592 TRANS. ST. LOUIS ACAD. SCIENCE. 



anther-scales 10-12, rounded, entire, each bearing usually 4 or sometimes 

 •6 anther-cells ; berries on straight peduncles, 1-2 -seeded; seeds angled, 

 mostly grooved, and often rough toward the upper end. — Spec. PI. 1471 

 (1753); Parlat. 1. c. 488. (See Fig. 8.) 



From the St. Lawrence to Florida and from the Atlantic to the North- 

 ern Pacific; it is not found in Southern Texas, in the greater part of 

 Utah and Arizona, and in the whole of California and perhaps Oregon; 

 in Washington Territory and British Columbia it associates with Sabina 

 and perhaps with occidentalism and in the Rocky Mountains south of Pike's 

 Peak with the latter species ; on the upper Missouri (Cedar Island) it 

 attains large dimensions. — Usually the berries are small, about 3 lines 

 thick, but in the Rocky Mountains forms occur with berries of 4 or 5 lines 

 in diameter, and with larger seeds ; among the foot-hills of Pike's Peak 

 the trees of this species have the size and shape of an apple tree, with a 

 rounded, spreading top. 



9. J. Bermudiana, Lin. ; A tree said to have been common on the Ber- 

 mudas, and also in other West Indian Islands, of which I have seen only 

 ■ a few specimens. Branchlets stout; leaves in pairs, oblong or linear- 

 oblong, obtuse, closely appressed, with entire margins and a well marked 

 linear gland or resin-duct on the back; anther -scales about 16, large, 

 rounded, smooth-edged, with about 6 cells ; berry with 2-4 seeds, much 

 like those of the last species. — Spec. PI. 1. c. — Parlat. 1. c. 490. (See Fig. 8.) 



y. Barbadensis, Lin. is said to be the same species, and Biota Melden- 

 sis, Gord. its acerose young state. Michaux, as well as Parlatore, quotes 

 Florida as its home, but all the specimens from that country which I have 

 seen, even those from Cedar Keys, and those of Michaux's Herbarium 

 in Paris under the name of y. Barbadensis, are nothing but forms of y. 

 Virg-iniana, with very small, rounded and strongly convex leaves. The 

 forms from the different West Indian Islands, all referred to y. Bcrtnu- 

 ■diana, require further examination, as we know that one at least, from Cuba 

 .(see p. 590), is certainly quite different from it. 



