Botanical Ecology of the Dry Tortagas. 133 



large Ficus hispida, planted in the courtyard at the light-house on 

 Loggerhead Key, drop their leaves during the coolest months, Decem- 

 ber and January. But this dormant condition lasts only a few weeks; 

 other plants, as the Hibiscus and oleanders, seem to bloom and flourish 

 throughout the year, and during the summer months particularly are 

 covered with showy crimson, scarlet, and pink blossoms. 



Flowering and fruiting. — These phenomena are not definitely limited 

 in the Tortugas region. Coconuts frequently bloom and ripen at the 

 same period, and the same is true of the geiger-trees, some of which have 

 showy clusters of scarlet flowers and bear on the same branch the hard, 

 pointed, white fruits. The SccBvola often bears its white blossoms 

 simultaneously with the large edible-looking but nauseating black 

 berries. Some periodicity, however, has been noted in the red man- 

 grove and the sea-grape, Cocolobis uvifera. Trees were seen with 

 blossoms and ripe or nearly ripe fruits, but never trees with blossoms 

 together with very immature or half-ripe fruits and fully ripened ones. 

 There must be some definite cycle of growth which prevents the occur- 

 rence of fruits at all stages in these trees. 



Herbs, such as Melanthera, Valerianoides, and Alternanthera have 

 buds, blooming inflorescences, and old dead and dry flower-clusters all 

 on the same plant ; other herbs, such as Cakile and some of the grasses, 

 have a well-defined growing-season. Cakile grows only during the 

 winter and spring; by June the plants are full of ripe seeds, and during 

 July the plants die and become tumble-weeds. Tournefortia and 

 Suriana seem to have cycles of growth not separated by long periods, 

 for some individuals are frequently seen having flowers and ripe seeds. 

 The Convolvulacese, represented by Ipomcea and Calomjction, appear 

 to have all stages of flowers and fruits, for even long periods of drought 

 do not seem to prevent Ipomcea pes-caprce from producing its showy 

 rose-purple funnels along the i*unners trailing 20 to 30 feet across the 

 burning white sand-beaches. 



Influence of soil. — There is not much field for observation under this 

 head, since the islands are essentially all ahke geologically — i. e., the 

 soil is all white calcareous sand, but at several places in the Tortugas 

 there is a slight variation in the soil. On Bush Key, near the small 

 tidal pool, the soil is lower than the adjacent sand and various unicellu- 

 lar and a few filamentous algae (e. g., Lyngbya), etc., grow among the 

 mangrove seedlings. In similar situations on Boca Grande the writer 

 observed Sesumum, Alternanthera, and Batis maritima in salt meadows 

 or marshes where young rhizophoras were also estabhshing themselves. 

 However, on Bush Key only an algal flora was observed on the surface 

 of the soil and on the mangrove stems. 



The other variations in soil were due to artificial conditions. On 

 Loggerhead several carloads of soil had been brought from Maple- 

 wood, New Jersey, and this supports a few introduced weeds, such as 



