274 University of California Publications in Botany [Vol. 5 



before taking up the particular phase of the problem which is the 

 subject of this report. The various F^ hybrids between N. Tahacum 

 varieties and N. sylvestris are replicas on a large scale of the N. 

 Tahacum parent concerned in the particular cross. The dominance 

 of the many distinct vegetative and floral characters w^hich are summed 

 up in the general appearance of the N. Tahacimi parent is, for each 

 character, practically complete, w'hile the dominant nature of any 

 given parental character is made more striking through the general 

 increase in expression due to heterosis. The F^ species hybrids are, 

 then, distinguished from the N. Tahacum parent almost solely in an 

 increased expression of all characters peculiar to it and also by the 

 possession of a short-lived perennial habit which is characteristic of 

 the N. sylvestris parent. Their flowers are normal in every respect, 

 with the important exception that the open anthers contain a relatively 

 small amount of light, dry, almost entirely functionless pollen as con- 

 trasted with the mass of heavy, more or less sticky pollen which is 

 produced by the flowers of their parents. Almost every grain of the 

 parental pollen will germinate in its own stigmatic fluid. Examination 

 of the Fj hybrid pollen made at the start of the experiment and often 

 repeated thereafter showed that a very small percentage of what 

 appeared to be normally matured grains occurs in the mass of shriveled, 

 degenerated and undeveloped grains (cf. plate 36). These latter 

 correspond in appearance to those which have often been flgured as 

 characteristic of the pollen of sterile plants. 



One of the most striking peculiarities of the hybrid flowers is their 

 tendency to fall soon after anthesis. A flower under bag on a hybrid 

 plant will, in general, fall at approximately the same time that a pro- 

 tected, castrated flower on one of the parent plants will fall. The only 

 capsules remaining at the end of the season, and they are relatively 

 very few in number, contain a little viable seed. Similarly, if the 

 pollen of the corresponding parents is used to pollinate the F^ flowers 

 the latter usually persist and in all such cases a little viable seed 

 is formed. Again, in the case of a number of other attempted inter- 

 specific crosses in Nicotiana we have found that the foreign pollen, 

 although in some cases it actually germinates in the stigmatic fluid 

 of the female parent of the attempted cross, will not inhibit the falling 

 of the castrated flower. Lastly, the case of the F^ hybrids betAveen 

 N. sylvestris and N. Tahacum "Cuba" deserves mention in this con- 

 nection. N. Tahacum "Cuba" has been shown to exhibit a rather 

 striking parthenocarpy, holding its flowers after castration and 



