Queen Elizabeth I
15 of 601 portraits matching these criteria:
- subject matching 'Jewellery - Precious stones'
- Tudor and Jacobean Portraits Database
Queen Elizabeth I
associated with Nicholas Hilliard
oil on panel, circa 1575
31 in. x 24 in. (787 mm x 610 mm)
Key findings: The close association of the Phoenix portrait with the Pelican portrait was confirmed.
Acquired by the Gallery from Colnaghi in 1865. First recorded in 1839 when the Rev. W. Newcome was Warden of Ruthin, Denbighshire. The portrait may have been bequeathed by Gabriel Goodman, Dean of Westminster (1529?-1601) with his household stuff to Christ’s Hospital, Ruthin, which he founded in 1590.
Referred to as the ‘Phoenix’ portrait after the jewelled pendant worn by Elizabeth I (see detail 05). The phoenix jewel symbolises rebirth and chastity. It became associated with the queen in the 1570s as an emblem of virginity, and as reassurance that she would be able to regenerate the dynasty.
Notes on likely authorship and justification
The portrait is closely related to another portrait of Elizabeth I, in which the queen is depicted wearing a pelican jewel (Walker Art Gallery). Both portraits can be compared to a miniature of Elizabeth I painted by Nicholas Hilliard in 1572 (NPG 108).
Hilliard’s treatise refers to the technique of oil painting and thus he appears to have been aware of the oil painting process. However, there is no categorical evidence that he painted in oil himself. Hilliard may have worked to establish the design and thereafter passed the execution to a practised oil painter. There is also evidence of more than one hand in the painting, so it is probable that the portrait was produced in a workshop under Hilliard’s supervision.
Commentary on painting style, technique
The painting is in a stable condition, although the thick and discoloured varnish layers affect the understanding of the paint layers and pigments.
A number of comparisons can be made between the technique of the Pelican and the Phoenix portraits of Elizabeth. The painting techniques display a very high level of knowledge and skill in the manipulation of oil paint, with striking attention to detail. This is notable in the jewels, where the extremely fine brushstrokes used to depict small fruits and gem stones can be compared to a miniaturist’s technique. The painting technique is also reasonably painterly, particularly in the fan, the ruff and some parts of the hair, where wet-in-wet blending can be seen. The small dabbed brushstrokes of the phoenix jewel compare very closely with the handling of the pelican jewel and the painting of the hair is also very similar in both portraits. However, the treatment of the ruff and cuff in the two portraits cannot be compared; the small dabbed brushstrokes, applied to soften the threads, are unlike the crisp handling of the ruff and cuff in the Pelican portrait, and were probably executed by a different hand.
The interesting palette and pigments also compare closely with the Pelican portrait. These include an unusual, very smooth, green pigment used in the rose, which has been identified as green verditer (synthetic copper green). Indigo was also used in both portraits. The use of pigments is carefully controlled; for example, vermilion is restricted to the wings and fire on the phoenix jewel.
Justification for dating
The technique and materials in use are entirely consistent with a work from this period. The portrait is painted on boards from the eastern Baltic region which came from the same two trees as the boards used for the Pelican portrait (Walker Art Gallery). There is evidence which suggests that the two panels might once have been a single piece prior to painting. Dendrochronological analysis revealed that the wooden boards derive from two trees that cannot have been felled before 1557 and 1561 respectively. This provides a conjectural usage-date range of 1561-93 for the panel.
This can be refined by the fact that the costume dates to the mid-1570s (Arnold, 1988, pp. 22-4). The close association between the materials used in the Phoenix and Pelican portraits suggests that they were produced in the same workshop at the same time, even though there are differences in the costume.
Drawing and transfer technique
Infrared reflectography and surface examination revealed extensive underdrawing in both the Phoenix and Pelican portraits, carried out with a sparkly carbon medium that is probably charcoal black. The face appears to have been traced from a pattern and then reinforced with a liquid medium and a brush. Comparison of tracings revealed that the same face pattern was used (in reverse) for both portraits. The position of the face in the Phoenix portrait was raised a little during execution. The initial drawing for the features and the outline of the forehead are clearly visible using infrared reflectography. After the initial flesh layer was laid in, the face pattern was traced again in a slightly higher position. The general shape of the hand in both portraits also appears to have been traced using the same pattern, but some changes were made to the position of the fingers at the painting stage.
Relevance to other known versions
Closely associated works:
- Mirror image of the Pelican portrait, Walker Art Gallery Liverpool, WAG 2994
- Nicholas Hilliard miniature (NPG 108)
- Rothschild Collection
- Anglesey Abbey
- Dumbarton Castle, DUM176
- National Museum of Wales, Cardiff, NMW A 3738
- Kunsthistoriches Museum, Vienna
- a later copy in the collection of James Berg, Chicago, Illinois in 1981
- a later copy in the collection of Lord Kilmarnock, bought by W. Sabin Christies 15 July 1955
Arnold, Janet, Queen Elizabeth’s Wardrobe Unlock’d, 1988, pp. 22-4, 74
Beard, Charles, ‘The “Pelican Portrait” of Queen Elizabeth’, Connoisseur, vol.92, 1933, pp. 263-4
Hearn, Karen, ed., Dynasties: Painting in Tudor and Jacobean England 1530-1630, Tate, 1995, pp. 80-1 (cat. no. 34), 235-7
Hilliard, Nicholas, The Arte of Limning, eds., R.K.R. Thornton and T.G.S. Cain, 1992
Strong, Roy, Portraits of Elizabeth, London, 1963, p. 60
Strong, Roy, Tudor and Jacobean Portraits, London, 1969, pp. 101-2
Strong, Roy, Gloriana: The Portraits of Queen Elizabeth I, 1987, pp. 79-83
Compare Images (what's this? )
Compare high-resolution images against the painting - mainly x-ray and infra-red photography images, but sometimes UV or raking light images - side by side with the ability to zoom in on details.
The panel has a number of old splits, which have been glued, filled and restored (see Support). The paint surface is in a stable condition, although a considerable amount of restoration is evident along either side of the splits and joins. The filling in these areas appears raised and the retouching is glossy. The varnish has a patchy, slightly cloudy appearance and has considerably yellowed. The mismatched restoration and degraded varnish layers are distracting features of this painting. The original paint on the face is considerably worn and overpainted.
Full treatment was undertaken in 2014.
Number of boards: 3
Panel Orientation: Vertical
Panel condition observations
A number of splits are present in the wood. The largest is approximately 340 mm in length and is present in the middle of the central board, through the sitter's face to the neckline. Additional splits are present in upper and lower right-hand corners and along the lower edge of the central board, approximately 30 mm from the left-hand panel join. The joins and non-original splits have been reinforced with wooden buttons on the verso. The central board has been thinned, exposing woodworm channels, particularly alongside the central split.
The boards appear to have been separated in the past, before being glued, re-joined and the wooden buttons adhered. It is clear that three horizontal (non-original) batons were once attached to the verso along the top, centre and lower edges of the panel. The removal of these, and the consequential sanding of the reverse, has partially exposed a wooden dowel in the upper portion of the left-hand join, from the back.
There are pinholes through the wood, and also through paint and ground, along the lower edge. It has been suggested that these may have supported a rectilinear net to aid the production of an engraving (see report from Tate: Rica Jones and Joyce Townsend).
An examination of tree rings, which can help to provide the earliest possible felling dates for the wood used for the panel. The technique can also indicate the geographical origin of the wood.
Number of boards: 3
Last date of tree ring: 1549
2005 analysis carried out at Tate by Ian Tyers, when the last tree ring identified, board B, was found to date from 1549. Boards A and C could not be dated at this time.
In 2010 further analysis was carried out by Ian Tyers. The boards all derive from trees from the eastern Baltic region. It was very interesting to find that the boards of this portrait and those of the Pelican portrait of Elizabeth derive from the same two trees (see Dendrochronology, 67/2010(1)).
For the purpose of analysis the boards were labelled A to C from the left, from the front. As no sapwood was present on any of the boards a terminus post quem can be applied to the panel. The series of rings obtained were compared and boards A and C were found to match material derived from boards A and C of the Pelican portrait, (67/2010(1)); these 4 boards all derive from a single tree. The ring sequence on board B was found to match well with board B of the Pelican portrait and undoubtedly also derives from the same single tree.
The last ring present on board B of the Phoenix portrait was dated 1549, but the last ring on board C of the Pelican portrait, (67/2010(1) (derived from the same tree as the Phoenix portrait's boards A and C), was dated 1553. Adding the minimum expected number of sapwood rings to these suggests that board B on both panels was derived from a tree felled after 1557, and that boards A and C from both panels were derived from a tree felled after 1561. As this picture is undated and it is unlikely that the boards have been trimmed, it is appropriate to apply an eastern Baltic 8-40 year LEHR-usage range to this panel. This provides a conjectural usage date of 1561-93 for both this portrait and the Pelican portrait.
A technique used to identify changes in composition beneath the surface of the paint layers and to understand the physical structure of a panel or canvas.
Four x-ray plates were made to record the composite whole, but as two plates cut through the head, a separate plate was made in order to show the whole head.
X-ray shows that there are dowels present between the boards, two across each join, towards the upper and lower edges of the panel. On the left board, seen from the front, there is a dowel 83 mm from the top and 75 mm from the bottom edge, on the right board there is a dowel 75 mm from the top and 9.3 cm from the bottom edge (see x-ray mosaic 01).
Broad diagonal brushstrokes from the application of the priming layer are evident in x-ray, particularly in the background. The x-ray also clearly shows the overall density of the flesh paint to be greater in the face than elsewhere. This supports additional evidence that the face was initially laid in with flesh colour at a very early stage, before the face pattern was moved upwards with additional underdrawing. A number of changes are evident in the positioning of the fingers at the painting stage. It appears than an initial reserve was left for the body of the hand, thumb and forefinger. The fifth finger was then applied after the costume and multiple changes can be seen in the positioning of the third and fourth finger. This confusion has lead to the presence of six fingers in the x-ray, which were then painted out before the final position was decided upon.
A technique used to observe the layers beneath the paint surface which can reveal underdrawing and changes to the initial design.
Infrared reflectography shows extensive underdrawing in a carbon-based medium (see IRR mosaic 01). The face appears to have been traced from a pattern and subsequently reinforced with liquid medium and a brush. The head was initially underdrawn in a slightly lower position than is seen in the final painting. After the initial underdrawing was applied above the priming, the face was laid-in with a layer of lead white-rich flesh colour. The initial drawing for the features and outline of the forehead are clearly visible as dark lines using infrared reflectography. After the initial flesh layer was laid in, it is clear that the face pattern was traced again in the slightly higher position, but does not appear to have been reinforced with liquid medium except in the face outline. The altered outline of the pearl on the forehead can be seen using infrared reflectography. The hair appears to have been underdrawn after the shift in the position of the face was made.
Examination of tracings has revealed that the same face pattern was used, in reverse, for both the Phoenix and Pelican portraits of Elizabeth (see 67/2010(1)). The drawing in the hair in the Pelican and Phoenix portraits also appears very similar, suggesting that a pattern was used. The general shape of the hand appears to have been transferred using a pattern. No distinct underdrawing is evident in the hands using infrared reflectography, but the similarity in shape of the reserved area in both the Phoenix and Pelican portraits suggests the same hand pattern was used for both works. Infrared reflectography shows a number of changes occurred in the position of the fingers at the painting stage. Following on from the initial reserve, it is clear that the index finger and thumb were extended. Shifts in the positioning of the third and fourth fingers are also clearly visible. In addition, it appears that the fifth finger was initially painted curled around the rose stem, close to the third and fourth fingers, but was altered during the painting process, above the costume paint to the more open position as seen in the final painting.
Individual pearls and jewels are carefully and comprehensively underdrawn using liquid medium. Freehand liquid underdrawing is also evident in the costume, delineating the folds and clasps in the sleeve puffs. It is clear that the drawing for the folds or clasps were not particularly closely followed at the painting stage.
Reserves for the rose and headdress are clearly evident using infrared reflectography.
Investigation into microscopic pigment samples or samples of other media in order to help with dating, to reveal the order of the paint layers and help to understand the painting techniques used.
Paint samples were taken to analyse the pigments and layer structure of the painting in July 2010.
The painting was prepared with a natural chalk ground, followed by a lead white priming layer. A trace of red lead was identified in one sample of the ground, which most likely relates to the formation of a lead soap, rather than added pigments; it was not seen in additional samples taken. A fragment of black underdrawing was present in a dispersion from the hair (sample 7). This has a silvery grey appearance and may be graphite.
Vermilion was used in a very restricted manner, forming bright points of scarlet and details such as the wings and fire on the phoenix jewel. Red lake was used in a number of areas, as a glaze on the rose and for details in the jewels.
Sample 1: Cross-section taken from the golden embroidery on the dress which showed the preparatory layers and paint above. Above the priming, the black of the dress could be seen, followed by a thick layer of red lead and a little red ochre. A highlight of lead-tin yellow was present above this. Sample 2 also showed a similar structure. The dark grey/black costume layer appears to have a purplish tint, and appears to contain a little red pigment with the black.
The hair appears to have been first laid in using sienna, followed by a dark tone of lead-tin yellow, perhaps mixed with yellow ochre, for the hairs. Pale lead-tin yellow was then used for the highlights in the hair.
Sample 5: The bright orange found in the fan jewel was sampled and found to contain a mixture of yellow ochre and sienna.
Sample 4: The unusual light green pigment in the rose was very smooth, with occasional dark green spots. In dispersion, this had the appearance of smoothly rounded and fairly evenly shaped particles. This was identified as the artificial copper green pigment green verditer. This unusual pigment was also found on the Pelican portrait of Elizabeth I (67/2010(1)). The darker green particles were identified as verdigris. An interesting discovery was the use of azurite in the lower part of the leaf stalk. Although not sampled, this blue is clearly identifiable and may be mixed with yellow ochre or an organic yellow to give a greenish hue. It is puzzling why azurite was used for the lower end of the stalk, and verditer and verdigris for the remainder.
Jewel on headdress
Sample 6: Given the presence of azurite in the lower portion of the rose stem, one might have expected the blue jewel in the headdress to have been carried out using the same pigment. However the sample taken from this area found it to consist of lead white and indigo. It is also worth noting that indigo was employed extensively employed in the Pelican portrait as the blue underlayer for the background.
The opaque background is largely composed of red ochre with a dark red lake and black.
An examination of the construction of paint layers, glazes and condition often using a microscope. This method provides important evidence concerning an artist’s technique and paint handling and can reveal a specific artist’s characteristic painting style.
Painting style and method
The techniques adopted in this painting display the artist's knowledge and skill in the manipulation of oil paint to a very high level. The attention to detail is noteworthy, particularly in areas such as the jewels, where extremely fine brushstrokes were used to depict small fruits and gemstones. Despite the level of detail, the technique is reasonably painterly, particularly in the fan, ruff and some parts of the hair, where wet-in-wet blending can be seen. It is likely that the painting is the result of workshop production, rather than having been painted by a single hand.
The panel is prepared with a white chalk ground followed by a lead white priming. There is extensive black sparkly underdrawing present and some of this can be seen through the thinly applied paint above (see micro 13).
The position of the forehead and the features have been raised a little. These were placed a little lower and slightly to the left when the underdrawing was carried out. The outline of the face on the left side was drawn and painted slightly further to the left (see x-ray mosaic 01). The hair and ruff were painted over this initial left outline after the position of the face was raised a little. The flesh was laid in at an early stage in the painting process. Underdrawing can be seen beneath the paint layers and the underdrawing plays a role in defining the line of the nose. The face was fully painted in the first position with detailed eyes, mouth and nose. This can be seen with surface examination, x-ray and infrared reflectography. The first position of the eyes is noticeable in the paint surface where the paint has become more transparent with time (see micro 14). The face paint appears noticeably more dense in x-ray than the rest of the paint surface due to the additional paint layers applied when the position of the face was altered. The face is painted with a mixture containing lead white, red lake, vermilion and charcoal black and there may be a little smalt also. The flesh paint is very pale, due probably to fading of the red lake. The current line of the chin is somewhat confusing where the paint surface has a spotty appearance due to overpaint and restoration applied over the detailed lace edges to the ruffled collar.
Surface microscopy suggests that discoloured smalt is present in the whites of the eyes (see micro 03). Understanding the paint layers and pigments is problematic, due to the thickness and discoloured nature of the varnish. There is considerable brown restoration in the eyes. The original paint mixture in the iris appears to contain smalt with a thin, medium-rich brown over the priming layer. This appears to contain black, red lake, perhaps some vermilion. The eyes are defined with a thinly applied mixture of earth pigments, charcoal black, red lake and a little vermilion. The paint in the tear duct area contains a high proportion of red lake, vermilion and black, applied wet-in-wet to the flesh paint below (see micro 02).
The hair was painted soon after the face. The paint was laid in first with a thin, medium-rich warm brown containing sienna, red lake and black. The outline of the forehead was raised slightly at this stage, when the position of the head was raised a little. Some underdrawing for the curls lies under the current edge of the forehead. The curls were painted at a later stage, after the headdress and the background, using a very fine brush. A thicker application of medium-rich brown, with a high proportion of black, was used for the shadows. The highlights and individual hairs were created with lead-tin yellow and with lead-tin yellow mixed with sienna, yellow ochre, red lake, vermilion and black. The brushstrokes were applied wet-in-wet using a combination of very fine blending and precise, unblended strokes (see micro 05). The hair seems to have been finished at a late stage in the painting process.
The black paint of the dress was applied with a mixture containing lamp black and fine red particles which have the appearance of red ochre. A reserve was left for the white 'puffs' on the sleeves and bodice, and for the hand, rose, fan, ruff and cuff (see IRR mosaic 01). A very thin flesh colour lies under the blackwork partlet. Above this, a thin dark grey, with black and white, was used to create shadows and folds, before the blackwork was painted. The blackwork appears to vary in tone, the areas in slight shadow appear to contain more red ochre particles than the fully lit parts, such as over the shoulders, where there appear to be less red particles. The fine white network over the blackwork partlet was painted with lead white applied with a fine brush, after the background was painted. The network is convincingly three dimensional, in particular where it stands up against the background on the shoulder on the left (see micro 09). The ruff, cuff and underskirt were laid in with a thin mid-grey, after the blackwork detail of the partlet was applied. The priming shows through as a mid-tone in grey passages of the costume (see micro 25). The lace on the ruff and cuff was painted with a small brush, often using small dabbed strokes which softens the threads (see micro 07). This is unlike the technique used for the ruff and cuff on the Pelican portrait (67/2010(1)) from Walker Art Gallery, Liverpool, where the lace is executed with crisp brushstrokes. The gold embroidery on the dress was applied over the black layer, and after some other details had been applied. This can be seen below the rose, where the embroidery paint slightly overlaps the green sepal. The decoration of the dress is finely executed and convincingly three dimensional. The embroidery threads appear to have been laid in thinly with a mixture of red lead, earth and possibly a little black. Above the first layer a red lead mixture with lead-tin yellow in varying proportions was used to create the outline and detail for the embroidery; above this, lead-tin yellow was added for highlights. The red lead has deteriorated and become more white, which has changed the overall tone of the embroidery from a warm orangey red, to an orangey brown (see micro 28). The pearls were enlarged slightly with a final thin paint layer which overlaps the black costume paint. The white puffs and the jewels, the pearls and the underskirt were painted after the black costume. Red lake was applied over the grey underlayer on the underskirt. Much of the red lake has been removed, but some of the crimson colour remains at the lower edge, where it has been protected by the frame rebate (see micro 24).
A reserve was left for the rose when the dark costume paint was applied. The rose was increased a little in size and the outline is painted over the costume paint. The rose was painted by laying in lead white, and then red lake was applied over it. A mixture of black and red lake was applied over this, wet-in-wet. The stamens were dabbed in using lead-tin yellow in two shades, one pure and the other with some yellow ochre added. The stem and leaves are painted with verdigris and green verditer, which was also identified on the Pelican portrait (67/2010(1)) in the enamel on the carcanet at the neck, on the armlet and jewelled enamel on the fan (see micro 19). Unlike the rest of the stem, the green of the end beneath the finger and thumb was painted with a mixture of azurite, lead white, a little black and some yellow ochre (see micro 18). One of the leaves was painted over with a white 'puff', before a finger was painted over the puff (see micro 22). A partially painted out stem, with the same green as the rest of the stem, can be seen beneath the much larger final stem (see micro 18).
With surface examination, the underdrawing can be seen faintly beneath the paint surface. Both infrared reflectography and x-ray show that there have been some changes in the hand, but the changes are most clear with infrared reflectography; changes can be seen with surface examination but are far less obvious. The shape of the reserve for the hand is very similar to shape of the hand in the Pelican portrait (67-2010(1)). The flesh paint was applied after the costume was painted, but several changes were made in the position of the fingers when the flesh paint was first applied. The tip of the thumb was extended, but then shortened and painted over with dress paint (see micro 21). The tip of the forefinger was extended. The third finger was extended in order to curl up to hold the rose stem, and the first position for this finger was painted out with dress paint. The little finger is not evident in the underdrawing and no reserve was left for it when the costume was painted. The little finger was painted over the lowest white costume puff. The curled third finger is painted over the white puff above (which had itself been painted over a green rose leaf). The outline of the hand was extended a little with the final paint layer, and overlaps the costume paint.
Jewels and phoenix
On the jewelled collar, the enamelled roses were laid in with grey and white, and a red lake glaze was applied over the outer petals; the red lake glaze has flaked off in most parts (see micro 11). The small jewelled fruit and the enamel work on the carcanet (jewelled necklace) are very finely painted with careful detail with a small brush (see micro 29). In some parts the priming layer was left uncovered in order to vary the tone. The underdrawing plays a part in the definition of some details. The phoenix is subtly painted with small dabbed brushstrokes, in a characteristic technique which compares closely with the execution of the Pelican portrait (67/2010(1)). The red of the wings and the fire are painted with vermilion (see micro 17). A blue mixture of lead white and indigo was identified in the headdress and necklace (see micro 29).
The fan appears to have been laid in with a pale grey with brushy strokes. These are painted with lead white and the shadows with dark grey, all applied wet-in-wet (see micro 27). Some of the threads of twisted gold embroidery on the dress were added above this paint.
The priming shows through as a mid-tone in the headdress, where the dark brown/black has been very thinly painted. This paint is thickly applied in areas of shadow. The red jewels were first underpainted with vermilion, with red lake over it. The highlights were painted with lead-tin yellow. The blue jewel in the headdress was painted using a mixture of indigo and lead white. The pearls were painted with black and white. The headdress was painted after the background and after the initial thin paint layer of the hair were applied.
The very fine fabric of the veil was painted using a thin skim of lead white over the background colour.
The background was painted directly over the priming with mixtures of red ochre and black. It was painted after the costume.
Order of construction
- Face painted and first layers for hair laid in
- Position of the face raised; eyes was drawn in new position and then painted
- Black dress
- Ruff and cuff laid in
- Hand, rose, jewels and phoenix
- Embroidery and pearls on black dress
- Changes in hand, rose leaf and stem, and costume in that area
- Further paint applied to black costume later in the painting process for modelling, local detail and outlines
- Also probably further paint on some outlines on the background
- Headdress painted after the background and the initial laying in of the hair
- Curls on hair painted at a late stage, after the headdress and the background
- Fine veil painted over the background, and fine open lace over blackwork undershirt applied after background
The palette used is interesting and compares closely with the Pelican portrait of Elizabeth (67/2010(1)): lead white, lamp black, vermilion, red lake, red lead, azurite, smalt, indigo, green verditer, verdigris, lead-tin yellow, yellow ochre,earth pigments
Changes in composition/pentimenti
The position of the head was raised after it had already been painted in. A rose leaf was painted over with a white puff. Fingers were also altered, and one painted over the white puff.
Abrasion and loss in the face has been restored to a greater extent than elsewhere in the painting. Overall the level of restoration present is low. Restored losses down the panel joins and splits are poorly matched and glossy.
A method which helps to reveal past restoration and the extent of varnish layers.
Examination in ultra violet light revealed the extent of the varnish layers present; the varnish is very opaque and thick and has a greenish fluorescence (see UV 01). Retouchings along the panel joins and splits are clearly evident in ultra violet light, as are a number of small, isolated losses in the face, chest and hand.