117 
nia Nutt., Torr. & Gray, FI. i. (>89: referred to jMyrrhishy Beiith. 
h Hook. Gen. Plant, i. 897. Myrrhis ^ Glycosma Gray, Proc. 
Am. Acad. vii. 840. 
This genus is well circumscribed, not only by its fruit structure, but 
also by its general habit. In the original description of Glycosma Nutt, 
the authors note its close relationship to Osmorhiza, pointing out that it 
differs “in its glabrous fruit, depressed stylopodium, and the absence of 
involucels.” The discovery of species since then has broken down these 
characters completely; for G. ambigua has fruit sometimes bristly, G. 
oecldentalis often has a conical stylopodium, O. h ra chypod a a somewhat 
depressed one, O. nnda I’arely has involucels, and in G. occidentaUs they 
are sometimes found. In the description of 0. aadu Torr. Pacif. E. Kep. 
iv. 9:i, the author remarks that “this species is intermediate between 
Osmorhiza and Glycosma. The two genera should perhaps be united.” 
This was said in view of the combination in the new species of the bristly 
fruit of the former with the stylopodium and absence of involucre of the 
latter. Bentham & Hooker, in Genera Planfarum, refer the Glycosma 
species to Myrrhis Hcop., but fail to modify their description so as to in- 
clude them. The wing-like ribs, prominent involucels, numerous and much 
cut leaves of Myrrhis do not apply to Glycosma; a difference which is still 
further emphasized by the internal structure of the fruit. Dr. Gray, to 
reconcile this discrepancy, in Proc. Am. Acf.d. vii. .349, proposed a ^ Glycos- 
ma under Myrrhis, characterized so as to include our Glycosma species; 
but in Proc. Am. Acad. viii. JlSfi, he withdrew Glycosma from Myrrhis, and 
affirmed a much closer alliance tor it to Osmorhiza, (also, see Watson, Bot. 
Calif, i. 263), although still retaining it as an independent genus. This 
view was obtained from the study of the new Glycosma ambigimm. That 
Glycosma is distinct from Myrrhis is abundantly evident for the reasons 
given above, strengthened as they are by the fact that the prominent 
groups of strengthening cells are not found in the pericarp of Afi/rr/iis as 
in Osmorhiza and Glyco.sma, but almost in the tips of the wing-like ribs. 
In reference to the distinction between Glycosna and Osmorhiza, Dr. Gray 
(1. c.) says “the cardinal characteristic of Osmorhiza is the caudate atten- 
uation of the base of the carpels. Glycosma has none of this, but the base is 
abrupt and obtuse.” While this distinction is apparent between such 
extreme species as O. brevistylis and G. occidentaUs, the “caudate atten- 
uation” of Osmorhiza becomes a very short stout base in O. brachypoda, 
while the “abrupt and obtuse base” of Glycosma becomes acute in G. 
ambigimm. A generic line drawn between 0. brachypoda and G. ambi- 
gniim would be extremely arbitrary. 
* Fruit with prominent caudate attenuation {2 to 4 lines 
loner) at base., very bristly. 
1. O. nuda Torr. Pacif. R. Rep. iv. 98. Rather slender, 
somewhat jiubescent or glabrous: leaves twice ternate; leaflets 
