2. Allium canadense L. (wild garlic, wild onion)

Pl. 99 a, b; Map 397

Bulbs 1–3 cm long, ovoid to narrowly ovoid, the outer coat a dense network of anastomosing, hairlike fibers. Aerial stems 10–70 cm long, not inflated, erect at flowering. Leaves mostly basal or in the lower 1/3 of the aerial stems, 10–50 cm long, 1–7 mm wide, flat or sometimes folded longitudinally toward the base, linear, not tapering to a petiole, the base pale green to white when fresh. Umbels with 0–90 flowers, some or all of the flowers sometimes replaced by sessile bulblets. Flower stalks much longer than the flowers. Perianth bell-shaped, the sepals and petals 4–8 mm long, narrowly elliptic, the tips pointed or blunt, pink, less commonly white. Fruits (when produced) 3–4 mm long, depressed-globose to broadly obovoid, 3-lobed, lacking triangular projections. 2n=14, 21, 28. April–July.

Common throughout Missouri (eastern U.S. and adjacent Canada west to Montana and Texas). Present in a variety of habitats, including open forests, bluffs, glades, prairies, and various types of disturbed ground.

The bulbs, bulblets, and herbage of this species are edible and may be eaten raw, pickled, or cooked. However, the plants can impart an onion flavor to milk produced by cows that feed on them.

The biosystematic relationships among members of the A. canadense complex were studied in detail by Ownbey and Aase (1955), who classified the bulblet-forming and floriferous races as six varieties of a single species, with three of these occurring in Missouri. Steyermark (1963) and some other subsequent authors, impressed with the ecological weediness and superficial distinctiveness of the bulblet-forming race, disagreed, and continued to maintain two species in the complex. The level at which these taxa formally should be recognized is perhaps a matter more of personal taste than science. The classic studies of Ownbey and Aase (1955) are the more detailed of the two approaches and the classification of the complex as a single species appears to be more soundly supported. In Missouri, the more difficult taxonomic question, however, is not whether the bulblet-forming race should be recognized as specifically distinct from the floriferous races, but whether the two floriferous varieties treated below are truly separable morphologically. Although extreme forms are easily recognized, there are a number of specimens intermediate for one or more characters. Further studies to resolve this problem are needed.