3. Quercus ellipsoidalis E.J. Hill (northern pin oak, Hill’s oak, jack oak)

Pl. 414 c, d; Map 1841

Plants trees to 25 m tall. Bark medium to dark gray, divided into persistent ridges, the inner bark orangish. Twigs 2.0–2.5 mm wide, dark brown to reddish brown, glabrous. Buds 3.5–4.5 mm long, brown, the lower scales glabrous or nearly so, the upper scales pubescent. Petioles 30–45 mm long. Leaf blades 11–15 cm long, 8.5–12.0 cm wide, obtuse or broadly obtuse at the base, divided 80–90% of the width, the lobes 3 or 4 per side, evenly spaced, with well-developed lobes 32–45 mm wide, broadened outward, tapered apically, with 1 strong secondary lobe on the lower margin and several teeth, each with 5–9 bristles 4–5 mm long (the whole blade with 25–35 marginal bristles), the strongest secondary veins reaching the margin at the tips of the lobes and ending in bristles, others reaching toward sinuses and turning aside before reaching the margin; the upper surface rather shiny, glabrous or with a few hairs scattered along the main veins; the undersurface green, glabrous, smooth to the touch, the vein axils with small tufts of 6–9-rayed, stalked hairs, sometimes a few hairs also scattered along the sides of the midrib. Acorn cups 13–14 mm long, 14–17 mm wide, covering 40% of the nut, bowl-shaped, the base prolonged, the inner surface smooth, sparsely pubescent near the nut scar, the outer surface with the scales thin and plane, sparsely pubescent. Nuts 14–17 mm long, 10–11 mm wide, ellipsoid, without concentric grooves around the tip. April.

Uncommon, known only from Harrison County (north-central U.S. and adjacent Canada, from Ontario and Ohio west to Manitoba and Missouri). Mesic upland forests and savannas; also pastures.

Most Missouri specimens previously determined as Q. ellipsoidalis are actually the northern race of Q. velutina, which has a similar growth form (see discussion under that species). Additionally, at least one specimen from Harrison County (Hershey s.n. on 12 Sept. 1980, at the Missouri Botanical Garden Herbarium) is close to Q. ellipsoidalis, but appears to represent the result of past hybridization between that species and Q. velutina. Quercus ellipsoidalis is morphologically variable and confusing throughout its range (Overlease, 1977; R. J. Jensen, 1977, 1986; Hipp and Weber, 2008). R. J. Jensen (1977) showed that puzzling specimens intermediate between Q. velutina and Q. ellipsoidalis are found across the Great Lakes region. In their genetic study of Q. ellipsoidalis and its relatives, Hipp and Weber (2008) discussed evidence documenting past hybridization between this taxon and Q. velutina. Until more ample material is collected from Missouri, it is impossible to evaluate the status of this apparently very rare species in the state.