File:Late Saxon coin (FindID 500347).jpg

English: Hoard of 23 coins, four silver brooches and two silver strap-ends.

Coins: Twenty three silver pennies of Eadmund of East Anglia, covering most of the types known for this king. They were found close together and clearly represent a dispersed hoard. The lack of any later coins is compelling evidence for the deposition of this hoard being intimately connected to the invasion of the Danish Great Army into East Anglia in 869 and the succeeding murder of king Eadmund in the same year.

1) North 456, moneyer Baeghelm (fragment, weight 0.35g)

2) North 456, moneyer Bea[...], probably Beornheah blundered (fragment, weight 0.90g)

3) North 456, moneyer Beornheah (same dies as no. 4, weight 1.35g)

4) North 456, moneyer Beornheah (same dies as no. 3, in two pieces, weight 1.04g)

5) North 456, moneyer Eadberht (fragment, weight 0.37g)

6) North 456, moneyer Eadmund (fragment, weight 0.47g)

7) North 456, moneyer Eadmund (fragment, in two pieces, weight 1.08g)

8) North 456, moneyer Eadwald (1.09g)

9) North 456, moneyer Eadwald (0.97g)

10) North 456, moneyer Eadwald (fragment, weight 0.91g)

11) North 456, moneyer Eadwald (fragment 0.75g)

12) North 456, moneyer Sigered (weight 1.16g)

13) North 456, moneyer Sigered (incomplete, weight 1.06g)

14) North 457 or 458, moneyer Ethelwulf (fragments, weight 0.51g)

15) North 458, moneyer Aethelhelm (weight 1.29g)

16) North 456, moneyer Aethelhelm, obverse blundered (weight 1.18g)

17) North 459, moneyer Beornferth (weight 0.95g)

18) North 461, moneyer Baeghelm (weight 1.33g)

19) North 461, moneyer Baeghelm (weight 1.01g)

20) North 461, moneyer Baeghelm (same obverse die as no.21, incomplete, weight 0.98g)

21) North 461, moneyer Baeghelm (same obverse die as no. 20, weight 1.11g)

22) North 462, moneyer Twicga (weight 0.97g)

23) North 462, moneyer Twicga (weight 1.35g)

Artefacts: Six silver objects. Two cross-shaped brooches, (a) and (c); one openwork object, probably from a brooch, (b); one fragment of brooch, (f); and two silver strap-ends, (d) and (e).

(a) Silver disc brooch

Description: Part of an openwork silver disc brooch in the form of an equal-armed cross with splayed terminals to the arms. About a quarter survives, now in two joining pieces; one piece with the central roundel and a fragment of surviving cross-arm, and a second joining fragment with a complete cross-arm and parts of two openwork panels between the arms. The breaks are all fresh and granular.

The cross is divided into a number of decorative panels, each containing Trewhiddle-style decoration reserved in silver on a field of niello, which substantially survives but is missing or decayed in some places. Where the niello survives well, the flat polished surface can clearly be seen; the brooch was perhaps not polished down quite far enough, as much of the ornament is a little unclear. The narrow bands of silver separating the fields are in most cases decorated with a line of punched circular dots, with engraved guide lines visible in one place.

The cross has outward-curved ends to the arms and inward-curved angles between. In the spandrels of the cross (the angles between the arms) two incomplete circular or near-circular openwork panels survive, one adjacent to the other. They each appear to contain the same motif, unclear because incomplete; it consists of a central long triangle or drop shape outlined by niello inlay, with the wider end towards the centre of the brooch; this is flanked by a pair of triangles with inward-curling circular terminals at the apexes, both of which point towards the centre of the brooch. These may represent wings, or leaves, as their niello-inlaid contour lines each have a niello-inlaid spiral line, perhaps a stalk, curling outwards and passing underneath the triangle. The relationship of this line and a short length of additional groove parallel to but outside the triangle forms an alternative, smaller pair of wing-like elements; if these are to be read as the bird's wings, perhaps it was enmeshed in foliate interlace. Alternatively the design may be that of a pair of profile birds, each with one larger round-ended wing; or the complete motif may have been entirely vegetal, and the bird-like elements merely coincidental.

At the centre of the brooch is a silver rivet, the head missing, with a blank space around it perhaps originally covered by a boss or domed rivet head. A second silver rivet close to the edge of the brooch holds a fragment of silver in the shape of a stylised animal head, with incurved edges forming the snout, the rivet passing through the top of the head, and a broken edge behind; this is probably (from the parallels of Elmsett and Pentney, see below) the remains of the otherwise missing pin or catch-plate.

The central rivet is surrounded by four panels forming a roundel. Two of the opposing panels contain plant motifs of two branches issuing from a central stem, but they are not the same. One has outward-curling leaves on each branch, forming a double acanthus leaf pattern; the other (adjacent to the surviving cross-arm) has two triple-branched stems interlacing with a figure-of-eight.

The other opposing pair of panels each contains a profile beast; these are both the right way up when the brooch is held with the complete arm to the left. The lower panel has a beast with its head turned back to look over the shoulder, with an excessively long tongue which issues from the open mouth and forks as it crosses over the body. One end crosses over the hind leg and ends in a scrolled terminal; the other under the foreleg, which clearly has two toes, and ends in a simple round-ended terminal. The animal has a round eye, a blunt-ended upwards-pointing tail, and a nick into the body to either side of the foreleg.

The beast in the opposing panel is less clear, and appears to have been less well designed. It also appears to be looking over its shoulder, and again has a tongue emerging from its mouth, but this then appears to merge with what may be a hind leg, with two nicks at its top, but no foot; it instead ends in a scrolled terminal. The head is detached by a groove passing across the neck, and has a round eye. In front of the head are some space-filling elements (a pellet and a V shape) that look a little like leaves; behind the head (in front of the curving chest) is what looks suspiciously like the tail on the other panel, but it is entirely in the wrong place for a tail. A two-toed foreleg emerges from the lower edge of the body, and there are two grooves across the body.

The panel at the start of the broken cross-arm is filled with a foliate motif, with a central stem from which two branches emerge. The lower branch ends in three leaves and the upper in two, and the stem ends in a trefoil consisting of two curved leaves and a central round leaf, showing that the break must have occurred at or close to the end of the panel. The panel is the right way up when the centre of the brooch is held downwards.

The complete arm is divided into three panels. The first contains a profile beast with open mouth, round eye and long ear. It has no tongue, but the lower jaw is extended and curled downwards to end in a scrolled terminal in front of the chest. The beast has two legs, each ending in three-toed feet, and a long tail which curls over the body and under the rear leg, and ends in a scrolled terminal behind the rear foot. There are space-filling elements around the animal; a reserved triangle under the body, a single reserved line in the corner above the tail, and several curved lines and shapes in front of the animal. This panel is the right way up when the brooch is held with the outer edge downward.

The outer two panels each contain a beast looking back over its shoulder, with slight differences between the two. Each has an open mouth biting at a pellet, a round eye, ear, toed feet, a large trefoil tail with double grooves around the base of at least the outer two foils, a double-grooved collar around the neck and multiple circular dots or 'pecking' (using the same tool as the dots on the lines between the panels) on the body. The left-hand animal has a round ear, a round pellet, the foreleg bent up in front of the neck and a pair of inward-facing scrolls as space-filler below the body. The right-hand animal has a long thin ear, a triangular pellet, the foreleg bent down under the body, two niello-inlaid grooves within the body and a reserved lozenge as space-filler below the body. These panels are the right way up when the brooch is held with the outer edge downward.

Between these two panels the edge of the brooch is recessed, with a reserved border around the riveted-on element. To either side of this, the surviving original edge of the brooch is beaded.

Dimensions: Original diameter (reconstructed) 68mm. Thickness less than 1mm. Weight 7.2g.

Analysis: Non-destructive X-ray fluorescence analysis of the surface indicated a silver content of c. 95%, with c. 3% copper and traces of gold, lead and zinc (plus traces of iron, probably from the burial environment). The presence of niello (containing gold as a trace element) was confirmed during the scientific analysis.

Discussion and Date: The overall design is very similar to an example from Elmsett, Suffolk (West, 1998, 140, fig.24, no.6). The Elmsett brooch has almost-circular openwork elements enclosed by the arms of a cross, and a separate pin with riveted animal-head attachment. The detail of the ornament on the cross is, however, more closely paralleled on the Beeston Tor brooch (Webster and Backhouse 1991, no. 245(a), Wilson 1964, no. 2), which has remarkably similar Trewhiddle-style animals, with distinctive curved grooves on the body; one also has a large trefoil tail, a motif which may be related to the peacocks' tails found on the Ethelwulf ring (Webster and Backhouse 1991, no. 243). The Beeston Tor brooch was found in a hoard coin-dated to c. 875 and must have been made not long before; the Ethelwulf ring probably dates to before 858.

Three fragments of another brooch of this type are recorded on the PAS database at BM-138695 and LIN-9B7487, and there are also similarities to the largest brooch from the early 9th-century Pentney hoard (Norfolk HER 3941; Webster and Backhouse, 1991, 230, 187e) which again has animal-head clips holding the pin and catchplate.

(b) Incomplete silver openwork object, probably part of a brooch

Description: Incomplete silver openwork object, probably part of a brooch, which can be read as being in the form of an equal-armed cross with splayed arms with concave sides and convex ends. About one third survives, now in two joining pieces.

The centre of the object has a tiny undecorated square panel with a separate dome-headed silver rivet through the centre. On the reverse this rivet is almost flush to the surface and was therefore probably wholly decorative. Around the central rivet are four square apertures. The square apertures are each defined by a line of niello, which returns at the junction with the central panel to join up with its neighbour. Outside this is a line of tiny pecked dots which run between and around all the apertures, then a further parallel nielloed line follows the outside of the group of squares and extends at the corners to define the rest of the ornament. Under the microscope the pecked dots can be seen to be long wedge-shaped marks.

At each corner of the central group of squares is a roughly semi-circular openwork panel. Only one survives semi-complete and can be reconstructed; there is a central drop-shaped element outlined in niello, the apex pointing towards the centre of the object, with a sub-triangular aperture to either side and an undecorated rounded shape linking the drop-shaped element to the rest of the edge of the object. Fragments of two other panels suggest that all were identical in design.

In between the semi-circular panels are what may have been intended as cross-arms. Again, only one survives in a near-complete state, but the fragments of two others again suggest that all were identical in design. The surviving arm has four apertures created by two opposing long triangular panels set radially, apex to apex, and two opposing comma-shaped panels, all four decorated with a contour line of niello. The motif here may be a single highly stylized animal head, with the comma-shaped elements forming the eyes and the apex-to-apex triangles the brow and nose, as found on the Strickland brooch (Webster and Backhouse 1991, no. 189; Wilson 1964, 152). If so, the semi-circular panels could possibly represent even more stylised heads facing inwards, again as on the Strickland brooch, which dates to the 9th century AD.

The brooch is made from very thin sheet silver. There is much copper corrosion over both front and reverse and all the breaks appear fresh. The reverse has areas of matt grey surface, perhaps solder. This suggests that the object may be the thin openwork front-plate of a brooch with a separate, perhaps copper alloy, back-plate.

Dimensions: Original diameter (reconstructed) 54mm. Thickness less than 1mm. Weight 3.7g.

Analysis: Non-destructive X-ray fluorescence analysis of the surface indicated a silver content of c. 87%, with c. 7% copper and traces of gold, lead, tin and zinc (plus traces of iron, probably from the burial environment). The presence of niello was confirmed during the scientific analysis.

Discussion and Date: The object is most likely to be the front-plate of a composite brooch. The smallest brooch in the Pentney hoard (Webster and Backhouse 1991, 230, 187f ; Norfolk HER 3941) had an openwork silver front attached to a gilded copper-alloy back-plate, but in this case the attachment was done with a quincunx of five functional rivets rather than the solder which appears to have been used in this case. Although the decorative style of the Pentney brooch is altogether different, being based on curvilinear plant motifs and datable to the late 8th century, the technique is similar and the dates of the two objects cannot be widely separated.

(c) Incomplete silver cross-shaped brooch

Description: Incomplete silver brooch in the form of a cross with splayed, convex ended arms, now very corroded. It bears Trewhiddle-style animal and possibly foliate decoration, reserved in silver and with the engraving inlaid with niello. One edge has an integral catchplate surviving complete, and the opposite edge has a barely-visible projection c. 4mm wide, probably the filed-down remains of an integral spring and pin which was lost in antiquity. There is also a rivet hole, perhaps secondary, although there is so much corrosion over the decoration that it is impossible to tell; this rivet hole presumably attached a separate pin as part of a repair. The probable presence of solder was detected as part of the scientific analysis (see below). This axis of the brooch, parallel to the pin, is noticeably longer than the perpendicular axis.

The corners of the ends of the arms terminate in scrolls which meet, fully enclosing the internal angles or spandrels between the arms of the cross. Each of the spandrels contains a teardrop-shaped panel, apex outwards, with a sub-triangular aperture to either side. The teardrop has a groove around the edge and two transverse grooves, all originally filled with niello.

In the centre of the brooch is a separate but wholly decorative silver rivet. Four narrow lines of reserved silver extend from this towards the drop shapes, to form the divisions between each cross arm. The panels thus formed have decoration obscured by much silver corrosion and decayed niello, but it seems likely that one panel (uppermost when the brooch is held with the empty rivet hole to the right) contained a Trewhiddle-style animal, as a round eye can just be seen at the top with perhaps an ear and open jaws.

Turning the brooch so that the catchplate is to the right, a second animal can be seen on the right-hand panel, with a tail behind a haunch and a rear leg with two toes. A round eye can be seen at the end of this incomplete panel, and it seems likely that the head was turned around to face backwards towards the tail. Deep nicks can be seen into the body.

At one edge of the long axis a collared rectangular-sectioned projection extends, and is bent to form the integral catch-plate. At the opposite end of the brooch, the rivet hole is surrounded by a small amount of iron corrosion, particularly on the front, presumably from the (missing) pin. The reverse is corroded and undecorated. A corner is missing from three of the arms of the cross.

Dimensions: Length of plate 33.5mm. Length including catch-plate 39mm. Width 29mm. Thickness 1.5mm. Weight 3.9g.

Analysis: Non-destructive X-ray fluorescence analysis of the surface indicated a silver content of c. 97%, with c. 1.2% copper and traces of gold, lead and zinc. The matt grey corrosion on the reverse showed higher amounts of tin, copper and lead, which could indicate the use of lead-tin solder for an otherwise almost-invisible repair. The presence of niello was confirmed during the scientific analysis. Unlike objects (a), (d) and (e), the niello contained no gold.

Discussion and Date: The overall design, with outward-pointing drop shapes in the spandrels meeting the scrolled ends of cross-arms, is akin to that on the similarly sized brooch from the Heckington area, Lincs, found in three fragments and recorded on the PAS database at BM-138695 and LIN-9B7487. The use of Trewhiddle-style ornament dates this brooch to the 9th century AD.

(d) Silver strap-end

Description: Trewhiddle-style silver strap-end of Thomas's Class A, Type 1. It has a bi-convex split end with two dome-headed silver rivets surviving. A trefoil motif extends from between the rivet holes to fill a pelta-shaped field at the top of the strap-end; each of the long side leaves of the trefoil are narrow, curved and with pointed ends, and the central shorter leaf has a pointed base and a bulbous rounded end.

The sides of the strap-end are very slightly convex, and have multiple notches creating a scalloped edge which continues to the animal head terminal. This has large oval ears with central C-shaped recesses, each with a V-shaped centre. There is no niello visible in the recesses, but there is iron corrosion in the left-hand one.

The brow and snout are essentially cruciform, with curved angles to fit around the ears above and lentoid eyes below. The eyes have counter-relief centres, each with a drilled dot in the centre; one of these dots retains a blue glass pellet at the centre, the other is obscured by iron corrosion. There is iron corrosion over much of the eye recesses.

Each of the arms of the cross shape are blunt-ended and in the centre is a lozenge shape outlined in niello. Four short grooves running towards the centre, all with niello inlay, form a quatrefoil motif. Three of the foils are lozengiform with pointed ends; the fourth foil extends to follow the shape of the snout and has a transverse groove a little above the blunt end, also inlaid with niello.

In the centre of the plate, to be viewed with the split end held to the left, is a decorative field with convex upper and lower edges and concave ends, containing a beautifully made profile beast reserved in silver on a field of niello, now decayed in many places. The beast is crouched, the legs terminating in feet tucked under the body. The forefoot appears to have two main elements, a bulbous upper one with a pair of nicks and a more conventional lower one with a comma-shaped foot and a tiny lobe behind; the rear foot appears to have three toes.

The head is turned backwards to look towards the tail, and has a large lozengiform ear filling the corner of the field, a blind hole for the eye (perhaps originally filled with niello, or a glass pellet) and curved brow with a curved groove forming an eyebrow. The open mouth has round-ended jaws, and a narrow tongue extending which ends in a large triangular terminal. Around the neck is a collar formed from two grooves. The long tail extends from the rump, which has a curved groove beneath the tail. The decay of the niello makes detail unclear, but the tail appears to bifurcate, having one short straight upper fork ending in a rounded terminal, and a lower tendril that passes under the rear leg and perhaps ends in a further bifurcated terminal with round ends. Alternatively this V-shape may be part of the toes of the rear foot; there is a pellet behind the rear leg which could possibly form an alternative terminal to the tail.

On the reverse of the terminal, more iron corrosion adheres in a ring around the edge of the terminal. There is also copper corrosion inside and around the split end.

Dimensions: Length 45mm. Width 6.5mm. Weight 9.4g.

Analysis: Non-destructive X-ray fluorescence analysis of the surface indicated a silver content of c. 92%, with c. 5% copper and traces of gold and lead. This is slightly different to strap-end (e), which is formally very similar. The presence of niello (containing gold as a trace element) was confirmed during the scientific analysis.

Discussion and Date: The use of the Trewhiddle style is widespread on 9th-century strap-ends, and the motif of a profile animal with backwards-turned head, orientated so that it is the right way up when the strap-end is held with the terminal to one side, is also very common (compare, for example, KENT-A58993). This particular strap-end is heavy and well made, and is a high-quality example of the type.

(e) Silver strap-end

Description: Trewhiddle-style silver strap-end of Thomas' Class A, Type 1. It has a bi-convex split end with two rivet holes, one retaining a silver rivet with flat lozengiform head. A trefoil motif extends from between the rivet holes to fill a recessed pelta-shaped field at the top of the strap-end; each of the long side leaves of the trefoil are narrow, curved and with pointed ends, and the central shorter leaf has a pointed base and a rounded end.

The sides of the strap-end are slightly convex, and have multiple notches creating a scalloped edge which continues to the animal-head terminal. This has large oval ears with C-shaped recesses filling their lower halves, each recess having a V-shaped centre. Above these recesses, each ear has two narrow curved grooves above. There is no niello visible in any of these grooves or recesses.

The brow and snout are essentially cruciform, with curved angles to fit around the ears above and lentoid eyes below. The eyes have counter-relief centres, each with a drilled dot in the centre; one of these dots retains a blue glass pellet at the centre, the other has a fragment of blue glass adhering. Each of the arms of the cross shape are blunt-ended, except the upper arm which ends in a circular terminal between the ears, with a central reserved silver circle outlined with niello. In the centre, between the eyes and ears, is a lozenge shape outlined in niello. Four short grooves running towards the centre, all with niello inlay, form a quatrefoil motif. Three of the foils are lozengiform with pointed ends; the fourth foil extends to follow the shape of the blunt-ended snout and has a transverse groove a little above the blunt end, also inlaid with niello.

In the centre of the plate is a decorative field with convex sides and concave ends, separated from the animal-head terminal by a reserved area with a row of tiny stamped triangular dots. The central field is divided into four by a cross with a reserved circular centre and reserved circular terminals to the arms; the upper and lower terminals are smaller than the other reserved circles. Each of the four resulting panels contains a profile beast reserved in silver on a field of niello, now partly missing and decayed. The beasts all have their hind legs in the outer corners and their heads towards the centre, but are all different; they fall into two pairs, those on the right-hand side of the plate (held with the split end uppermost) and those on the left-hand side.

The two beasts in the upper and lower panels on the right side of the plate are crouched, their front feet facing towards each other, their long necks bent so their heads face back over their rumps. Both beasts have open mouths with blunt-ended jaws, dots of niello for eyes, and short ears behind rounded brows. The upper beast is clearer, with a two-toed front foot, a single-toed rear foot, a deep nick into the belly just in front of the shoulder, and a pair of narrow grooves across its haunch; there is a circular pellet just above its nose. The lower animal has a similar pellet, but this appears to be the terminal of its curving tail. It also has a curved groove emphasising the base of the tail, and a pair of narrower grooves across the body.

The beasts in the left panels are also crouched, facing each other, with their heads held forwards and angled slightly upwards. The decomposition of the niello and some scrapes to the surface make the details a little unclear, but each appears to have an open mouth and short neck with a pair of narrow transverse grooves forming a collar. Each has a pair of narrow grooves across the haunch, and comma-shaped feet; the lower beast has a short tail.

A small part of the back of the split end, broken across a rivet hole, is missing. There are multiple scratches on the reverse, including one straight scratch that runs almost perfectly down the centre. There is also an irregular patch of green corrosion on the reverse.

Dimensions: Length 50mm. Width 17mm. Weight 10.9g.

Analysis: Non-destructive X-ray fluorescence analysis of the surface indicated a silver content of c. 89%, with c. 7% copper and traces of gold, lead and zinc (plus traces of iron, probably from the burial environment). This is slightly different to strap-end (d), which is formally very similar. The presence of niello was confirmed during the scientific analysis.

Discussion and Date: The use of the Trewhiddle style is widespread on 9th-century strap-ends, and the division of the main decorative area into constituent fields, as here, is also common, particularly in south-eastern England (Thomas 2003, 2 and fig. 1.4). This particular strap-end is heavy and well made, and is a high-quality example of the type.

(f) Silver brooch fragment

Description: Silver fragment, probably from a brooch, consisting of a large, apparently decorative, silver rivet through a fragment of flat plate with no original edge surviving. The decoration, reserved in silver, probably originally on a field of niello now missing, consists of four radiating lines decorated with tiny dots, dividing the area around the rivet into four curved panels forming a roundel, only one of which survives almost intact. This panel has an outer border also decorated with punched dots, and is filled with a pair of reserved outward-curling spirals either side of a reserved trefoil motif. The tiny surviving portions of the other three panels suggest that the opposite panel was decorated alike, and that a different design or designs filled at least one of the remaining two alternating panels. The front of the fragment is dark grey and corroded; the reverse is more shiny. All of the breaks appear fairly fresh. The base of the relief is keyed for niello inlay although none of this now survives.

Dimensions: Surviving dimensions 11 x 17mm. Weight 1.3g.

Analysis: Non-destructive X-ray fluorescence analysis of the surface indicated a silver content of c. 94%, with c. 3% copper and traces of gold, lead and zinc (plus traces of iron, probably from the burial environment). No trace of niello was detected during the scientific analysis.

Discussion and Date: Although containing similar decorative elements, this fragment does not join any of the other three incomplete brooches so far discovered from the hoard, and almost certainly represents a fourth brooch. The use of dotted reserved lines, silver and niello, as well as the decorative motif, dates the fragment to the 9th century.

Discussion of the artefacts as a group

The artefacts are all likely to have been produced within the 9th century, the openwork object (b) (possibly a brooch front) perhaps a little earlier than the others. This date is consistent with contemporary deposition with the coins. Although all four brooches are now fragmentary and incomplete, all the breaks seem to have occurred as accidental fragmentation in the ground after burial, probably as a result of agricultural activity, and not from the deliberate cutting of the artefacts into weights of silver. The objects therefore seem likely to have been valued as artefacts rather than bullion, perhaps indicating an Anglo-Saxon rather than Viking milieu, but with burial connected to the invasion of the Danish Great Army in 869.

Although details of the decoration differ between the brooches, they follow the same overall pattern of a cross with circular panels formed in the internal angles. Brooches (a) and (c) also appear well made, better in design and execution than (b). Not enough of (f) survives to judge.

The two strap-ends are both well made and, apart from the use of different motifs, in general quite similar. They are likely to have been products of the same workshop, and were perhaps produced as a high-quality set, although (d) is very slightly better designed and made than (e).

There are some decorative similarities between the strap-ends and the brooches; the use of short curved grooves, perhaps to indicate musculature, occurs on both strap-end (d) and brooch (a), and brooch (a) also has a trefoil of very similar nature to that on both strap-ends. The group of artefacts therefore shows some coherency, particularly between (a), (d) and (e).

Scientific analysis has shown that the niello in (a), (d) and (e) appears to contain gold, but that in (c) has none detectable. The metal composition of the two strap-ends (d) and (e) is slightly different, suggesting that a group of similar objects was put together some time after their manufacture, rather than that a group of objects was made from the same melt and then immediately deposited.

Erica Darch, Adrian Marsden, Gareth Williams and Helen Geake