The shaft is shouldered (G.I., fig. 1(d)), and slab-like in section.

A (broad): There is a wide outer edge moulding, plain and flat, and worn on the left-hand side. The inner plain moulding is one inch wide. The ornament is arranged in three panels, defined by the inner moulding. The lower part of the shaft contains a pair of narrow vertical panels with semicircular arched tops, while the upper part, above the shoulder, contains a single panel whose base is defined by a pair of concave arcs meeting at a point in the middle, the arcs being roughly concentric with the inner parts of the arched heads of the panels below. This shape is repeated by a plain moulding in the wide, plain band between the upper and lower panels; the band then narrows and fills the space between the two panels. Each of these contains interlace, the pattern being clearer in the right-hand one. This contains three registers, and the beginnings of a fourth, of encircled pattern E, linked by pairs of pattern C loops. The form of the terminal is unclear, but the narrow strands apparently ended in animal heads. The left-hand panel contains the worn remains of two registers and part of a third of an encircled pattern, probably a form of turned pattern C (a ring-knot). Again, the form of the terminal is unclear.

The upper panel of the shaft is filled with interlocking profile animals and interlace. The beasts are disposed diagonally with right-angle bends, and have elongated thin necks and limbs. The tail, ear, and tongue are extended into two thin fetters. The small snout is curled. The jaws hang open, the tongue passing between fangs. The fetter is a logical continuous band of loops and knots, filling space without crowding.

B (narrow): The upper section, above the shoulder, contains the worn remains of a panel defined by a broad outer and narrow inner moulding, disposed in a concave arc at the base to fit the shoulder of the shaft. It contains a now indecipherable interlace. The lower panel had similar mouldings and has a semicircular arched top, concentric with the base of the upper panel. It contains the worn remains of at least three registers of interlace, apparently a form of half pattern with outside strands. The strands are narrow, as on face A, and appear to have terminated as loose ends, one perhaps ending in an animal head, as on face A.

C (broad): Totally defaced. It is cut back into two rectangular, shallow troughs. No original carving survives.

D (narrow): The disposition of the panels, and the form of the mouldings defining them, are as on face B. The panels are filled with interlace formed from narrow strands, as on the other carved faces.

The upper panel contains two registers of half pattern F, turned; the terminal is apparently formed from pattern D loops and loose strands. The lower panel has an irregular upper terminal, with an encircled unpinned loop, followed by three registers of what is probably spiralled half pattern A. The form of the lower terminal is unclear.

This monument has an unusual form for Yorkshire. Shouldered shafts are quite rare; even at Whalley, Lancashire, the form is suggested by the ornament rather than the profile (Collingwood 1927, 108, fig. 132). A small group is concentrated in Ryedale: Levisham 3 (Ills. 644–7), Hovingham 3 (Ills. 490–3), and the present shaft. The sudden narrowing of their broad faces is accentuated by the pair of adjacent panels with curved tops, here and at Hovingham filled with Anglian ornament. Adcock has indicated (1974, I, 133–5) that some of these patterns are unique in Northumbrian sculpture, the encircled pattern E and C, but do occur in the Lindisfarne Gospels. The carpet pages on fols. 2v and 94v have calligraphic versions of the pattern (Backhouse 1981, 34, pl. 19 and 48, pl. 28). The turned pattern C, the ring-knot, does have a wider distribution in sculpture, Adcock pointing to Stanwick, North Riding, Chester-le-Street, co. Durham and Whithorn, Galloway. Her opinion of the quality of the interlace, despite its interest, is critical: 'The terminals show that the designer was not a person who understood the drawing of interlace . . . the arched terminals are space-filling nonsense' (Adcock 1974, I, 135). The variety of pattern F used is not found elsewhere in Northumbrian carvings.

A possible manuscript source is also possible for the animal ornament of the upper portion of the shaft. The beasts are interlocked and fettered in the Insular manner, with prominent nostrils and infilling interlaced trails, probably body extensions. Allowing for the different scale and medium, they belong to the genre typified by the canine profile beasts found on fols. 211r of the Lindisfarne Gospels (Backhouse 1981, 81, pl. 56; Haseloff 1987, 48–9, 50, fig. 9), where the trails even turn in pattern E terminals, like those of the monument. So traditional and long-lasting are these animals, since they result from construction on a diagonal grid, that they are no indication of a very early date for the shaft. Similar beast-chains appear as revivals on very late Irish sculpture (Henry 1970, 193, fig. 30), as well as in restricted panels of borders in books like the Macregol Gospels (Henry 1965, pl. 110; Hemphill 1912, pl. 1) where the proportions of the bodies and their parts provide close parallels for Kirby Misperton 1. Adcock postulates a late ninth-century date but that would place it in the Anglo-Scandinavian period, when local styles had developed to some degree. The animals of the beast-chain were not influential upon the tenth-century zoomorphic ornament of Ryedale. Neither are they typical of late Anglian beasts in York.