98 BOTANICAL GAZETTE [AUGUST 
Four of the five fruits consisted of an ovary surmounted by a 
sessile stigma. The ovary was not completely developed, in any 
case only a single carpel probably being represented. Each of the 
four larger inclosed fruits produced seeds. This 
may be seen from the figure, one of the fruits 
being placed to one side and turned: over to 
show the incomplete development and the pro- 
Fic. 2—Sketch of duction of ovules. The stigmas, instead of 
small secondary fruit 2 Sa i ‘ 
only partly: visible in being flattened and laciniate, as in normal fruits, 
fig.1; naturalsize. | were capitate, considerably swollen, spongy, 
and -with tuberculate surface. The smallest 
fruit has a very small ovary, without ovules, the capitate stigma 
being borne on an elongated style (fig. 2). The inclosed fruits 
were yellow, being somewhat paler than normal fruits. 
Fic. 3 Fic. 4 
Fics. 3, 4.—Fig. 3, portion of epidermis of secondary fruit showing stomata; 
160; fig. 4, portion of epidermis of normal fruit showing stomata; 160. 
A microscopic examination of the epidermis of these fruits 
(fig. 3) showed it to be made up of cells similar in shape but some- 
what larger than those of the epidermis of normal fruits (fig. 4). 
This similarity extended even to the presence of stomata. The 
guard cells were without chloroplasts. The only evident external 
difference in the epidermis of the inclosed fruits from that of 
normal fruits was in the absence of the coating of wax which the 
latter possesses in a marked degree. In cross-section the structure 
of the inclosed fruits resembles closely that of normal fruits, the 
