In 1973, a snippet of the Shroud was given to the Belgian textile expert Gilbert Raes for analysis. It was described as shaped as an irregular right-angled triangle, with sides 40mm, 13mm and 42mm. The hypotenuse was the actual cut, from the end of the Shroud, up one of the herringbone ‘spines’ and diagonally across to the side, between two of the ‘ribs.’ Most of the material was folded up in the rolled seam and the end-hem. Photographs taken at the time show the area remarkably clearly.
The area of the Shroud is instantly recognisable. We are looking at the warp-faced (image) side of the cloth, with a short edge immediately below (with a modern hem just visible in the upper photo), and the long edge to the left, here missing and represented only by the tabby-woven Holland cloth. The remaining piece of Shroud visible will mostly become the ‘reserve’ section of the radiocarbon sample.
Here the Raes sample, photographed on both sides, appears in Shroud Spectrum 38/39, March/June 1991, heading an article by Gilbert Raes himself called “The Textile Study of 1973-1974.”
The heavy rolled seam has been somewhat flattened, and the effect of the light, coming from the bottom edge (in both photos), makes it almost invisible compared to the top photo above. This is worth pointing out, as in 1983 John Tyler, also writing in Shroud Spectrum (Issue 6, ‘Looking at the Turin Shroud as a Textile’), had illustrated his article with this:
The upper picture is the weft-faced side, which on the cloth has no image, and very rarely thought of as the face, while the lower picture is from the image-side, which is very rarely thought of as the underside. Actually, in order only to raise one set of heddles at a time during weaving, the Shroud was probably woven with the weft face towards the weaver, so Tyler is not wrong, but the captions are confusing.
In order to investigate the sample thoroughly, Raes cut the thread holding the hem and seam and separated the pieces. Examining the photos above we can see that the hem (at the left and right side of the photos) is rolled. The warp face, with its longitudinal bars, is clearly visible turned over in Figure 1. The seam is more controversial, in that it is not clear what the weave of the underside of the seam is. It does not look like the hem. Assuming it was as described by Mechthild Flury-Lemberg, who used a diagram from Hero Granger-Taylor’s Masada report to illustrate it, we may reconstruct the folds thus:
The two pieces removed from the sample can then be reconstructed (approximately) thus:
Raes then measured the weft density of both pieces to be 25.7 threads per cm, and the warp density of Piece 1 as 38.6 threads per cm. Pulling Piece 2 away from the sample had distorted its warp threads too badly to be measured. He then extracted some pieces of thread from each piece, and fibres from each thread, which he photographed and commented on.
Thibault Heimburger (‘The Origin of Rogers’ Raes and Radiocarbon Samples,’ at shroud.com) summarises the history of the Raes sample. In 1976 it was returned to Turin, in 1979, Luigi Gonella sent a selection of 14 threads to Ray Rogers, and at an unknown date, Eugene Nitowski took photos as well. By the time of the 1988 radiocarbon sampling, the two pieces were in a sorry state, as can be seen in the video of the occasion. Before the radiocarbon sample was cut, the Raes pieces were carefully lowered into their place, for no apparent reason but we can use the event as an opportunity to check on their condition.
Here is the photo Ray Rogers took of the threads he was sent, still in their plastic bag.
These threads can be broadly divided into two types, long thin ones (#2, #3, #4, #7, #8, #11, #13), with little ‘bridges’ equally spaced along their lengths, and shorter, fatter ones (#1, #5, #6, #9, #10, #12, #14), with no bridges. On a piece of replica cloth handwoven for me, the first type are warp threads and the second are weft.
This corresponds well with any close-up view of the Shroud, where it is clear that the weft threads are more or less straight, while the warp weaves over and under them.
This also explains why the underside of the cloth looks rather different from the top.
Given Gilbert Raes’s assessment of 25.7 (weft) and 38.6 (warp) threads per cm, this implies a distance between the humps of each bridge of about 2.1mm, and an overall length of the longer threads between about 25 and 30mm. By comparison the shorter threads seem to be about 15cm, which is about right for a weft thread from Piece I.
All this information contrasts with Ray Rogers’s assessment, as quoted in Thibault Heimburger, ‘Cotton in Raes/Radiocarbon Threads: The Example of Raes #7,’ at shroud.com. Rogers also observed the two types, and that one type showed “distinct, periodic bends. They correspond to the 1:3 spacing of the weave, and they were compressed into the yarn segments. They are almost certainly weft yarns. The straight segments are almost certainly warp yarns (…) which were held under tension during weaving.”
Although he is correct about the tension, it is apparent that weaving does not produce the observed pattern, but encourages the warp to wrap itself tightly around the weft. Correspondence with Ruth Gilbert, who wove my piece, suggests that, somewhat counter-intuitively, the warp threads, which are under much greater tension than the weft, have a tendency to spring back when the tension is released, buckling as they relax, which is why the warp takes up the form of the weft rather than the other way round. What can be confidently deduced is that the distinct difference in the two threads is the result of a loom-weaving process, and certainly not of any interweaving using a needle and thread. It conclusively refutes the idea that this area was “rewoven” by hand.
Heimburger goes on to observe “7 or 8 bends” on #7, which he compares (via a strangely misaligned diagram) with the number of kinks in a weft thread on the photo of the back side of the Raes sample. I think he is mistaken.
However, the photo cuts an unknown length of thread off, on the right, and it would have been better to look at #2, #3, or #4, which are shown in entirety.
Individual threads were photographed by various people at different times. Heimburger publishes #1, #6, #9, #11 and an unlabelled pair, all taken by Rogers, the last three against rulers (©2002 Raymond N. Rogers Collection, STERA, Inc.). They can be matched against the collective photo above, but it looks as if #6 and #9 have been mislabelled. They are probably actually #9 and #8 respectively.
Whatever. Ray Rogers does not seem to have explored his threads much until 2004, when the suggestion that the radiocarbon corner was a reweave was made by Joe Marino and Sue Benford. The ensuing investigation led to a flurry of papers and presentations, the most notable being Rogers’s paper in Thermochimica Acta, ‘Studies on the Radiocarbon Sample from the Shroud of Turin.’ Rogers does not specify which threads he studied himself.
Next comes the 2008 Ohio Conference, Perspectives on a Multi-Faceted Enigma, at which Robert Villarreal presented his findings on three threads Rogers had asked him to investigate, in ‘Analytical Results on Threads taken from the Raes Sampling Area (Corner) of the Shroud.’ His PowerPoint presentation is at shroud.com.
Villarreal was first given Raes #1, which he subjected to a series of microscopic investigations.
He began with X-ray Photoelectron Spectroscopy (XPS), from which he derived a table of the proportion of elements at each end of the thread, called the ‘twisted’ and the ‘fuzzy’ ends.
This is not easy to interpret. Flax and cotton are largely made of cellulose, C6H10O5, which only contains one more atom carbon atom than oxygen. If the sample were pure cellulose, an oxygen reading of 24.37 should produce a carbon reading of 29.60. If, as seems the case, the sample were contaminated by quartz (SiO2) and limestone (CaCO3), the carbon value for that value of oxygen would be even lower. One explanation could be in the Gum Arabic (C12H36) which Rogers found in abundance on his threads, or even the terpene (C5H8), which was found by Robert Villarreal. As hydrogen was not measured, any addition of gum arabic would simply add carbon, although there would have to be quite a lot of it. Taking just the fuzzy end, every 100μg of cloth would have to be contaminated with 150μg of gum, 37μg of quartz and 16μg of limestone to produce the figures quoted. Even allowing for other natural constituents – the nitrogen for example – this does not seem realistic.
The next examination of #1 was by Time-of-Flight Secondary Ion Mass Spectrometry (ToF SIMS), during which the thread fell apart into three sections, a long fuzzy end, a short twisted end, and a sheath, holding them together. It is interesting to compare the photos we have of the broken pieces with three previous photos of the intact thread.
They have been resized and rotated so that the ‘fuzzy end’ is on the left, but not, as can be seen from the text, mirrored. From the top:
– Photo of #1 in pieces, from Robert Villarreal’s ‘Analytical Results’ presentation.
– Photo originally from Ray Rogers and Anna Arnoldi’s ‘Scientific Method Applied to the Shroud of Turin.’
– Photo of all the Raes threads, by Ray Rogers, from STERA.
– Photo of #1 intact, by Ray Rogers, from STERA.
On the basis of the second photo, Rogers and Arnoldi write: “The Raes samples show a unique splice. Raes thread #1 shows distinct encrustation and color on one end, but the other end is nearly white. The photograph was taken on a 50% gray card for color comparison. Fibers have popped out of the central part of the thread, and the fibers from the two ends point in opposite directions. This section of yarn is obviously an end-to-end splice of two different batches of yarn. No splices of this type were observed in the main part of the Shroud.”
The lower photo, which is much clearer, could also suggest colouration at the ‘twisted’ end, but if “the fibers from the two ends point in opposite directions” is meant to mean that they had opposite twists, then they are clearly mistaken. There is nothing to suggest that we are looking at any kind of splice of two separate threads.
The upper photo does not suggest three distinct pieces either. The longer end is almost as long as the unbroken thread, the shorter end about a third as long, and the ‘crust’ a little longer than that, which suggests that the two ends overlapped, and all that happened was that the fibres of the thread pulled apart. The ‘crust’ was not an adhesive. It may have been part of a stain of that part of the cloth, perhaps from an attempt to colour the Holland cloth to match the Shroud. Its presence on the twisted end may help to explain the huge proportion of carbon found in the XPS results.
Anyway, the split saving been observed, the three pieces were subjected to the Tof-SIMS analysis by Doug Farr, who noted “Both positive and negative SIMS spectra show simple, low mass species expected for cellulose and other organic fibres like cotton. Impurities detected were Na, F and Cl. No consistent significant differences were observed in the surface chemistry between the thin and fuzzy ends of the fibre.” This is quoted on Bob Villarreal’s PowerPoint slide. In his lecture, however, Villarreal quoted him, “This analysis that we are looking at looks very much like cotton. I don’t see anything that looks like flax.”
The best way to resolve the query was via Fourier Transform InfraRed spectroscopy (FTIR), a magnificently accurate and precisely targeted method of identifying chemicals made of the same thing – but in different proportions and in different chemical configurations. This is because the infrared rays interact with pairs of atoms and the bonds between them as a whole, not just the individual atoms. It can distinguish between, say, a carbon atom bonded to an oxygen atom by a single bond (C-O) and a carbon atom bonded to an oxygen atom by a double bond (C=O). Cotton, for example, being 90% or more cellulose, has very few C=O bonds, while linen, whose cell walls are intertwined with hemicellulose and pectin, has more.
The FTIR analysis of Raes #1 was carried out by Kevin Hubbard, and begins with a slide of Linen and Cotton ‘Standards,’ against which the fibres of the Shroud could be compared:
This is followed by two representative spectra from each piece of the Raes #1 thread.
At first sight this appeared strong evidence that the thread was made of cotton, in spite of some significant differences on the right hand side. However, from my own researches, it is apparent that the “standard” for linen (labelled “Warren” in the spectra below) is actually no such thing. It looks more like that of an artificial fibre. This began to dawn on Villarreal and his colleagues when two further Raes threads (#7 and #14) produced similar spectra, and then various other threads from processed linen (the “Hofstra” sample), the Dead Sea scrolls, a mummy and even a couple of Shroud threads from other sources (labelled “Tama 4” and “Riggi HF”) were all tested and found to look more like their “cotton” control than their “linen” one.
The last diagram is a composite of three that were compiled by Jon Schoonover, one of Villarreal’s colleagues, whose paper was read out by Villarreal at the St Louis conference in 2014. By now, it was apparent that the threads could not be distinguished as either cotton or linen on the basis of their FTIR spectra alone. Even Schoonover’s statement, “However the samples studied did retain the C=O absorption characteristic of modern unprocessed linen – a feature not observed in any of the cotton standards studied (nor in the literature),” must be regarded with a certain scepticism. C=O absorption peaks, although small, are indeed a characteristic of all cotton FTIR spectra.
Villarreal’s next investigation was by Radioisotope and Tube Excited Micro-spot Energy Dispersive X-ray Fluorescence Spectrometry. A tiny scrap of the #1 thread was analysed for individual elements, and found to be uniformly distributed with Calcium, Iron, Potassium, Phosphorus, Silicon and Sulphur. what if anything, this might mean was not explored. Aluminium, which would have been seen if present, was not present. That instrument was unable to say anything about Sodium, but Villarreal said it had been detected by a different means. Of the first seven elements, about three quarters was Calcium, a tenth was Potassium, and less than 5% was, in decreasing order, Iron, Silicon and Sulphur, Phosphorus, and Aluminium (0%). What it all meant was, and is, anybody’s guess.
Meanwhile in 2005, John Brown was sent what he describes as #7 and #14 for microscopic analysis. This is his picture of #7:
He describes it as “a weft thread, R7, at an original magnification of 28X. The thread has a yellow-brown coating with the exception of indented regions which are white. These indented regions are at the intersection with the warp thread.”
This cannot be true. As Rogers’s first labelled photo of all the threads shows, #7 is a warp thread with distinct ‘bridges’ every 2mm or so. The thread above is a weft thread, with very slight indentations where the overlying warp crossed it. Perhaps it is #14, and mislabelled.
Brown only gives a UV-photo of what he calls #14:
This also lacks the distinct bridges of the warp, and is also a weft thread.
In 2008 Thibault Heimburger was given the same #7 to analyse. The first thing he writes of it is “Raes #7 (R7), as received, was about 1 cm. in length and was very tight.” Tight it may have been, but if it was #7 it was originally two or three times as long, so he was only given a fragment. In fact, he describes his thread as “similar to the thread as seen in Brown’s paper. In particular, I was able to see easily the “yellow-brown coating with the exception of indented regions which are white.” If this is true, then he was actually looking at a weft thread, perhaps #14.
Heimburger pulled fibres in clumps from the thread and mounted them on three slides, then counted the number of flax and cotton fibres in four places on each one. On the first two slides, he found a flax/cotton ratio of 60:40, and on the third only flax and no cotton. However, many of the cotton fragments were very short, so that he was sure many of them came from the same original fibre. Overall, he estimated an original ratio of about 85:15.
By squeezing one end of the thread flat, the fibres were distributed into a central mass, and outlying fibres on each side, which Heimburger distinguished as a kind of core surrounded by a sheath, and counted a slightly greater proportion of cotton (about 15%) in the sheath than in the core (about 10%), although the uncertainties in the counting and distinguishing between the two kinds render this conclusion meaningless. Anyway spun threads do not have cores and sheaths.
In an appendix to Heimburger’s paper Giulio Fanti analyses another weft thread from the same corner, this time from the reserve portion of the radiocarbon sample. Here is the thread, about 12mm long:
And here is the end, snipped off and squashed apart.
Fanti thinks he detects 4 cotton fibres among a total of 188, but I can identify several more.This thread, it seems to me, is not unlike Heimburger’s.
From all this it is apparent that the Raes sample is a flax/cotton blend, quite possibly deliberate, the warp almost entirely flax and the weft with 10%-15% cotton blend, combining the strength of flax and the softness of cotton. The question is, is how does it compare with the rest of the Shroud? Unfortunately, no fibres from rest of the Shroud have been analysed in the same way. Heimburger quotes Ray Rogers, however, with reference to sticky tape slides 3AF and 1HB, commenting “Absolutely no cotton could be found in this sample” (in ‘Supportive comments on the Benford-Marino ’16th century repairs’ hypothesis,’ BSTS Newsletter 54, at shroud.com). However both these slides were extensively photographed by Eugene Nitowski and Joseph Kohlbeck, and both sets of photos show cotton fibres.
A paper by Giulio Fanti (‘Statistical Analysis of Dusts taken from Different Areas of the Turin Shroud,’ at https://www.shroud.com/pdfs/ohiofanti3.pdf) enumerates fibres derived from the micro-vacuumings carried out by Riggi di Numana of areas between the Shroud and the Holland cloth in 1978, and of the radiocarbon area in 1988. This is the weft side of the Shroud. The dust extracted was filtered through a cotton pad, and sampled off the pad by sticky tape. No blank pad was sampled, so it is not possible to say what proportion of cotton derived from the Shroud (if any) and what from the pad (if any). Fanti attributes it all to the pad. Different sample areas produced very different amounts of cotton, but the different contributions made by the actual amount of cotton in the Shroud thread, variations in the sampling of the Shroud, variations in the cotton filters, and variations in the sampling of the filters are difficult to assess. From the face, feet and buttocks, for example, about 300 fibres each of various kinds were counted, plus no cotton, 110 cotton and 6 cotton respectively. From the hands and the radiocarbon corner came 120 and 169 non-cotton, and 5 and 14 cotton respectively.
Overall, Fanti counted 135 cotton fibres and 946 ‘Shroud’ fibres, presumably all flax. This is a ratio of flax to cotton of 88:12.
During the Middle Ages reference is often made to a material called fustian, which from Tudor times was very much associated with the working class, and was usually quite a thick, coarse, warm textile, with a flaxen warp and usually a double cotton weft, one layer of which was raised and trimmed like corduroy or velvet. It seems to have had an Egyptian origin, however, and was extensively produced in Italy from the twelfth century, beginning as a simple twill, lighter, and of higher status. An inventory of St Paul’s Cathedral from the end of the thirteenth century includes a “casula de fustian” and a “casula alba de fustian.” (Sir William Dugdale, The History of St Paul’s Cathedral in London, from its Foundation, 1818) See also The lexis of Cloth and Clothing in Britain Project of Manchester University, which has several early references to fustian.
If the Shroud was made of two types of thread, then it is not surprising that the STuRP sticky tape threads, which lifted material from the warp side, had very little cotton, while the weft threads analysed by Heimburger and Fanti had considerably more.