Apr. 5 ,1924 Spiny-Fruited Gooseberry from Florida 73 
In the new species the sepals when reflexed reach to the summit or sometimes to 
the base of the ovary, in curvata far beyond its base. 
On the pollination of Grossularia echinella only a few observations were made. 
Isolated plants did not set as much fruit as plants growing in groups and there¬ 
fore better situated for cross-pollination. Only a single insect was seen pollin¬ 
ating the flowers, a carpenter bee, Xylocopa virginica. To suck the nectar it 
alighted in an inverted position on the pendent flower, grasping the stamens 
and style with all its legs in such a manner that the anthers and stigmas were 
brought into contact with the middle of the underside of its body, a very effec¬ 
tive position for the transfer of pollen from one flower to another and from 
one plant to another. The bee assumed its inverted perch instantly, without 
hesitation and without slipping, drew the nectar rapidly, and proceeded 
promptly to another flower. In maintaining a firm grip on its perch it must 
have been greatly aided by the long hairs on the filaments of the flower, a 
characteristic very unusual in the genus Grossularia, and occurring in no other 
of the more than 40 American species except curvata and nivea. Both of these, 
like echinella , have long protruding stamens which, if devoid of hairs, would 
furnish only a slippery hold to a pollinating insect. 
The cup-shaped glands at the tips of the coarse hairs that densely clothe the 
ovary were not yet yielding an exudate either in the newly opened flowers or in 
the older withered flowers in which the hairs had begun their elongation into 
prickles, a condition that suggested, at the time of the first observation of the 
inflorescence, that the function of the glands was related to the fruit, not to the 
flower. This suggestion was confirmed by the observations made on March 27 
and by those made by Doctor Kurz on April 7 and later, when some of the berries 
were half mature. As the fruit enlarges, the glands at the ends of the prickles 
give off a sticky and unpalatable secretion. The species presents therefore 
the apparent anomaly of a sweet and succulent berry well adapted to the dis¬ 
persal of its seeds by fruit-eating animals, yet barred from this means of dis¬ 
persal by its spiny covering and offensive exudate. A little consideration, 
however, discloses the fact that these obstacles, though effective against such 
animals as insects, rabbits, and squirrels, would present only an interesting 
conundrum to such wide-awake and investigative fruit-eating animals as 
mocking birds, catbirds, and thrushes, whose long bills would enable them to 
open the berries with ease, and whose digestive limitations would insure a wide 
distribution of the living seeds. The structures that appear at first, there¬ 
fore, as a bar against seed dispersal provide in all probability a very special 
and very effective means of advancing the distribution of the species. 
At the present time the new species is known only from the type locality, 
on the north side of Lake Miccosukee, Fla. Grossularia curvata occurs in northern 
Georgia, northern Alabama, Louisiana, and eastern Texas. It may be questioned 
why a shrub so vigorous, so well protected against grazing animals, and so well 
adapted to dissemination by fruit-eating birds as Grossularia echinella , has such 
a limited geographical range, a strip of country about a mile in length and only 
a few rods in width. The manner of occurrence of the bushes in this area 
indicates that the establishment of the species here is of comparatively recent 
date. The plants occur in a definite center of abundance, with many individual 
bushes forming thickets, and farther away individual younger plants more 
widely separated from each other. The situation is exactly what would be 
expected if seeds of this species had been first introduced into this locality a 
few decades ago from some other and older center of distribution. The present 
known area of the species appears to be an advance colony, not a remnant, 
and it is to be expected that another and parent area will sometime be 
88285—24t-6 
