436 cook: pomegranate flowers 



are still completely covered by the infolded petals, an arrangement 

 obviously favorable to cross-fertilization. 



In the other type of flowers, with the narrow tapering base, 

 shown in figure 2, the ovaries are poorly developed and have 

 only minute rudimentary ovules that degenerate and shrivel, 

 sometimes even before the flowers open, leaving the ovarial cham- 

 bers apparently empty. A corresponding reduction appears in 

 the styles, which are often less than half as long as in the per- 

 fect flowers and remain hidden under the stamens. This arrange- 

 ment might be considered as favorable for self-fertilization, were 

 it not that the short-styled flowers have lost all but the staminal 

 functions, as shown by the reduced ovaries and abortive ovules. 

 In the staminate flowers the whole surface of the ovary becomes 

 bright red like the upper part of the calyx tube, but the perfect 

 flowers often fail to show the bright color on the ovary. 



The second or staminate form of flowers was much more 

 numerous on most of the bushes and some of them had none of 

 the perfect flowers, so that no fruit could be set. Thus the pome- 

 granate may be considered as a polygamodioecious plant, to the 

 extent that a large proportion of the flowers no longer produce 

 functional pistils. On the other hand, the perfect flowers show 

 no apparent tendency to lose the staminal function. It is as 

 tho a simple dimorphism of short and long styles had been 

 followed by a further reduction of the pistils and ovules of. the 

 shortstyled flowers, until the reproductive functions were lost. 



The existence of ornamental varieties of the pomegranate that 

 produce no fruits, but are propagated entirely from cuttings, 

 may be taken to indicate a further dioecious tendency. In vari- 

 eties that have double flowers sterility might be due to loss of 

 function in the stamens, but it now appears that absence of per- 

 fect flowers with functional pistils may be another cause of un- 

 fruitfulness. At Bard, California, where these observations were 

 made, several of the bushes with ordinary single corollas had only 

 the staminate form of flowers and buds. But as this might not 

 be true in another season, the extent of the dioecious tendency 

 remains in doubt. The bushes were flowering abundantly at 

 the middle of May, 1912, and the beauty of the floral display 



