344 



BOTANICAL GAZETTE 



[NOVEMBER 



some nonnal plants, two mutant hybrids, and will give the constitu- 

 tion of their progeny, together with that of two normal individuals 

 of 0. gigas, in table V. The numbers of seedlings have been very 

 small in this case, owing to the small degree of fertiUty of the pollen 



of O. gigas nanella. 



TABLE V 



Dwarfs in the first generation of 0. gigasXO. gigas nanella 



The first two seed-bearers had evidently about one-half of their 

 egg cells mutated into nanella, which by the fertilization with the 

 pollen of dwarfs must, all of them, become nanella specimens. The 

 two last-named plants, although externally not differing from the 

 others, had only very few mutated sexual cells, and therefore pro- 

 duced only about 3 per cent of dwarfs. 



TABLE VI 

 Dwarfs in the second generation of 0. gigasXO. gigas mut. 



nanella 



The experiment showed at the same time that hybrids between 

 0. gigas and O. gigas nanella have the features and the stature of the 

 former type, and thereby justified the assumption made above in 

 the explanation of the behavior of mutant hybrids. 



I made the reciprocal cross in the same year, fertilizing some 

 dwarfs of my race by the pollen of normal plants of 0. gigas. The 



