Note to the reader: No material presented here may be reproduced by any means whatsoever without permission by the author.  Further information may be obtained by contacting the author. All photographs unless otherwise indicated are by the author.

To avoid confusion, the reader is reminded that the first part is the first report concluded in April 2001. The successive parts contain additional material and some clarifications in light of new information. Footnotes are at the bottom of this web page. I have added a few notes which are in parentheses.

I refer to this discovery as "Teri's Find". In my view it is appropriate as the owner Teri Horton deserves every credit for it. That it is now presented in this fashion is solely a personal gesture and a free gift to the owner Teri who has to my mind persevered in her brilliant and single-minded pursuit of what she perceived as being right - the recognition that the thrift shop painting she purchased for $8 is indeed the work of Jackson Pollock.

This painting and this research has received front page coverage in the international media, TV and radio on the major networks. Some links are provided below.

For security reasons, several images in this report are watermarked in a way that is not apparent to the observer .

The fingerprint images have also been reduced in resolution so as to render them unusable except for illustration.

I advise against evaluating the fingerprint images illustrated in this report as if they were the actual source material. Any attempt to do so is pointless as no valid conclusion can be drawn from such low resolution images except to say that they are too low in resolution to evaluate. In addition, as an added security and control measure, the image of the fingerprint comparison (figure 10) is also watermarked - the nature of which is proprietary and is known only to me.  

The actual fingerprint comparison being used in this case has not been published anywhere for privacy and security reasons. That fingerprint comparison is independently confirmed. To reiterate: any attempt to evaluate the fingerprint comparison published here in figure 10 and to draw conclusions from it is irrelevant as it serves as an illustration only and, for reasons explained in section II, I chose to discard that comparison  in figure 10 back in 2001 and seek other comparisons - the ones not published here.

Although this study is published on the Internet some parts are proprietary and therefore not 'public domain'.

The purpose of this presentation is explained in its title: a Forensic Study in Authentication. For information regarding the fullness of this project contact the owner.

Initially, I have been retained to conduct this investigation. Since May, 2001, I have worked on this project pro bono.

PPB.

Part I.

Teri’s Find

Research conducted by:

Biro Fine Art Restoration

& Forensic Studies in Art

3014 St. Antoine St. West, Montréal, Québec, Canada, H4C 1A5

Phone and fax: (514) 933-2885

e-mail: artsleuth@sympatico.ca

www.birofineartrestoration.com

ABOUT THE AUTHOR

I grew up in the art world. My father was art restorer to the Budapest Museum of Fine Arts, Hungary, for many years and he is also a recognized painter. Early on, I became familiar with the work of many artists, techniques, art materials and styles including Jackson Pollock’s. I now have some 30 years of experience in fine art restoration, forensic and art-historical research.

From 1970 to 1990, I worked in a family enterprise called Center for Art Restoration, in downtown Montreal, Canada. By 1990, my father retired. I continue the vocation with a concentration on scientific and forensic investigation to assist in matters of conservation, authorship and authenticity. Since 1990, I have been concentrating almost entirely on the use of forensic techniques and technology to aid in art-related studies.

Other areas of specialization are optical and electron microscopy which aid the precise identification of art materials[i]. I have been pioneering the use of fingerprint identification in attribution of works of art since 1984[ii] with numerous attributions and discoveries to my credit[iii]. I am also a specialist in scientific photography and digital image processing.

My research on art materials and techniques has been published in such scholarly journals as Mankind Quarterly[iv]. A new major paper will be in print in Antiquity[v], in the June, 2001 edition. I am co-founder of an international interdisciplinary group of scientists[vi] exploring some of the oldest paintings of humankind. In October of 1998, I gave an important paper at the International Sixteenth Century Studies Conference in Toronto on Italian Renaissance techniques of composition and the use of geometry. In addition to traditional connoisseurship, my specializations embrace a forensic approach coupled with modern scientific methods of detection that are not specific to any individual artist but rather encompass elements universal to them – the artistic process. I have given talks numerous talks and was panelist at symposia, most recently at the annual conference of the Appraiser's Association of America, New York, at the annual conference of the Royal Microscopical Society, London, National Portrait Gallery. I will be speaking at the University of Glasgow international conference on Conservation and Authenticity March 24th, 2006.

I have recently published an important paper in the Proceedings of the Royal Microscopical Society, Oxford, March 2006 edition. My forensic research work is now part of a collaboration with the Pigmentum Project, Oxford University, Department of Archaeology.

My fingerprint and investigative work is known to forensic scientists from the RCMP, the FBI[ix] and various police departments as well as to major museums and universities. I am presently working on the manuscript of my first book exploring the potential of fingerprint science in attributing works of art. Further information on my fingerprint work can be found on my web site.

2006 saw the completion of two important documentaries focusing on my work. One, entitled Who the #%&% is Jackson Pollock chronicling this present investigation is already being screened at various film festivals. The other, a two hour Discovery Channel presentation traces the story of a possible new Leonardo da Vinci discovery. It is due to be aired in May, 2006.

A more detailed biography is on the About Us page.

It has been my privilege to work on this project.

Peter Paul Biro

March , 2006

1.      SUBJECT

A painting of an abstract composition in the poured style, executed in emulsion paint and acrylic on cotton
canvas (also called cotton duck) , 66 ¾  x  47 5/8 inches. (Figure 1).

2.      MATERIALS AND CONDITION

Figure 1. Teri's Find

 

 

a)       The composition is painted in emulsion paint (sometimes referred to as enamel) [x] and acrylic. Visual examinations suggest this; however, the medium could be further identified scientifically[xii]. See Revision IV.

For now, that the paint is of the enamel type was determined through visual assessment of the high level of gloss and through prior experience.

 

b)       The painting was never varnished. It is free from mechanical damage, wear, prior restorations or alterations. /The painting is relatively clean.It does not suggest the need for cleaning or any conservation./

 

c)       The canvas is stretched and stapled to a stretcher that appears somewhat weak relative to the dimensions of the painting.

3.      DIMENSIONS and CONSTRUCTION

Figure 2. The arrow may point to the top of the painting.	Figure 3. The calliper jaws at 1 inch placed on the canvas.

a)       The canvas measures 66 ¾  x  47 5/8 inches.

b)       The painting may have been intended to be vertical, indicated by a small arrow marked by a blue ballpoint pen on the stretcher[xiii].  (Figure 2)

c)       The stretcher is of a ‘workshop variety’ constructed from standard 1x2 inch pine stock reinforced with four approx. 8x9 inch squares of 1/8-inch thick plywood at the corners. (Figure 5)

d)       The canvas is white cotton and the full width has been used – a standard 54 inch wide stock. This is evidenced by the presence of the finish on both sides of the weave. The canvas weave is 28 single threads per inch by 36 double threads per inch (11 threads per cm by 14 threads per cm) with an average thickness of 0.022 inch. (Figure 3)

e)       The painting appears cut from a roll after it was painted. The painting had to have been done with the canvas being level with the ground. The cutting is evident along the two long sides. The pouring of paint extended beyond these edges.

Figure 4. Pollock painting, 1949. Pollock normally used white cotton duck for painting. It is interesting to note the canvas facing away from us in the top right corner leaning against the wall. The folded edge is distinctly 'white' and is probably primed. The canvas looks much darker than cotton duck and it is probably natural linen. Compare it to the verso of the painting in figure 4b. That is how cotton duck looks - almost white. It is important to note that there are always exceptions to what one assumes to be 'normal practice' by an artists. Photographs 4-4b: Arnold Newman.	Figure 4a. Pollock and Krasner in 1949 in front of stretched canvases. The stretcher bars appear very similar to the stretcher of Teri's find.	Figure 4b. Pollock in front of a (vertical)   stretcher reinforced just as in Teri’s Find.Note the rectangular form left and right of his shoulders. They are pieces of wood used to reinforce the joints of the stretcher.	Figure 5. Teri’s Find, verso. The corner joints are reinforced with pieces of plywood and the canvas stapled on top. Along the right edge some paint may have bled through the canvas or was picked up from the surface on which it was painted.

f)         Because of the above implications (e), I illustrate here on figure 4 how Pollock painted on a roll of canvas tacked to the floor. He stretched the canvas almost without exception after fully completed[xiv].

g)       Figure 4 also illustrates Pollock’s technique in overextending his pouring movements and thereby the trajectories of flying paint extend beyond the bounds of the canvas. The patterns made on the studio floor are a direct result of this. When cut, the actual painting becomes the ‘central portion’ of the original design. This ‘cropping’ helped, in part, to create the effect of a unified and balanced painting.

h)       On figure 4, to the left and behind of Pollock are visible strips of wood stock, no doubt for constructing his stretchers.  They appear to be 1x2 and 1x3 stock.

i)         On figure 4a, we observe to the right of Lee Krasner, a stretched canvas on »1x3 stock. On the floor is an unrolled run of canvas.  The stretcher is of a simple workshop construction.

j)         Figure 5 shows Teri’s Find from the verso. Here, we see the corners of the stretcher reinforced with pieces of plywood. These would serve to stabilize the frame before the canvas is attached.

k)       Figure 4b shows Pollock in front of his paintings in his barn studio, The Springs, Long Island. Here, on the back of the canvas behind him, we find an analogy for the rectangular reinforcements on the stretcher seen of Teri’s Find. The canvas, here too, is stretched over the rectangular attachments, again, as we see in Teri’s find.

4.      SAMPLES COLLECTED

a)       I collected four scrapings (labelled A-D) of paint from the folded-over edge of the canvas with a surgical probe for later analyses.

b)       I collected six samples[xv] of hair and fibre. The hair samples appear to be human, dark brown in colour[xvi]. Some are covered with paint. Judging their colour is difficult without cleaning[xvii] but generally appears to be dark brown. Some were embedded in the paint layer.  They could be used for comparison should reference material become available later. Their examination is still cursory.

c)       Vial 7 is water used and 8 contains the swab used to remove contamination from the painting for possible later testing.

d)       Vial 9 contains a staple removed from the folded edge of the painting.

e)       The sample collection has been recorded on photographs and on Dictaphone.

5.      Provenance

The painting’s history is sketchy at best. I am informed by the owner that it had been purchased at a country sale in California in 1992 or 1993 by Teri  for very little. Nothing of its ownership is known before this. Nevertheless, a tantalizing clue emerges from correspondence with her son Bill [xviii]:

“Purchased at an antique store in San Bernardino, Ca. The dealer bought it at an estate sale in the Victorville, Ca. area. It was included in a room upstairs in a house at the estate sale. It was wrapped in a big piece of carpet/burlap loaded with dust and spider webs according to the daughter of the owner of the antique store who has since gone out of business and we understand the owner has passed away. About 3 years ago, my mom went back to the location of the store but it was no longer there. She says she talked to the daughter but did not get her name. Mom says she went back there last month but could not locate anyone.

The brother Charles Cecil Pollock supposedly lived in the Apple Valley area close to Victorville area and according to the Social Security index the final check was paid in 1991 or 1992. We believe he died in 1991. The painting was bought in 1992 or 1993. Mom doesn't remember.”

The possible connection with a Pollock family member is certainly a lead worth following up, though with Charles Cecil’s death any information is likely to be circumstantial.

In their catalogue raisonné, Thaw and O’Connor write[xix]:

“Other obviously authentic works could not always be readily traced to their origin. The artist himself, or others, gave some away or sold them without record. Others were privately transferred or inherited. Later such examples entered the art market with their previous ownerships blurred."

This may have been the wayward route of Teri’s Find to California[xx]. Ultimately, we will likely never know for sure.

6.      FORENSIC DATA

Figure 4b. This print proved to have a sufficient number of characteristics for comparison.	Figure 4a. This print, though clearly left when the paint was still wet, proved too poor for use.	Figure 5. Using the 3D approach, it was possible to examine remaining ridges in detail from any angle.

a)       I located and documented forensic evidence on Teri’s Find such as two fingerprints left in wet paint, i.e. coinciding with the point in time of the painting’s creation. One was located along the folded edge of the canvas on the verso. This print is not clear enough for comparison, as it does not have enough discernible characteristics. (Figure 4a)

b)       The second fingerprint was discovered on the verso. Its location has been recorded and photographed. It was deposited with a fingertip that was coated with paint of various colours. The print was photographed with a macro lens at 1:1 reproduction ratio with a medical ring flash unit. (Figure 4b)

c)       The latter recorded fingerprint was found to have 16 identifiable characteristics and it was considered usable for comparison. DIP[xxi] was used to enhance the print and to verify that the observed characteristics were not artefacts of the canvas’ weave. Additionally, sophisticated imaging software[xxii] was used to render the fingerprint image in simulated three-dimensional space to ascertain that all possible recognizable ridge characteristics are recorded. (Figure 5)

d)       During this research, I located and examined numerous fingerprints on reproductions of Pollock’s paintings.

e)       Because of its excellent photographic quality, I concentrated on Varnedoe’s catalogue[xxiii]. On illustration 184-190, (page 282) I noticed numerous fingerprints[xxiv]. The illustrations were scanned from the book at 1200 dpi in full colour and the images examined on the computer using a sophisticated image-processing program[xxv]. I was able to document fingerprints clear enough for comparison. (Figure 9a)

f)         At a scene of crime, or in our case on a work of art, usually only partial prints are found and the presence of all the characteristics is not required. When two prints are compared, they have to reveal characteristics, which are not only identical but are also similarly located in relation to one another. Just as in a judicial proceeding, a point-by-point comparison has to be established. The number of points that have to match is not universally agreed on, however, 6 to 8 points are often sufficient[xxvi].

g)       Upon examining the Berlin images, I noticed that a small partial fingerprint from painting #4 in that series contains ridge patterns that closely resemble a portion of the fingerprint from Teri’s Find. These are reproduced in exploded views on figure 9a.

 

Figure 9. Exploded view of the portion of the fingerprint under examination from Teri’s Find.	Figure 9a. Exploded view of the fingerprint under examination from Jackson Pollock: Untitled (Red Painting Number 4) c. 1950, oil on canvas. Private Collection, Berlin. Number 4 in a series.

h)       When reproduced to the same scale and in the same orientation, in all, 12 characteristics appear identical and in the same relative position to one another. The print from Teri’s find is clear in this particular area but the comparison print, being reproduced from a book is not as clear and it is also partial. Notwithstanding, I attempted a comparison.

i)         First, it was necessary to perform DIP on both images. In the case of the fingerprint from Teri’s find, it was important that the ridge patterns are not confused with the contribution of the background, principally the weave of the canvas. This was achieved through following the colour index of each ridge’s paint-deposit – on the microscopic scale. For better visualization the principal paint colour was then tinted dark green.

j)         In the case of the Red Painting Number 4 fingerprint, it was necessary to remove the ‘noise’ created by the offset dots from the reproduction. This was achieved through DIP, particularly through the implementation of FFT[xxvii] following colour segmentation.

k)       The characteristics I found equivalent are reproduced below on figure 10. Fingerprint A is from Pollock’s Red Painting Number 4 and B is from Teri’s Find.

l)         The characteristics are: 1 bifurcation, 2 small island, 3 bifurcation, 4 small island, 5 bifurcation, 6 bifurcation, 7 ridge ending, 8 ridge ending, 9 bifurcation, 10 small island, 11 bifurcation, 12 bifurcation.

m)     The correspondence of 12 characteristics is confirmation that the fingerprint in A was left on the painting by the same finger as what left its mark in B.

n)       The fingerprint chosen for comparison comes from an undisputed Jackson Pollock. It is recorded and catalogued by O’Connor and Thaw as 304 [Red Painting 4], c. 1950, oil on canvas, 21 ½ x 11 ¾ inches. Its history is recorded as “estate of the artist”.

7.      STYLE AND TECHNIQUE

a)      It is not my intention to duplicate research and eloquent analyses already in print, I shall therefore constrain myself to aspects that are of particular interest in this present case. It is difficult, if not impossible, to articulate the sublime, complex and the chaotic that Pollock created and are so visually evident in Teri’s Find. The closest comparison appears to be Number 5, 1948. This painting is also of vertical format and of similar width (with a difference of 1/8 inch).

b)     Pattern similarity is highly analogous. Both paintings carry final traces of a predominant colour. These appear as the final acts[xxviii] of pouring of the painting process – they are the topmost layer. In both paintings, these trajectories of paint appear as geometric inversions of themselves placed opposite each other in proximity.  In Teri’s Find this appears in silver (aluminium paint) and in Number 5, 1984, it appears in yellow.

c)     Detail density and complexity are virtually identical. This was measured through digital Image processing and the counting of features in arbitrarily chosen equal areas from both samples.

d)       Colour similarity is also highly analogous between the two paintings compared here except for the more generous use of black in Number 5.

e)       Behaviour of paint, mixing and marbling (not only) on the macroscopic scale are comparable or identical on both paintings. This is revealing of the similarity or identity of paint materials used to create both compositions.

f)         Figure 13 demonstrates just how harmonious Teri’s Find is with Number 5, 1948, on the level of small detail as well. The left half of the figure below is an arbitrarily chosen area from Teri’s Find, digitally stitched to a similarly chosen portion of Number 5, 1948 on the right hand side /not to the same scale, which actually would have worked better/. The two details merged, side by side, create an essentially whole and new ‘composition’.

Figure 13. Montage of two details of equal size and orientation. On the left half Teri’s Find, on the right half Number 5, 1948. The right hand side image is somewhat blurred in comparison as it was reproduced from a book illustration. Selection of colours and their relationship to each other are highly comparable.

 

An other important comparison demonstrates the highly similar behaviour of paint (figures 13a, 13b and 13c). For this marbling effect to occur, paint materials have to be of similar or identical composition and quantity of solvent added. Figure 13c is of an area 24x36mm. It is clearly seen here that marbling occurs on a large scale especially in the mixing of the brown and yellow paint.

 

Figure 13a-b-c. On the left, detail of Teri’s Find 24x36 mm. In the middle, an approximately identical surface area from