No. 557] PARTHENOGENESIS IN NICOTIAN A 293 



Whether the pollen tubes in these experiments reached 

 the ovules is not known, but probably not, since no fertile 

 seed was produced. The production of numerous seed 

 normal in appearance indicates, however, either that the 

 pollen tubes must have stimulated the nucellus tissue in 

 some way, or that normal seed development was started 

 but not finished, for no seed of any kind was produced in 

 the decapitated blossoms where pollen grains were not 

 applied. 



The total abortive seed produced by the pollination of 

 the decapitated styles included two from N. tabacum 

 where the stubs were covered with 50 per cent, cane 

 sugar solution and self-pollinated, four from the same 

 species where the stubs were covered with stigmatic fluid 

 and self-pollinated, twenty-seven from A r . paniculata 

 where the stubs were covered with stigmatic fluid and 

 self-pollinated, and fourteen from N. tabacum where 

 N. Forgetiana pollen and no fluid was applied. 



In connection with the decapitation experiments, an 

 experiment on the grafting of pistils 8 was conducted. 

 One hypothesis for the non-crossing of certain species r 

 as has already been mentioned, is the extraordinary 

 length of the style. By removing a portion of the style 

 and grafting the stigma end of a pistil of either the same 

 or another species to the stub, the style was shortened 

 from one to one and a half inches. Immediately after 

 grafting, the stigmas were pollinated. From one of the 

 five grafted N. tabacum blossoms was produced one abor- 

 tive seed. The development of this one seed may or may 

 not have been due to the penetration of one or more 

 pollen tubes, as in the cases where pollen grains were 

 applied directly to the decapitated pistils. 



"The grafting technique is simple, nevertheless, the operation is difficult, 



