx influence. And in several of these families, 

 notably No. 78, the leaves on many plants bear large yellow blotches. 

 It is not obvious what the significance of this conspicuous feature may be. 



In the same manner, when ruhricalyx x grandiflora is crossed back 

 with ruhricalyx (Cult. 70, fig. 21), the majority of the offspring were 

 intermediate between the F^ hybrid and ruhricalyx. Six plants were 

 almost identical with the latter (he), though one was nearly ps. Blending 

 and variation are the only conceptions applicable to these results. 

 The idea of fixed units undergoing segregation and recombination 

 completely breaks down. 



I. Inheritance of time of flowering. 



I have referred elsewhere (p. 216 and Gates, 1913 c) to the striking 

 physiological differences between grandiflora and ruhricalyx. These 

 differences also are inherited in intermediate or blended fashion. Wlien 

 grown in this climate as an annual, grandiflora differs from ruhricalyx 

 in two respects; (i) the rosette stage is more evanescent and stem- 

 formation begins earlier, (2) blooming is delayed until much later. 

 These differences come out in a peculiar way in some of the hybrids. 



