354 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1942 



hanging on with grim determination as though they would not give up 

 until the rocks themselves disintegrate. Every narrow canyon, dark 

 and damp, harbored bromeliads of the more delicate type, while above, 

 braving wind, sun, cold, and heat, were the xerophytic ones. Villa 

 Velha is a botanist's and geologist's paradise. 



THE BEOMELIAD CHARACTERISTICS 



When you enjoy the sweet, juicy fruit of a pineapple, you are eating 

 a bromeliad. Ananas comosus (sativus). When you sink into a soft, 

 well-cushioned automobile seat, the filling responsible for your com- 

 fort may be a bromeliad, Tillandsia usTieoides^ or Spanish moss. In 

 manner of growth these two represent the two extremes : the pineapple 

 is strictly terrestrial, while the Spanish moss is wholly epiphytic, even 

 going so far as to dispense with roots. Between these two extremes, 

 bromeliads exhibit a great variety of plant characteristics. 



It would not be difficult to surmise that these two forms might be 

 the latest development, eacli in its own type of fruiting method — the 

 appendaged-seed type (represented by the Spanish moss) and the 

 berry-seed type (represented by the pineapple). The pineapple has 

 had all its fruits fused into one big "berry." No other fruit-bearing 

 type in this family has the individual berries that hold the seed more 

 completely welded than in the pineapple fruit. On the other hand, 

 the Spanish moss, with its appendaged seeds in a pod, grows with such 

 a fusion of leaves in one continuous growth that there is no evidence 

 of roots (which, I believe, have been absorbed) or of the usual maturing 

 of individual plants, characteristics which probably make it the latest 

 development in the appendaged-seed division. We might say that this 

 "freak" of a plant is certainly the most modern, for it travels entirely 

 by air. 



The bromeliad flower pattern is formed in multiples of three. Its 

 flowers generally are formed in spikes or racemes with brightly colored 

 bracts. In the botanical descriptions of this family every flower has 

 been said to be monoecious or perfect, even though in some of the Hech- 

 tias (of Mexico), the flowers of which have both stamens and pistils, 

 only one of them functions. 



An exception appeared during January 1942, when I observed that 

 in several of the species of Cryptanthus the flowers were not monoe- 

 cious, but dioecious, for there were separate male and female flowers in 

 the same plant. (One of our Grypthanthus species, not yet determined, 

 however, does have all perfect flowers.) This condition of separate 

 sex flowers has apparently not been noted before, as there seems to be 

 no record of it in the literature. 



Flowers throughout the family range in size from tiny, almost micro- 

 scopic blossoms as in the giant HoJienhergia ougusta^ in which the 



