The Mode of Inheritance of Semi-Sterility in the Offspring of Certain Hyhrid Plants. 305 



grains mostly shruuken and anthers which did not dehisce. After many 

 artificial pollinations, one descendant of tliis plant was raised. Tliis 

 was fertile, and had full pollen-grains and deliiscing anthers. Jesenko 

 subsequently repeated the pollinations on a largei- scale, and obtained 

 several first-generation hybrids; but all attempts to fertilize these inter 

 se failed, though a few successful back-crosses were made with wheat 

 pollen. He now (8, p. 316) regards the second-generation plant above- 

 mentioned as the product of such a back cross. Sutton (15) crossed 

 a wild pea from Palestine (Pisum humile) with many varieties and 

 species of cultivated Pisum. In the few crosses which set seeds, the 

 hybrids were usually quite or partially sterile. The few second-gene- 

 ration plants obtained from these hybrids were however fertue. This 

 was confirmed by Bateson (1). Other undoubted cases are known of 

 some of the progeny of nearly sterile hybrids being fertue. 



But if among the descendants of a partially sterile hybrid there 

 are some quite fertile plants, in what proportion are these? This can, 

 sometimes at least, be ascertained by raising several hundred plants in 

 the second filial generation, and determining the abortion of poUen-grains 

 and embryo-sacs with the microscope, or by counting the seeds and the 

 aborted ovules in the seedvessels. 



The Parent Plants. 



The results to be described form a part of an investigation into 

 the improvement of certain Florida forage crops by crossing, which has 

 been in progi-ess for four years. The four parent plants are grown for 

 forage on a large scale in Florida. The first-named plant, in my 

 opinion, belongs to a different species from the other three, whatever 

 limitation of species we accept. 



The Florida Velvet "bean" {Stizolohium deeringianum Bort) of 

 unknown origin (3), is a large ornamental tropical twiner, wliich has 

 been found to be an excellent cattle food. Many thousand acres of it 

 are grown in Florida and other Southern States. It forms very long 

 vines if it has the chance to climb up a tree or a tall pole. I have 

 seen it twine up a string to the top of the Experiment Station building, 

 about forty feet high. When grown on eight-foot poles and prevented 

 from running on the ground, it branches profusely, and forms many 

 racemes of violet-purple flowers and hundreds of pods. The pods (Fig. 1) 

 are about 6V2 cm. long, and are covered with a soft brownish-black 

 "velvet", or wool, of flat and twisted hairs. The seeds (Fig. 1) have 

 a definite pattern of dark mottling or streaking. 



Induktive Äbstammungs- und Vererbungslehre. XII. 22 



