42 MAINE STATE COLLEGE 



brought under observation, it was noticed that while the fruits 

 were abundant and well filled out, the\' began to wither when about 

 half ripe and not one produced perfect seed.* 



The instance of the mulberry is conflnned by Claypolc who 

 cites| a case within his own observation in which a pistillate tree 

 bears fruit abundantly every year though no staminate tree is in 

 the vicinity, and no staminate flowers have been found on the tree 

 itself. The "seeds" in these fruits, as in the other instances, con- 

 tain no embryos. Whether this is a case of development in the 

 entire absence of pollen, as circumstances would indicate, or 

 whether there may have been a limited supply of pollen at hand, it 

 is evident that the ovaries developed independently of any action 

 on the ovules. 



Dr. Masters is authority for the statement that certain varieties 

 of pears habitually produce seedless and eoreless fruit. J In the 

 same way it is not uncommon to find the capsules of many herba- 

 ceous plants fully developed while the seeds are absent. M. Jean 

 Sisley, a well known French horticulturist, found this to ocear 

 with great frequency in case of the geraniums and pelargoniums. 

 Of one hundred flowers of Geranium platypetalum artificially pol- 

 linated, not one produced perfect seeds, and of a large number of 

 capsules sent by another party, nearly all were without seeds. § 



Naudiu, as a result of his studies of the genus Cucurbita, sug- 

 gested the possibility of a specific effect of pollen in exciting 

 growth of the ovary ; and this theory is supported by Focke, who 

 says: "Pollen has two actions on the female organs, one on the 

 seeds, and one in exciting the growth of the fruit."! The theory 

 seems plausible, and in view of the many examples of well devel- 

 oped but empty seed pods, it would seem that the stimulating 

 action is alone exerted in some instances. These examples are 

 specially common among peas and beans. 



The accompanying photograph, Figure 4, represents the natural 

 size of a Lima bean which failed to develop seeds — the undevel- 

 oped ovules may be seen at the right. Similar instances are very 

 common in all varieties of this class. 



*B. J. Camerarii Opuscula Botanici Argumenti, cited by Sachs, Hist, of Bot. 386. 



fRep. U. S. Dept. of Ag. 18S7, 318. 



JNature, XXXV, 12. (Nov. 4, 1886.) 



§Garcl. Cliron. N. S. IV, 654. 



IIFocke, Die Pflanzen roisclilinge, 447. 



