456 REPRODUCTION. 



1177. Ill reciprocal hybridization tlie pollen of A is effective 

 when applied to the stigma of JB ; and, conversely, the pollen of 

 B is potent when applied to the stigma of A. But it sometimes 

 happens that the rule will not worls both ways. Thus the pollen 

 of Mirabilis longiflora was found by Kolreuter to produce hybrids 

 when applied to the stigma of Mirabilis Jalapa ; but the pollen 

 of the latter was without etfect upon the stigma of the former. 

 Other cases are known, but the cause of this extraordinary 

 preference is not understood. 



1178. Hybrids are produced artificially by the transfer of 

 pollen from one species to the stigma of another species, care 

 being talsen to exclude all pollen of the second species from its 

 own stigma. The pollen is best transferred by means of a sable 

 or camel's hair pencil.' Exclusion of the pollen of the flower 

 to be fertilized must be secured by removal of the anthers before 

 the flower opens. This is easily effected by tlie use of delicate 

 forceps, an incision being carefully made in the side of the 

 corolla. After the application of the pollen to the stigma, the 

 plant or blossom must be covered b}- some close netting. 



1179. Following the application of the pollen, changes take 

 place in the fertilized flower. But, as Nageli has pointed out, 

 these changes in many cases fall far short of yielding satisfac- 

 tory results to the experimenter. Nageli describes several grades 

 of partial fertilization: (1) that in which the ovary, and per- 

 liaps the persistent calyx, grows somewhat without appreciably 

 affecting the ovules ; (2) that marked by greater growth of the 

 ovary, and by slight enlargement of the ovules, which after- 

 wards shrivel up ; (3) that with small imperfect fruits with 

 empty seeds ; (4) that having good fruits with empty seeds ; 

 (5) that with normal fruits with apparently perfect seeds which 

 have no germs ; (6) that producing good fruits with seeds which 

 have only minute germs incapable of further development. 



In successful fertilizatiou there are produced good fruits hold 

 ing sound seeds. 



Some of the cases iu whicli hybrids have been produced between the 

 species of different genera are given by Nageli (Sitz. d. k. Akad. d. Wiss. z. 

 Miinchen, 1865 and 1866), as follows : Rhododendron and Azalea, Rhododen- 

 dron and Rhodora, Rhodora and Azalea, Rhododendron and Kalniia. Of those 

 above mentioned, Rhododendron, Rhodora, and Azalea are now placed by 

 Bentham and Hooker in a single genus, — Rhododendron. 



1 It is of great importance that the pollen should be applied at exactly 

 the proper period for impregnation. This is usually indicated by the moisturs 

 of the stigmatic surface. 



