Figure 2 represents a first record from the Canadian Carboniferous of a medullosalean-impression fragment, in part preserved in siderite (and therefore from within a short distance from the roof). It is approximately 15 cm long and 8 cm wide, and some sort of a cuticular structure is preserved, judging from the brownish-looking surface that is finely granulate (Basinger et al. 1974). The significance of this find remains to be seen and cannot be properly interpreted until more material can be collected.

Physically associated with the flora is a plethora of fragmentary axes of various dimensions from 0.5 to over 13 cm in width, and up to 130 cm in length, after fragmentation by open-pit mining methods. They all show longitudinally striated or ridge-valley structures that extend into the petioles to form medial ridges (see also Gastaldo 1990). These represent cortical sclerenchymatous fibres or bundles that are characteristic of younger medullosalean stems and axes, and played a supporting role (Stewart and Delevoryas 1956; Basinger et al. 1974; Remy and Remy 1977; Pfefferkorn et al. 1984; Laveine 1986; Gastaldo 1990). None of the axes has organically attached foliage, and all show submillimetre-size rosettes of iron-copper sulphide (pyrite, chalcopyrite: indicative of locally strong reducing conditions, Eh ca. -0.3, Garrels and Christ 1965). These minerals are absent on the structure depicted in Fig. 3.

Of particular interest, because of its large recovered size, is a petiolate stem that shows 12 fibres/linear cm near the top and 14 — 15 near the bottom, and which has rare remnants of a coalified-stem compression 10 — 13 cm long (Fig. 3). The stem is ca. 130 cm long, appears uniformly 13 cm wide (but thickening below petioles is observed), and shows 7 petioles, numbered "1 to 7" in Fig. 4. They range in width from 4 — 6 cm, are entirely "naked", i.e., without laminate (leafy) structure, and do not show a dichotomy for their preserved length of up to 60 cm. This, and an additional much shorter stem (Fig. 5), is impression-preserved on a secured 155 cm by 110 cm by 30 cm silty-shale slab (weighing ca. 315 kg). Spacing between petioles is seemingly uneven, from 12 — 45 cm. The latter figure is probably biased, as between petiolar locations "4" and "7" (Fig. 4) the erosional surface is below the stratigraphic level on which the stem is imprinted. Petioles "1, 3, 4 and 7" clearly are in life position, whereas "5" could be associated with "6" (Figs. 4, 6), or it represents attachment to the hidden side of the stem. The presence of petioles "2" and "6" is each represented by a scar, as inferred from convergence of the cortical fibres. Proximally on "4" is a disruption of such structure but this does not necessarily have to be interpreted as a scar. The down-bent petioles in the upper part of the stem probably signify position after death. Closer examination of the edges of this silty-shale slab shows the presence of many additional striated, 4—5 cm wide axes with medial ridges, and indications of abundant foliage.

Two architecturally different rachides are recognizable among the many found, collected, and degaged axes. One is exemplified (Fig. 7) by a tripinnate, 65 cm long, striate, fragmentary primary rachis, 2 cm wide near the base, and 1.4 cm near the top; secondary rachides are oppositely placed, also striate, 1 — 0.7 cm wide, and preserved for a length of 13 cm. These show suboppositely-placed bases of ultimate-pinna rachides, ca. 2 mm wide, that are separated by 3 — 6 cm, depending on position on the primary pinna.

The other configuration is represented by shorter, fragmentary, "Y"-shaped axes. These show that one of the two ramifications is consistently narrower than the other (Figs. 8, 9). Although the difference is measured in terms of millimetres only (Table 1), it is deemed significant to signify morphological asymmetry, as experimental methods have demonstrated that compression results more in a biased length than width (Rex and Chaloner 1983). Asymmetry is furthermore confirmed by the unequal exterior angles between the "Y" prolongation. The lack of mirror-image symmetry (see Frye 1993, p. 106) precludes interpreting the "Y's" as dichotomously-branching axes. In specimens with narrower ramifications, asymmetry is comparatively less pronounced, forming a more normal-branching axis (compare Buisine 1961 pl. 33, fig. 2). Also, consistently observed on both ramifications, and on prolongations of asymmetric fragments, are 2 — 3 mm wide rachial bases that are generally preserved for less than 1 cm in length. On these there are also preserved, sporadically, oppositely placed axial bases that are 1 mm wide.

Figure 10 conveys a sense of the abundance of densely-occurring detached, fragmentary ultimate-alethopterid pinnae, and caducous linopterid pinnules that are associated with the medullosalean axes throughout the sample area.

Figure 11a shows the tip region of a primary or secondary pinna with 2 cm long pinnules homologous to ultimate pinnae, which then change to pinnatifid pinnules in which subsidiary veins tend to dominate. A short midvein is present with maximally two uneven forkings. The lateral-vein pattern is self-similar (see Heggie and Zodrow 1994) to that which occurs in near-tip pinnules of an ultimate pinna (see Fig. 11b). Ultimate pinnae are generally linear in outline and show undifferentiated terminal pinnules (Fig. 11b), and in life these pinnae must have been quite long, judging from the fragmentary specimen that is preserved for 15 cm. Towards the base of an ultimate pinna, pinnules developed neuropteroid attachment (having lost both basiscopic sinus, and acroscopic constriction; Fig. 11c).

Pinnules reach a maximum length of just over 2 cm and 0.7 — 0.8 cm width, being slightly convex acroscopically, with a round apex, and near parallel margins, except in the distal part where they are not. Macerated-only specimens and cuticles show strongly developed compression margins that are 800 — 900 µm wide bordering a pinnule. Moreover, the midvein is demonstrably non-decurrent (Fig. 12a), and basal veins arise directly from the ultimate-pinna rachis (Fig. 12b). Adaxial and abaxial midveins of one pinnule also show different widths, epidermal morphologies, and differing degrees of cutinization. On adaxial surfaces (Fig. 13a), the midvein does not vary much in width, from 17 µm near the base, tapering slightly to 15 µm near the apex of a pinnule. Its epidermal cells arise directly from an ultimate pinna rachis (see also Fig. 12a), are aligned, mostly rectangular-elongate, rarely spindle-form in outline, and are 30 — 156 µm long and 11 — 27 µm wide. Anticlinal cell walls are 8 — 11 µm thick; some lateral walls are concave, and highly cutinized. On abaxial cuticles (Fig. 13b), the midvein varies in width from 700 — 1,000 µm near the base, from 400 — 600 µm in the upper part, and is 40 µm wide where it breaks up into simple lateral veins ca. 1,500 µm below the apex. Its epidermal cells also arise directly from the ultimate pinna rachis, are aligned and comparatively very irregularly shaped, as spindle-forms, sphenoidal, sausage-like, rectangular with angular and rounded corners, and near-square; length varies from 34 — 133 µm, and width from 12 — 30 µm. Trichomal bases are abundant, 27 — 42 µm in diameter (larger ones are oval in shape), and show "ringed cellular structure". Trichomes are not preserved.

On the basiscopic side of the larger pinnules, lateral veins emerge obliquely from the midvein, fork, and fork again closer to the margin than to the midvein to reach the margin at an oblique angle (Fig. 14). The venation pattern represents maximum two-vein forkings, which reduces to 1-1/2, once forking, to simple veins near tip and base as rachial veins. On the acroscopic side of the same pinnule, the lateral veins are much more horizontally inclined due to compressed space, or taphonomic influence. Vein density per marginal cm is 28, i.e., a rather loose pattern.

Adaxial cuticles are thick and very cutinized, and epidermal structure is strongly differentiated between intercostal and costal fields. Intercostal cells are very irregular in shape, from near iso-diametric, quasi-square, elongate polygonal, doubly to tribulbous with curved to semisinusoidal walls (Figs. 12b, 13a, 15a). The long axes are irregularly aligned with each other (Fig. 15b). Length varies from 49 — 103 µm, and width from 27 — 57 µm. Anticlinal walls are 3 — 8 µm wide. Costal fields show mostly aligned, elongate, subrectangular cells with round corners and straight or curviflexuous anticlinal walls. Length varies from 50 — 144 µm and width from 15 — 35 µm (Fig. 15a).

Abaxial cuticles are comparatively very thin and delicate, and much less cutinized. Epidermal structure is not known to be differentiated, as intercostal cells were not observed, but rare trichomal bases are present in intercostal fields (ca. 26 µm). Costal cells are similar in size and shape to adaxial costal cells. Structure of stomata could not be determined.

The second most abundant co-floral component (Table 2) comprises the combined linopterid forms L. brongniartii and L. obliqua. Wnuk and Pfefferkorn (1984, fig. 14) proposed a quad-ripinnate frond structure for the latter. L. obliqua was diagnosed and erected as a new species by Bell (1938), and both species are morphometrically and morphologically compared by Zodrow and McCandlish (1978). During the present study, these two forms occur notably as abscised pinnules, very rarely as ultimate pinnae of which three were collected. Cuticles were not prepared, but previous preparations showed thin surfaces that fragmented along aeriolae which makes separation of abaxial from adaxial surfaces next to impossible.

Detached medullosalean ovules are common. Based on their magnitude and external topography of what is preserved, they are categorized as follows: (a) Rhabdocarpus form; 8 — 9 cm long, 4 cm wide, and pointedly oval-shaped; one prominent medial rib is shown which has 8 — 10 parallel thin lines on either side, separated from each other by 2 mm (Fig. 16a, b). Observed in several collections are ovules that occur in single file, as if arranged along either side of an axis (Fig. 16a). Most of the ovules show "slickensided" margins that probably resulted from these large (heavy?) ovules dropping into the mud flats. (b) Trigonocarpus-Holcospermum form (Zodrow 1986, pl. III); ca. 3 cm long, 2 cm wide, ovally shaped; one prominent rib with 3 — 4 less prominent ribs on either side (Fig. 16c); and (c) 1.5 — 2 cm long, subtriangularly shaped and pointed, 1 — 1.5 cm wide, ovules (of linopterid affinity?) (Fig. 16d).

Occasionally, isolated 4 — 11 mm wide, up to 25 cm long fragmentary linear structures occur, without carinae. These could be construed as medullosalean roots. However, one specimen, 22 cm wide, with intact attached adventitious roots (Fig. 17), is the only known root mantle from the Canadian Carboniferous Maritimes Basin. The base of each root at the stem is bulbous, about 2.5 cm wide; the roots are maximally 12 mm wide, and they are arranged in dense, overlapping configuration.