Home Prev Next

Historiography of the Apparition of Guadalupe

Daniel J. Castellano (2013)


The Image in the Twentieth Century

The Image in the 1920s
Retouches to the Face
Retouches to the Mantle or Tunic
Chemical Analyses
Photographic and Ocular Studies in the Mid-Twentieth Century
The Figure in the Eyes
Callahan Infrared Study
Jorge Sol Rosales' Examination and Restoration
Dubious Claims of Leoncio Garza-Valdés
Synthesis of Evidence

The Virgin in 1897For the first four centuries of its existence, the Image of Guadalupe remained in an excellent state of preservation. Its original features and colors, still vivid and brilliant, remained intact, with the notable exception of the field of golden rays and the disappeared crown. Otherwise, the Virgin of Guadalupe was exactly as she had appeared since around 1600 or earlier.

In the twentieth century, however, there were significant changes to the Image, by natural or human means. We will document these alterations, in order to better interpret the findings of modern expert examinations of the Image. There have been many more such examinations than is generally recognized in English language literature. Some of these have, to date, only been poorly documented or even misinterpreted in English. A scrupulous compilation of these studies has been made Fernando Ojeda Llanes, who transcribed or cited all the contemporary primary sources available. [Fernando Ojeda Llanes, La tilma guadalupana revela sus secretos, Mexico: Porrua, 2005] We will resort to these primary sources, as transcribed by Ojeda Llanes or in their original publications, in order to paint a more complete picture of modern Guadalupan testimonies than has yet been done in English.

The Image in the 1920s

The decade-long Mexican Revolution established a socialistic, secularist regime which brazenly asserted the primacy of the state over the Church. The Constitution of 1917 abolished the privileged status of the Church and banished religion from public education, but otherwise guaranteed religious freedom. In 1920, however, the government began to implement aggressively anti-clerical statutes, at first only locally, limiting the Church's ability to own property, reducing the number of priests per region (sometimes to zero), and forbidding ecclesiastics from supporting political parties or even wearing religious garb in public. Some of these restrictions, including the suppression of religious broadcasting and clerical attire, remained in effect through 1992. More hostile statutes were enforced only locally, but the government's anti-ecclesiastical stance was clear to all.

Aftermath of 1921 explosionIn the context of this rising anti-Catholic sentiment promoted by the leftist government, we should not be surprised that the long-venerated Virgin of Guadalupe was subjected to a terrorist attack, and that the government made no attempt to prosecute the culprit. On the morning of November 14, 1921, a redhaired man among a group of workers left a bouquet on the floor beneath the Image. Moments later, a loud blast was heard in the basilica, and the area surrounding the icon was filled with a cloud of dust. A bomb had apparently been placed on the floor beneath the picture frame, where the marble was pulverized. The force of the blast knocked over brass candelabras and bent a heavy bronze crucifix, still on display today. Amazingly, the icon was completely undamaged. The protective glass (not bulletproof) was not broken or cracked. Yet the concussive wave had penetrated the wall, badly damaging the icon of St. John Nepomucene on the other side, just centimeters behind the Virgin. [Nicolás Mariscal, La Voz Guadalupana 18(7), Sept. 1952, p. 25.] The survival of the tilma was immediately acclaimed as miraculous.

Bent crucifix from 1921 explosionA photo taken after the blast shows a dark spot where the explosive was placed, in the corner between the marble floor and the wall. The candelabras and triptych have fallen forward, showing the direction of the force. The crucifix originally stood on the small altar in the foreground, with its back to the blast. It is bent forward where the pole meets the base, and backward in the middle, making an S-shape. The bent, twisted pole is bronze, though its decorative base and cross are brass. An inscription next to the crucifix reads, "The Son protected the mother."

Within minutes of the explosion, some churchgoers apprehended Luciano Pérez, a railroad worker believed to have placed the bomb (later found to have been Hercules brand dynamite). He was soon taken away by plainclothes police or soldiers, who numbered among the supposed workers visiting the basilica. Pérez was later released without charges, and the government made no serious attempt to identify the culprit. The procurator Eduardo Neri made the government's stance clear, coldly suggesting that the Church benefited from the terrorist act:

...the act in itself does not favor anyone but the clerical element, because this now appears to play, as it has done other times, the role of victim in order to gain public commiseration; now religiously, because a new miracle is exploited; now financially, because they have found, and who knows if not prompted, the Knights of Columbus affiliates, a new basis for organizing pilgrimages which will surely leave them large quantities of money. I esteem that all religious beliefs deserve an absolute respect, but it is repugnant to use them for ignoble ends.

This callous blaming of the victim, peppered with the usual liberal anti-Catholic canards, coupled with Neri's complete inaction, only strengthened Catholic conviction that the government was behind the attack, as evidenced by the presence of plainclothes agents disguised as workers accompanying the bomber. It is not clear whether President Álvaro Obregón, who was relatively moderate toward the Church in comparison with the more strident anti-clericalists, had any prior involvement in the act. A rumor soon arose that the president had raised the idea of such an attack to his stenographer Juan Esponda, but this has never been substantiated.

The providential protection of the tilma did not imply that the relic is absolutely indestructible, nor that precautions should not be taken to preserve it from future attacks. Government persecution of Catholics escalated in 1924, when the unelected atheist president Plutarco Elías Calles took office. Violent groups of liberal democrats were given a free hand to desecrate altars and murder priests, in the name of liberating people from "superstition." In 1926, Calles imposed punitive laws on the clergy, and began to confiscate Church property, close monasteries and convents, and eradicate religious schooling. This resulted in a mass revolt known as the Guerra Cristera (1926-29). During this period of open warfare, the various state governments became more virulently anti-Catholic. In Sonora, all churches were ordered closed. In Tabasco, a priest was required to marry in order to officiate Mass, effectively abolishing Mass. In Chihuahua, only one priest was allowed to minister to the entire state. In a few years, tens of thousands were killed, thousands of priests were expelled, and dozens were tortured and martyred for offering sacraments.

In light of this total war against the Church, the custodians of the Guadalupan Image decided to remove the venerable icon from public view for safekeeping. On July 31, 1926, before a notary and witnesses, the Image was wrapped, sealed, and removed in secrecy from the basilica. Meanwhile, a painted copy by the Pueblan painter Rafael Aguirre was displayed in its place. The original Image was not returned to the basilica until June 1929, when a truce had been agreed in the Cristero War, and the government ceased to enforce its more aggressively anti-religious laws. Other forms of persecution continued, however, for years afterward, including hundreds of executions of rebels and religious schoolteachers.

It is sometimes claimed, without sourcing, that the modernist Mexican painter known as Dr. Atl (Gerardo Murillo, 1875-1964) examined the tilma in 1928, but this is plainly impossible, as the relic was wrapped and sealed at this time. Any comments Dr. Atl may have given in 1928 would have been based on the Aguirre replica, or memories of the original when it was on public display. His purported claim that the tilma was made of not very old cotton is patently untrue of the original, which has a continuous documented history back to the seventeenth century, and has had its age confirmed by modern analysts. This suggests that he was really looking at the Aguirre copy, which had been secretly substituted for the original in 1926. This supposition could account for his singular disparagement of the icon's aesthetic quality: "The painting of La Guadalupana is a parody of an image preserved in Fuenterrabia, Spain; parody in turn of decadent Byzantine images. The Virgin of Guadalupe is a purely decorative work, executed by a mediocre imagination. It has standard technique and painting..."

Virgin of Fuenterrabia The obvious rebuttal to these remarks is to view the original Image, whose quality speaks for itself, and has been praised by numerous artists. Further, we may note that the Virgin of Guadalupe does not look at all like that of Fuenterrabia (shown). As for the judgments that it is "decorative" and of little imagination, these are informed by a modernist bias against realism and natural beauty in art. Religious iconography does not share the modernist aim to present disturbing philosophical ideas, or to be as daring or iconoclastic as possible. To expect iconoclasm in an icon is absurd. It is doubtful that Dr. Atl ever made the inept remarks attributed to him, as they are unsourced, and are used to support the patently false conspiracy theory (contradicted by all expert examinations) that the tilma was replaced by a new cloth in 1895.

When the original Virgin of Guadalupe was restored to her rightful place in June 1929, a notary and witnesses were again present, to attest that the seal had remained intact, and that the Image was the same as that which had been removed in 1926. Once on display, the Image was again available for photography, and some dozens of photographs taken in 1929 by Alfonso Marcue González were published in early 1930.

Retouches to the Face

In the late 1970s and early 80s, Guadalupan investigator Rodrigo Franyutti claimed that the photographs published in 1930 and 1931 showed evidence that the Virgin's face had been altered some time after 1926. In his 1983 pamphlet El verdadero y extraordinario rostro de la Virgen de Guadalupe), Franyutti compares photos from the 1920s with later photographs, showing that there has indeed been visible alteration to the Virgin's face, and not for the better. His claim that all these changes were made by 1931, however, is more doubtful.

Up through 1926, the most detailed photos of the Image were taken by the basilica's official photographer Manuel Ramos. An especially high quality photograph, certified as an authentic copy of the Image, was taken on May 18, 1923. By comparing this photo with those taken in 1929 and later, we may evaluate Franyutti's claims about retouches to the face.

Franyutti correctly notes that, prior to 1926, the Virgin's face was noted by all observers to be radiant. This lustre was tentatively identified by Bartolache's painters as akin to a varnish or soap grease. Whatever the cause, all agreed that there was a brightness to her countenance. Second, the facial features and outlines were delicately and lightly painted. As seen in closeups of the 1923 photo, the Image was so subtle that the weave of the cloth remained clearly visible. Outlines, where used, were only a few threads in breadth, while some features had no outlines at all, but only the most delicate shading. This gave the impression of an Image that had not been painted at all, but imprinted. Lastly, the subtlety of the eyelids and other features gave an expression of tenderness that had long earned the admiration of master artists.

By 1979, all of these qualities had been effaced, due to apparent retouches. These retouches darkened or emphasized certain outlines and other features, detracting from the Image's exquisite subtlety, and losing something of her tender expression. Since the original Image seemed as though it was not painted at all, any added paint gives the impression of uneven coloration. In black and white photographs, this can be seen as a partial obscuring of the cloth's weave. This results in what Franyutti calls a "darkening" of the face, not meaning that the Virgin was originally fair-skinned, but that she has lost a certain luminosity to her face. The features of the eyes have been emphasized with darker shading, losing something of their delicate tenderness. Further, the shading of the chin seems to have been replaced by a second line, making the face look plumper and older than the young maiden depicted in the original. In modern color photographs, it can be seen that the paint added to the Virgin's left cheek is a rouge that contrasts with the rest of her complexion.

Other changes include the widening of the lips and the addition of black paint to the hair, obscuring the hair lines. Also, a line of paint has been applied to extend the outline of the nose, where once the outline was only suggested by the weave of the cloth. In general, the retouching of outlines has given the Image a more artificial and aesthetically crude appearance.

The retouches identified by Franyutti are quite subtle, and difficult to discern in the photographic evidence he produces. Although these retouches all certainly existed by 1979 (as we will see in a study to be examined later), it is far from obvious that they were all present as early as 1929, as he claims. In any event, the face of the Image has certainly been retouched in the twentieth century, at one or more times from 1926 onward.

The Virgin in 1923 The Virgin in 1929 The Virgin in 1931
The Virgin in 1943 The Virgin in 1963 The Virgin in 1979

From left to right, we have (1) a detail of Manuel Ramos' authenticated 1923 photo; (2) an authenticated print from the original negative of Alfonso Marcue's 1929 photo; (3) the official color lithograph issued to commemorate the quadricentennial of 1931; and in the second row, (4) a detail of a 1943 photo (in Franyutti, op. cit.); (5) a 1963 photo; and (6) a detail from the 1979 Callahan study (to be discussed later).

From the above, we can first see that, pace Franyutti, the 1923 and 1929 images are nearly identical, except for differences in exposure. The cheeks have the same colorations, the hair remains light, and overall the face is that of a young maiden with a delicately rendered expression. The 1931 lithograph retains the light hair, but darkens the orbit of the eyes and accentuates the chin outlines, making the face look somewhat plump. The mouth has lost something of its tender expression. Franyutti laments that this inferior image was publicized as that of Guadalupe, but some of these defects may be attributable to the medium. In the 1943 photo (lower left), we see that the original Image has retained something of its youthful, tender expression, especially in the mouth. Still, there appears to have been some enhanced shading around the eyes, especially in the lower eyelids. Similarly, the contour of the nose has apparently been extended on the (viewer's) right side.

The early retouches were quite subtle, but more noticeable changes are visible by 1963. The face is now markedly different, with uniformly dark hair, an apparent double chin, and more pronounced outlines around the nose and lips. The cheeks appear to have more paint applied, so that the Image for the first time has the look of artifice. As recently as 1943, by contrast, we can see that the Image was so subtle as to be scarcely distinguishable from the visible weave of the cloth. It was this effect that caused so many expert witnesses to believe that the Image was miraculously imprinted. By 1979, however, this aspect of the Image was further obscured, by application of paint on the cheek, including an unsightly rouge that can be seen even today.

These alterations during the mid-twentieth century have had a largely negative aesthetic effect, making the face look older and fatter, with an expression that appears somber rather than tender. The featureless black hair and rouge cheek detract from the realism of the original. Most regrettably, the emphasis of outlines and application of paint obscure something of the tilma's greatest marvel, which is its ability to suggest the image of a beautiful face with only the slightest additions to the cloth, even making use of the irregular weave itself to render outlines and shadows.

The worst of these additions appear to have been done from the 1940s onward, somewhat later than what Franyutti claims. Assuming that the motivation for the retouches was restorative, it follows that the Image began to show some signs of decay during this period. By the same token, the fact that these retouches are so noticeable shows that no restoration of the face was needed for the first four centuries. Any retouching would have obscured the weave of the cloth, which was plainly visible throughout the face in 1943.

Retouches to the Mantle or Tunic

The art restorer José Antonio Flores Gómez claimed to have worked on the Image twice, in 1947 and in 1973. At the age of seventy-eight, he recounted his experiences to Rodrigo Vera for Proceso magazine (No. 1343, July 28, 2002, pp. 17-18.). Although he did not work on the face, his testimony complements the probable timeline of retouches we have gathered from comparisons of photographs.

Flores Gómez was first asked to restore the Image as a young man in 1947. When he examined the painting, he immediately noted signs of damage caused by humidity and mold, as well as previous retouches. This deterioration persuaded him that the Image was not miraculously imprinted, though Juan Diego may have been a real person. As Fray José María Tellez Girón noted in 1792, however, the questions of miraculous imprinting and incorruptibility are properly independent, for one does not imply the other. Flores Gómez's testimony really speaks only to the latter question.

The abbot of the basilica, Feliciano Cortés, asked Flores Gómez to fix some noticeable cracks (cuarteaduras) that had appeared in photographs of the painting. Flores Gómez noted that there was a long vertical crack from head to toe, as well as some less visible horizontal cracks. He was certain that the latter were caused by folding, so that at some time in the past the cloth had been folded and this cracked the paint.

The cracks noted by Flores Gómez do not imply any deficiency in the durability of the painting per se, since they are all caused by mechanical stresses on the canvas. The head-to-toe vertical crack is certainly at the junction of the two pieces of cloth, as noted by other examiners. We will find that the remounting of the canvas applied excessive tension at the seam, resulting in cracking. If the horizontal cracks were caused by folding, as Flores Gómez attests, the most recent opportunity would have been when the Image was removed from 1926 to 1929. Flores Gómez unwittingly implies an even more marvelous durability to the painting, suggesting there was a time when it was not mounted on its stretcher frame. It is possible, however, that the horizontal cracks he mentions were actually caused by the horizontal beams of the original stretcher frame, marks still visible today.

Flores Gómez noted that other restorers prior to him had retouched the Image, and he is certain that there have been other interventions since. When asked by Vera how many restorers would have worked on the Image, he answered: "It is very difficult to know that. Most difficult. But I calculate around 20." Naturally, it would be impossible for him to determine whether or not multiple retouches were done by the same artist, so all we can really gather from this testimony is that the Image has been retouched in at least twenty places, on several occasions before and after 1947.

The task for Flores Gómez in 1947 was to remove paint in some parts and restore the aforementioned breaks in the Image. He painted only in the damaged parts where he removed cracked paint, which is why he clarifies to Vera that this was a restoration, not a repainting.

Flores Gómez says that he worked on a part of the tunic (túnica) but not on the stars stamped on it (no a las estrellas estampadas en ella), because those had already been repainted. He has evidently confused the tunic with the mantle, as only the latter has stars. The vertical seam goes through both the mantle and the tunic, so Flores Gómez may have worked on both garments.

Elaborating on his comment about the stars, Flores Gómez showed various photographs to the reporter Vera, revealing how some of the stars had different numbers of points over time. One star, for example, had five points at one time, and six points at another. This proves that restorers had added or removed points at various times. We know from eighteenth-century testimony that all the stars originally had eight points. The fact that a restorer would make stars with five or six points implies that the restoration occurred only after at least two points had long disappeared without a trace. This implies that all such retouches took place in the nineteenth and twentieth centuries.

Interestingly, Flores Gómez says he had to use water-based paints, not oil, for his retouches, in order to match those used in the original. This suggests that he was not involved in retouching the face or hands, which have the appearance of oil. The baroque painters agreed that the tunic and mantle were made with a tempera technique, but were not more specific. Flores Gómez now tells us that the paint there is water-based, in which case it is properly a distemper. If this is true, the Image's centuries-long endurance becomes all the more remarkable, as distempers do not age as well as oil-based paintings, and such paints are easily removed.

In 1973, Flores Gómez again was asked to do restoration work on the Image. The abbot at that time, Guillermo Schulenburg, contracted him to clean the cloth and reline (reentelar) it, so that no more "alterations" (i.e., retouches) would be required. In the twenty-six years that elapsed, the painting had apparently deteriorated more. Flores Gómez notes that the icon was then still exposed to the ambient air in the old basilica, so it must have acquired a patina. As evidence of deterioration, he points out the peeling paint on the golden rays. He used a brush rather than a vacuum to clean the canvas, in order not to remove peeled paint.

The relative fragility of the golden rays is already well established, as we see signs of their deterioration even in late nineteenth-century photographs. If the entire Image were comparably fragile, as Flores Gómez seems to imply, this must have been a recent development, or the icon could hardly be expected to have survived one century, much less four. Yet Flores Gómez gives no specific evidence of extreme deterioration throughout the painting. Having been raised in a Guadalupan household, he was apparently scandalized to find that there was any decay at all.

Flores Gómez relined the painting with a new canvas backing. In the process, he measured the tilma's weave with a thread counter, and found that it was too fine to be made of ixtle (woven maguey). Rather, it was comparable to the fineness of cotton. Here we find nothing new, as it was already established in the eighteenth-century that the tilma had the fineness of medium-grade linen. This does little, however, to specify the material of the cloth.

Flores Gómez also speculates about the composition of the Image's paints. He posits that they are vegetable and mineral pigments dissolved in water, and applied in tempera. The golden rays, he judges, are likely made from ochre.

Lastly, Flores Gómez laments that other restorers have covered parts of the Image with new paint, creating an unfortunate contrast in colors, since the additions lack the patina of the original. He never met any of the other restorers, since they are likely loath to admit to such work. In his own case, he was never told by church authorities to remain silent, but he still felt he had to do so for his own security, since everyone in Mexico believes in the miracle.

It is at least questionable whether most restorers would share Flores Gómez's fears about coming forward. For one thing, anti-apparitionist opinions have been publicly proclaimed in Mexico for over two hundred years, without anyone suffering violence as a result. For another, educated guadalupanos have already known that the Image has been occasionally retouched. No one argues that the tilma, or indeed any religious artifact, is absolutely and perpetually incorruptible. Further, the eventual destruction of the tilma would have no bearing on the question of its miraculous origin. Given the weak basis for such fear, it hardly seems likely that there have been as many as twenty restorers who remained silent until death. In any event, we should be more concerned with the work of the restorers than their number, and try to obtain a more precise assessment of the extent of the retouching.

Chemical Analyses

Although many experts have given educated opinions about the composition of the cloth and its paints, none of these carry much evidentiary value without histological and chemical analysis. At most two such studies have been conducted, in 1936 and 1946.

Our only definite source for the 1936 study is the testimony of the well-known Mexican forensic scientist Ernesto Sodi Pallares (1919-1977), who devoted much of his time to Guadalupan investigations. One version of his account is mentioned in a February 1976 report that he and fellow criminologist Roberto Palacios Bermúdez gave to the Guadalupan researcher Manuel de la Mora Ojeda. [Carlos Salinas and Manuel de la Mora, Descubrimento de un Busto Humano en los Ojos de la Virgen de Guadalupe (Mexico: Editorial Tradicion, 1980), pp. 37-69. First published online in 2004 at http://www.interlupe.com.mx/Dictam.htm, now archived.]

In 1936, the basilica of Guadalupe's abbot, don Feliciano Cortés Mora, gave to Francisco de Jesús María Echavarria, bishop of Saltillo, some threads from the tilma for his reliquary. Bishop Echavarria gave two fibers, one red and one yellow, to Ernesto Sodi Pallares early that year to have them analyzed outside the country. At that time, Sodi Pallares was only in the second year (of three) of preparatoria (age-wise, like high school, but content-wise, like early college), yet he was valued for his good knowledge of German and for his close friendship with Marcelino García Junco, professor emeritus of organic chemistry at UNAM. García Junco, for his part, had obtained his doctorate in Germany and was a friend of the biochemist Richard Kuhn, the eventual Nobel laureate of 1938. He wrote a letter of introduction and recommendation to Dr. Kuhn on behalf of Fritz Hahn, a German academic residing in Mexico, who had been invited by the Nazi party to attend the Olympic Games in Munich. Hahn would bring the fibers to Kuhn for analysis.

It is not known if Kuhn was personally involved in the analysis of the fibers. As director of the chemistry division at the Kaiser Wilhelm Institute for Medical Research in Heidelberg, he may have delegated the task to other scientists or technicians. Decades later, Dr. Kuhn's son attested that he could find no record of such a study being conducted. [Jody Brant Smith, The Image of Guadalupe (Mercer Univ., 1994), pp. 77-78.] Nevertheless, Sodi Pallares says that results were sent back to Mexico, though he does not specify the means by which they were conveyed.

According to the lab results, the chemical composition of the fiber coloring did not match any natural mineral, vegetable, or animal pigment. The results were consistent with a variety of synthetic pigments, yet these were all invented in the late nineteenth century or later.

Naturally, these findings have been interpreted by guadalupanos as proving that the Image is of supernatural origin, as it is composed of materials that otherwise could not possibly have existed in the sixteenth century. Before the twentieth century, however, no one ever claimed that the paints were made of some otherworldly substance. They were assumed to be oil and tempera paints, and even the traditional legend suggests that the coloration came from flowers. The miracle of the Image was not in its substance, but in the means by which it became imprinted on the cloth. All previous analyses suggested that no human painter would have been competent to paint the Image, on account of technique rather than materials.

At any rate, the findings are questionable at best. In the first place, Sodi Pallares' testimony is too indirect to establish that Dr. Kuhn or anyone under his direction actually conducted the test. Second, even if the test was conducted, it could be that the fibers really did contain modern synthetic pigments. The fibers may have been taken from a retouched part of the Image, or they might not be from the tilma at all. The colors of the fibers, red and yellow, are not properly found on the Image. The tunic is pink or salmon, while the various adornments are a dark gold, not yellow. If salmon and gold are the intended colors, there is a decent possibility that both fibers were taken from retouched portions. The gold rays have suffered the most visible damage, while the gold stars have been repeatedly retouched. As for the tunic, the most likely place from which to extract a loose thread would be from along the seam, which is also the area most in need of retouching. Recall that the threads were originally given as relics, and need not have been selected for their suitability for analysis.

Better documentation is available for a 1946 chemical and microscopic analysis, done at the National University of Mexico (UNAM). A signed letter from the biologist Isaac Ochoterena (1885-1950), then honorary director of the Institute of Biology, is still available in the library of the Basilica of Guadalupe. The purpose of this study was to determine the composition of the tilma itself, rather than its pigments.

Previously, the consensus among experts had been that the tilma was likely composed of iczotl, botanically known as Palma silvestre (now Yucca aloifolia), rather than maguey proper (i.e., the genus Agave). The reason for this is that the cloth was considered to be much too fine to be woven maguey. The 1946 UNAM study, however, would revive the older magueyista tradition.

At the behest of the illustrious art scholar Manuel Toussaint, the biologist Isaac Ochoterena, a highly accomplished specialist in plant histology, analyzed fibers from a piece of the tilma measuring 8 mm by 5 mm. The cloth fragment was from a reliquary in the care of Lic. D. Benigno Ugarte, accompanied by professor D. Abelardo Carrillo y Gariel, D. Alfonso Marcue González (official photographer), and Lic. D. Manuel Garibi Tortolero (a Guadalupan scholar). Ugarte opened the reliquary in the presence of the others, and Ochoterena took about 3 milligrams of fibers from the relic. A photograph was taken of the cloth.

Relics of the tilma had been in common circulation since the last half of the seventeenth century, though it is not clear which of these were authentic pieces of the original. [See fray Isidro de la Asunción, Itinerario a indias (1673-1678).] We know of at least one authentic relic issued before 1946. In 1941, Mexico's archbishop Luís Martínez gave a half-inch square from the tilma to the Archdiocese of Los Angeles, in gratitude for Archbishop John Cantwell's visit to Mexico City, which alleviated church-state tensions. In a recent article, Los Angeles archbishop José Gómez says this is the only relic of the tilma outside of Mexico City, since the original resides in the basilica there. He seems to have no knowledge of other authentic relics. [Mons. José H. Gómez, "Un tiempo de María," ACI Prensa, 7 Dec 2012] Ochoterena's letter does not indicate the provenance of the relic he examined.

Ochoterena reports that the cloth appeared coarse like a sack, and the removed fibers had a dressing (aderezo), part of which must have been flaxseed (linaza). We note that flaxseed or linseed oil is an ordinary component of oil paint, but is not used as sizing, since it causes fabric to rot. Since Ochoterena perceives the cloth to be quite old, this evidence of paint without sizing is remarkable.

The few histochemical tests he was able to perform yielded the following results. Reaction with zinc chloro-iodide (Herzberg stain) resulted in dark yellow. This excludes most non-lignin fibers, which would turn red. Fibers with lignin, such as hemp, should turn bright yellow. The darker color is consistent with some (but not all) agave species (e.g., Agave americana, which have a low lignin content. Reaction with iodine and sulfuric acid yielded bright yellow. This excludes cotton, linen (except New Zealand flax) and hemp (except Manila hemp) which should turn blue. It is consistent with yucca, palm, agave, and aloe. [Color Trade Journal and Textile Chemist, 1923, 12(5), p.224.]

Lastly, the fiber swelled markedly when reacting with copper carbonate in ammonia. This reagent, much like Schweitzer's solution (which uses copper hydroxide instead), dissolves cellulose. [Horace G. Deming, "Some new solvents for cellulose and their action on this substance," J. American Chemical Society, 1911, 33(9), pp.1515-1525.] This eliminates most synthetic fibers. The fibers did not dissolve in any of the reagents, which eliminates cotton, silk and linen.

The chemical analysis was followed by a microscopic examination of the plant fiber cells. These were measured to be 1.5 mm in length, and 0.02 to 0.03 mm wide (i.e., 20 to 30 microns). This eliminates cotton, which has unicellular fibers, as well as flax and hemp, which have much longer cells. (Flax fiber cells are 4 to 77 mm long, with a mean of 33 mm. Hemp cells are 5 to 55 mm long, with a mean of 25 mm.) These dimensions fall within the range for agaves. Agave sisalana, for example, has cells 0.8 to 8 mm long (mean of 3.3 mm), and 7 to 47 microns wide (mean of 21 μm). [Menachem Lewin, ed. Handbook of Fiber Chemistry, 3rd ed., CRC Press, 2007, p. 476.] Agave americana, however, has a much lower diameter, with an average of 3.1 microns. [Yassine Chaabouni et al. "Morphological Characterization of Individual Fiber of Agave americana L." Textile Research Journal, 2006, 76(5) 367-374.] Judging from the size, rectilinear configuration, and structure of the cells, Ochoterena concluded that the plant was an agave of indeterminate species.

To this day, this has been the only technical analysis competent for determining the composition of the cloth. All other expert opinions, given before or since, have been based on ocular or low-power microscopic examination, without any chemical analysis or precise measurement. In a word, they are educated guesses, based on the weave and texture of the cloth. Only Ochoterena has made proper cellular and histochemical analysis, a task for which he was eminently qualified, and his verdict was that the cloth is made of a species of agave, commonly known as maguey.

The only feasible way to avoid this conclusion is to suppose that the relic was not an authentic piece of the tilma. In a 1949 article, the basilica's abbot Feliciano Cortés Mora says that it is beyond doubt that the tilma is of maguey. When the icon's glass was opened in August 1946 for photographs to be taken, a modernist painter claimed that the cloth was European. Subsequently, a thread from the tilma was analyzed as described and found to be agave. [La Voz Guadalupana, Feb. 1949, xv, 12, pp. 6-8.] Although Mora is certain that the thread was from the tilma, a more explicit documentation would be desirable.

In their 1976 report, Drs. Ernesto Sodi Pallares and Roberto Palacios Bermúdez proposed that the species of agave used in the tilma was A. popotule, a variation of A. lechuguilla. They note that past investigators may have confused this plant with the yucca palm called izote (iczotl). Its strong, flexible fibers were used to make hats and blankets for the poor. The finished woven product is rough, hard and resistant. Notably, the side that receives more light becomes smooth and soft over time. This is consistent with what has been observed of the tilma's texture on its painted side.

Photographic and Ocular Studies in the Mid-Twentieth Century

Jesús Cataño Wilhelmy (1946)
Alfonso Marcué González (1946)
Don Joaquín Flores (1946)
Coley Taylor (1947)
Manuel Garibi Tortolero (1947)
Feliciano Cortés Mora (1949)
Helen Behrens (1951)
Nicolás Mariscal (1952)
Francisco Camps Ribera (1954, 1963)

From 1946 to 1976, the Image was subjected to a variety of expert examinations, aided at times by enlarged photographs. We present a summary of their results, excluding the issue of purported figures in the Virgin's eyes, which we reserve for the next section.

Jesús Cataño Wilhelmy (1946)

In March 1946, the commissioned photographer Jesús Cataño Wilhelmy took 34 panchromatic plates of the Image over two nights, some through the glass, and others with the glass open. The purpose of his study was to reproduce the exact colors of the Image, so he took many of these photos through color filters.

Using this technique, he could clearly distinguish the original image from the retouches, since (as Flores Gómez explained) the restorers could not match the colors exactly. His photographs showed that most of the face was still original, except for retouches on the forehead, nose, cheekbone, and tip of the chin (barbilla). These were perhaps justified by the darkening caused by candle smoke in the upper part of the Image. Similarly, he saw that the cloak, though mostly original, had been retouched in the area of the head. Through the blue filter, he could discern a horizontal line, curved downward, of a change in tone, probably corresponding to where the crown had once been. He also saw a reddish spike extending from the apparent crown, matching a color found in some of the surrounding rays, which were now mostly overlaid with gold.

Remarkably, Cataño found that the face of the Virgin had a yellow refulgence which appeared in photographs, but was not visible to the naked eye. The yellow filter captured the entire pattern of the face, while the blue filter, which ordinarily should pick up shading, captured nothing of the face's relief. This implies that the painter used bright rather than dark colors to show relief or texture, which is contrary to established technique for shading. The photographer considers this to be miraculous, since the human eye (i.e., that of a painter) would not be able to distinguish shades of yellow sufficiently well to make a textured image thereby.

Another marvel, reported by Cataño in a later interview, was that the color of the Virgin's face changed with the distance of the observer, even under constant illumination. He and his coworker, the lithographer Alfonso Mártinez de Velasco, independently observed this phenomenon. The latter noted that the face seemed much more reddish when one approached it up close, than from the floor of the sanctuary. The painter Eduardo Cataño, brother of the photographer, confirmed this observation. Jesús Cataño further noted that an altar rail height's change in distance (i.e., two feet), was sufficient to produce this effect. Remarkably, white light caused the colors to turn gray rather than brighten. [La Voz Guadalupana, xv, 5, July 1949]

Cataño Wilhelmy did not live long enough to complete a technical study investigating the changes in color with distance, which he could not reproduce in other paintings. We note that the reddish portion of the face is said by other witnesses to have been a later retouch, so we should hardly expect this paint to have any paranormal properties. The phenomenon of color change is more likely to be related to the material or texture of the fabric, whose coarseness is much more pronounced in the area of the lightly painted face.

Alfonso Marcué González (1946)

Alfonso Marcué González, who had been official photographer of the basilica since the 1920s, produced a report on the Image's physical attributes in 1946, titled "Cómo es Nuestra Señora de Guadalupe de México," which is still preserved in the basilica archives. He says that the tilma was 105 centimeters wide and 168 centimeters tall, not counting the parts folded at the edge of the stretcher frame. We note that the width exactly matches the precise measurement by Bartolache in the eighteenth century, but the tilma has lost two centimeters in height. This might be attributable to the folding at the edges, if the tilma was partly remounted at some point.

The stretcher frame, González reports, was made of cedar. It consisted of four beams plus two additional horizontal beams to prevent the frame from bending. These last two leave visible prints in the Image that show up in photographs.

The Virgin's height, from hair-line to sandal-tip, is 143 centimeters (4 feet, 8 inches). This and other measurements were taken on March 21, 1946.

The two pieces of the tilma were woven together by a thin thread of the same material. It has resisted not only the weight of the cloths, but also countless times it has been touched by icons, medals and rosaries of the devout.

The entire work is painted in an unknown technique, but it has some resemblance to tempera paintings. It has no sizing or primer other than that of the colors themselves, incorporated in the rough cloth. The cloth is too loose and coarse to be painted upon.

Evidently, over the centuries, the Image has been subjected to "light retouches, made surely with a sense of mistaken piety". Fortunately, these have not fundamentally affected the aspect of the original. Nonetheless, Miguel Cabrera may have confused these later retouches with the original when he identified additional styles or techniques.

The Virgin's hair seems blotted from a distance, but under a magnifying glass one can discern its silky locks.

In a copy of the image using X-ray plates, the "outline (trazo) of the drawing appears intensely under very black lines, as if it had been projected on the cloak." When the lines are amplified photographically, "the outlines appear diffuse, distinct from other lines traced over the original."

Under powerful photography, the original colors can be distinguished from opaque retouches. In the area of the crown, one can still see the delineations of some rays and a gold band that united them, though this is not readily visible to the naked eye.

Don Joaquín Flores (1946)

In 1946, Don Joaquín Flores made a meticulous copy of the Image in its current state, using high quality photographs. He reported his detailed observations in December. [Juan Diego, 1946, viii, 85, pp. 29-31.]

On the right side, Flores saw the stain from where nitric acid had been spilled in the late eighteenth century. Some reddish coloration in the lower left may be a remnant of some painted roses that are said to have been added to the lower portion as an embellishment, much like the cherubim that were later erased. These areas (i.e., of the now erased roses and cherubim) have been painted over. There are also various small round stains, that seem to be of candlewax.

The field of rays is made of a dirty yellow ochre, and is covered with countless stains, from the patina of age or accidental maltreatment. Around the head the stains are larger and seem to form figures. The reddish background of the rays has deteriorated in the lower part, especially on the left.

Flores counted about 122 rays, for those at the feet and head were completely erased. Some are straight and others wavy. They are not equidistant, as those on the right are closer together than those on the left. Their thickness is not uniform either. They are said to have been adorned with gold a long time ago, and this may account for what has fallen away, especially on the right side and at the feet.

The face is slightly browner than pearl, just as Cabrera observed. It has very little chiaroscuro (i.e., contrast of light and dark), so that even subtle light and dark stains alter the face's aspect, especially on the forehead and cheek. Still, this is one of the better preserved parts of the Image.

The hands have many tiny horizontal stains. There is a dark stain on the left index finger and a light stain on the left ring finger.

The belt, which seems to have been purple (morado) previously, is now brown-black (negro pardo).

Flores notes that you can see the impression of the two crossbeams of the stretcher frame. He mistakenly thinks that this frame is no longer in use, having been replaced by the silver plate backing.

The tunic has still retained its salmon color, but the darker part has many small round stains, especially below the knee. The bottom of the tunic has the color most maltreated and muddied.

Astonishingly, Flores confirms Cabrera's observation that the tunic's floral decoration has a hair's breadth outline on either side. He says the decoration consists of a gold line flanked by two black lines, each as thin as a hair. He says this was examined by "some experts" who at first denied these existed, yet they confirmed their presence and declared that the Image must have been supernaturally made, as no human could have made such delicate lines on a surface like the ayate without sizing. Flores does not identify these experts.

Partial figure 8 In the lower left, Flores notes there is a figure of an 8. Unlike Cabrera, who saw an 8 as part of the gold decoration, Flores finds a black, partial figure 8. P. Mariano Cuevas suggests that this may have been the mark of the majordomo who handled the bundle of cloths from which Juan Diego's tilma was made. [Juan Diego, Jan. 1947, p.11.] Such a mark, however, ought to be beneath the painted Image.

According to Flores, it is said that the moon was once adorned with silver, for which reason it has browned and blackened. It has many places where the paint has fallen away.

The angel has the nose and eyebrows of an adult, but is childlike in other respects. Its face has many small damages, light and dark. Its dress is well preserved and is almost the same color as that of the Blessed Virgin. Its wings are blue, white, and reddish, not the Mexican tricolor. In one of the arguments over the Image's supernatural origin, a wing was scratched in order to see if there was sizing beneath it. Only a little hair (pelucilla) came out, leading them to declare that the Image was miraculous. Again, Flores does not indicate who "they" are.

Despite the numerous stains and damages that can be seen at close range, the Image of the Virgin retains a beautiful aspect from a distance. The deterioration of the Image, especially in its lower part, is attributable to the practice of allowing the public to touch it, kiss it, and light candles in front of it. Considering this treatment, combined with the fact that it had no glass for its first 140 years (counting from 1531), and that the area has such humid, saline air, Flores regards the Image's state of preservation to be miraculous.

Coley Taylor (1947)

The New York journalist Coley Taylor agreed with Flores' observation that the Image appears more beautiful from a distance. In 1947, Taylor noted that many details of the Image were more easily seen from a distance than up close, contrary to logical expectation. The stars are radiant from a distance, though barely discernible up close. Even the colors change, as the sea-green cloak becomes dark blue when viewed up close. The tunic is a faint pink up close, but becomes more intense in color from a distance. The face appears light up close, but becomes shadowed from a distance. The figure also seems larger when viewed from a distance, contrary to logic. None of the many masterpieces viewed by Taylor in New York museums and private collections has these aspects.

Manuel Garibi Tortolero (1947)

In a 1947 article, the Guadalupan researcher Manuel Garibi Tortolero comments on the photographic plates taken by Jesús Cataño Wilhelmy the previous year. He claims that the plates prove that there is no sizing or primer under the paint. Further, he asserts that all the baroque painters were mistaken in their belief that multiple techniques were used. Oil is used only in the retouches. The original Image is only tempera, but of a kind that was done in a single step. It was not properly painted, but imprinted or stamped on the tilma by the flowers, for which reason it has natural pigments. He agrees with the canon Don Luís Cabrera (who presented at the Interamerican Marian Congress in 1945) that the Image has no brush strokes, and its colors are incorporated in individual threads, sometimes forming figures thereby, as in a tapestry, while in other places the pigment dust seems to have been pressed into the cloth. [Juan Diego, Dec. 1947, pp.44-47.]

Feliciano Cortés Mora (1949)

In 1949, abbot Feliciano Cortés Mora gave a description of the Image in the basilica's official newsletter. [La Voz Guadalupana, Feb. 1949, xv, 12, pp. 6-8.] He himself had closely viewed and touched the cloth on various occasions, confirming its smoothness. He speculates that Juan Diego himself wove the cloth, providentially guided to use special care and the finest agave. He confirms that the area exposed by the frame is 168 by 105 centimeters. The stretcher frame must be of hard and incorruptible wood, as it is of unknown age, but it is not viewable because of the silver plate backing. The tilma must have been cut to fit the frame, as indicated by stray threads sticking through the silver lining. He also notes a tradition among certain Guadalupan writers that strips and fragments of the tilma were kept in a reliquary for many years at the chapel, though he does not know if this is well founded.

Notably, the abbot acknowledges that the Image has suffered some retouches, and these retouches have deteriorated over time. Examples of this are at the union of the two cloths and in the gold rays which someone had foolishly tried to brighten. In the neck and near the close of the tunic there is a little hole, a vestige left by some pious hand. Again, we are reminded of the astonishing amount of exposure and handling to which the tilma had been exposed in early centuries, making its preservation all the more marvelous. We further note that there has not been any attempt by the local church to suppress knowledge of the retouches, as the abbot published this article in the basilica's official newsletter in 1949.

Helen Behrens (1951)

In May 1951, the historian Helen Behrens, accompanied by a young painter and sixty other witnesses, observed the Image with the glass opened. Behrens agreed with earlier testimony that the colors seemed more vivid when viewed from further away. She also confirmed that the gold veins in the tunic had a dark border which made the gold jump out more. She notes that the face has rosy cheeks, and hair that is almost black. The shading of the tunic, which seems black from afar, is actually carmine. The type of painting cannot be classified, as it simultaneously exhibits traits of oil, tempera, watercolor and pastel. The beauty of the face is not replicated by any copy, and only the best photographs come close to capturing it.

Aside from her confirmation of other witnesses, Behrens' testimony is useful for dating some minor alterations to the Image. Retouches adding rose to the cheeks and darkening the hair have been made by now. More remarkably, she gives the cloth's measurement as 168 cm by 103 cm, a loss of 2 cm width since the 1940s. This suggests that the cloth may have been partly remounted.

Nicolás Mariscal (1952)

In a 1952 study of the Image, the architect Nicolas Mariscal noted the painting's seamless fusion of light and shade, finding a masterful moderation and subtlety in their use. This yielded the effect of volume, not of an opaque illuminated body, but of a refulgent one. He also notes the presence of chiaroscuro, creating multiple sources of light. The only black in the painting is in the moon, the bow ribbon, and the hair of the Virgin and the angel. (All these areas, perhaps not coincidentally, were likely recolored.)

Mariscal says the painting is without sizing, except in the areas of retouches, such as where the cherubim were once painted. He also says it has no varnish, making the aforementioned effects of chiaroscuro all the more remarkable. On these matters, however, he seems to rely on the testimony of the baroque painters, not his own observations.

Mariscal considers the painting to be far too masterful to have been painted by an Indian. A Spaniard, on the other hand, would not have chosen a crude ayate as a canvas, nor the Mexican type without a model, using unknown colors. Apart from the story of its divine origin, the author of the painting must remain a mystery.

Francisco Camps Ribera (1954, 1963)

In 1954, Professor Francisco Camps Ribera (1895-1991), an internationally known master painter from Barcelona, conducted a study of the Image at the behest of the historian Helen Behrens. He observed the icon on Holy Wednesday (April 14), with the glass case closed. He found the cloth had not been treated with any plaster, and judged that no human could have painted on the cloth without preparing it. Examining the Image with a strong magnifying glass, he could not find any trace of a brushstroke.

Having examined thousands of paintings in museums and private collections throughout Europe and North America, Camps Ribera could easily determine the technique of a painting—oil, tempera, watercolor, etc.—by visual inspection. Despite this experience, he could not say how the Image was made.

Drawing upon his knowledge, Camps Ribera deduces that no Spanish, Flemish or Italian painter of the sixteenth century could have produced the Image. No foreign painter residing in Mexico demonstrated sufficient sensibility or technique. He finds it incredible that any of three Indian painters who worked for the Franciscans—Marcos Cipac, Pedro Chachalaca or Francisco Xinamamal—could have represented the Virgin in such an authentic Christian spirit, as they were all recent converts from a very different religion.

The Image is over 400 years old. Any other painting that age is cracked, has lifted paint, and is darkened with a tobacco color. The Virgin of Guadalupe still has brilliant colors, "no major signs of age," and still gives a "sensation of a fresh and eternal youth."

The colors are vivid with sharp distinction between light and shadow from afar, yet up close with a magnifying glass they seem weak and blurry. Camps Ribera knows of no other painting in the world with this attribute.

Camps Ribera says the cloth is 66 by 41 inches. This converts to 168 by 104 cm. We cannot say if he agrees with Behrens' measurement of 103 cm or the older measurement of 105 cm, since either could be rounded to 41 inches. The Virgin is 56 inches tall, but seems larger from a distance.

The gold adornment is most difficult to conceive technically, as it is not attached with any fixative. It looks like gold dust, and is unlike any commercial gilding. There appears to be a lot of gold from afar.

Shortly after this examination, Camps Ribera acquired a cloth of the same material as the tilma. He prepared various paint colors and techniques used in the sixteenth century—oil, watercolor, pastel, tempera, ink, gilding and gold leaf—and found that the colors were just as vivid when viewed through the other side. He was informed that the Image, by contrast, had no visible colors on the other side, but we have found such reports to be ambiguous. Further, we cannot be sure that the test sample matched the material of the tilma, as the identity of the latter is disputed.

In a second examination in March 1963, Ribera noted that the black outlines appeared to have been added at a later date. He also claimed that the crown was not original to the Image. By looking at the few lines remaining of the crown, it could be seen that its gold, what little remained, was different from that in the rest of the Image. He further noted that, although it was atypical for the moon to be colored dark brown instead of gold or silver, the moon on the tilma was always this dark color, as seen by comparison with the bow beneath the Virgin's hands, which has the same color.

The Figure in the Eyes

From the 1950s through the 1970s, various experts in photography and ophthalmology have attested that the image of one or more human figures can be found in the Virgin's eyes. Ordinarily, we might dismiss such claims as mere coincidence. After all, the human brain is highly proficient at facial recognition, accounting for our propensity to recognize faces in clouds and other amorphous objects.

Purkinje-Sanson images Right Eye of Virgin of Guadalupe

Yet there are two considerations that might make us take the claims more seriously. First, the same figure seems to be in both eyes. Second, in the right eye, which is more clearly rendered, there are two or possibly three instances of the human figure, in the locations, sizes and orientations one finds in the reflections on a human eye, known as Purkinje-Sanson images. Both of these considerations greatly reduce the likelihood that we are dealing with mere coincidence. [See expert reports in Carlos Salinas and Manuel de la Mora, Descubrimento de un Busto Humano en los Ojos de la Virgen de Guadalupe (Mexico: Editorial Tradicion, 1980).]

Left Eye of Virgin of Guadalupe The strength of these assertions, however, is limited by the resolution of the images. While the primary image of a bearded man's face is clearly recognizable in the right eye, it is less clear in the left eye, which is generally less detailed and may have been retouched. Also, the second and third Purkinje-Sanson images are not obvious human forms. The latter, in fact, is little more than a bright dot. While this evidence is consistent with the sizes of Purkinje-Sanson reflections, we are dealing with a scale that approaches the resolution of individual threads.

Little credence is owed to claims that there are more personages, as many as thirteen figures, in the eyes. These more extravagant claims are not substantiated by expert analyses, and belong in the realm of cloud-reading.

Callahan Infrared Study

Arguably the most important study for reconstructing the Image's history is the infrared analysis conducted by Philip Serna Callahan in 1979, at the invitation of Professor Jody Brant Smith of Pensacola, Florida and Msgr. Enrique Salazar of the basilica. Dr. Callahan was a biophysicist at the University of Florida who was an expert in infrared photography and an experienced painter, familiar with natural pigments and with Renaissance painting techniques. His is the only scientific study of the Image that was duly published shortly after completion. [Philip S. Callahan, The Tilma Under Infra-Red Radiation (Washington, DC: CARA, 1981)]

Dr. Callahan was the sole author of the monograph containing his results, and he alone conducted the infrared photography and analysis. Although a publisher's foreword identifies Prof. Smith as a collaborator, Callahan himself only dedicates the monograph to Smith and Salazar, without acknowledging any contribution to the study. In fact, he says, "just as the scientific work contained in this report is mine alone, so also is the interpretation of that work." (Callahan, p. iv.) It is a mistake, then, to refer to this as the Callahan-Smith study, or to imply that two scientists rather than one were involved in the research.

This study is often described as a "NASA study" or conducted by "NASA scientists," on the basis that Callahan and Smith both had previously done work for NASA. The infrared study of the tilma was not sponsored by NASA, and the "Guadalupe Research Project" is not a NASA project, but a private company founded by Prof. Smith for the purpose of funding investigations. Callahan and Smith are not "NASA scientists" by virtue of having worked under NASA grants or contracts. This is not a slight, for in fact most NASA-sponsored scientific research is done by academics funded by grants or contracts. NASA staff with the title of 'Scientist' are usually program administrators or clinicians, with little expertise or research experience compared with academic scientists.

Callahan was permitted to photograph (but not touch) the Image, removed from its glass and metal reliquary, from 9 p.m. to midnight on May 7, 1979. The small room was lit by two 500-watt floodlights. Kodak High-Speed Infrared Film was used, with exposure times of 1/30 and 1/60 sec. The camera was hand-held due to cramped conditions, yet all forty photographs came out in perfect focus for the 0.7 to 0.9 μm infrared range.

The gold sunbeams were "in very bad shape," and there was also some cracking and fading of the gold in the mantle and the stars. A crack in the gold trim of the mantle reveals a black line that was apparently a guide for laying the gold. In some places, the embellisher failed to cover this line completely. Unlike the shadow lines in the mantle, which are integrated into the blue paint, this guideline is transparent to IR. The gold rays of the sunburst were opaque to IR, while the gold trim and stars were partly transparent. Callahan cautions that we cannot know the composition of the pigments without chemical tests, but guesses that the stars and trim were made of native clay or sandstone ochre, mixed with alumina hydrate to give a bright yellow. The sun rays appear to be metallic gold. Callahan considers all the gold components to have been added long after the original painting.

The moon and tassel, in Callahan's judgment, are elements of Spanish Gothic style that were later added to the Image. They turned dark brown with age, as did the angel's hair. In all these near-black areas, the paint is breaking away and "in an extremely bad state of repair." (p.7) The pigment is not carbon black, for that would not turn brown. It could be silver nitrate, which darkens with age. Much more likely is a ferric oxide known as Mars black. As a heavy pigment, it cracks with time if not properly bound to the canvas.

There are four horizontal fold lines (i.e., corresponding to the edges of two stretcher frame crossbeams), the upper two of which cross the tassel and body, but end at the edge of the mantle. Their non-appearance in the surrounding background implies that the background was added later, according to Callahan. (Another possibility, perhaps, is that plaster or another medium in this area prevented creasing.) By the same token, the tassel (and the contemporary moon and angel) must have existed before the background and its overlaid sunburst.

Callahan thinks "the tassel, and probably the moon, and even the angel since the angel's hair is also cracking away, were added by human hands...". He assumes that deterioration is evidence of late addition, which makes sense only if we presume that the original Image must be incorruptible. Yet even if we accept a miraculous origin for the Image, it does not follow that all its parts should be permanently imperishable. Callahan also says that the moon and tassel lack artistic technique and beauty found in the rest of the painting, though this judgment may be colored by his assumption that any corruptible portion must be non-miraculous and therefore non-original.

The black border of the mantle's gold trim is cracking, and it crosses the top fold lines, suggesting it was added after the tassel and moon, around the same time as the gold trim and sunburst background, which also obscure the folds. In some places, the black has gone over the gold rays. Stars consistently overlie the black border, so the order of addition was as follows: 1) sunburst, 2) black/gold border, 3) stars. The last two are judged to be very late additions; i.e., late 16th to early 17th century.

The mantle is turquoise blue, more blue than green. Callahan thinks it likely a Bremen lime blue, which is a copper hydroxide carbonate mixture. It resembles the blue in early Mayan wall paintings (copper oxide) or cured animal skin codices of the Aztecs. Although all such pigments are semipermanent and fade badly with age, the blue in the cloak is "bright enough to have been laid last week." (p.10.) There is some blue missing at the center seam of the tilma, likely due to folding or tearing when moving the relic. Otherwise, the blue is of even density and unfaded, which is physically inexplicable.

The pink robe is remarkable for its luminosity, being highly reflective to visible light, yet by far the most IR-transparent pigment. Unlike the blue, which fills spaces between the weave, the pink is barely on the weave's surface. As with the blue, the shadowing in the pink is blended with the paint layer, and there are no signs of underdrawing. The fold shadows, at first glance, seem to be thin sketch lines, but on close examination they are broad and blended with the paint layer, which is "uncharacteristic of undersketching." (p. 10.)

The pink pigment is inexplicable. Cinnabar, hematite, red lead and red oxide would be opaque to infrared, and there is no evidence of modern aniline colors. Organic pigments would be transparent, yet they are impermanent unless protected by varnish, which is not the case here.

One of the really strange aspects of this painting is that not only is the tilma not sized, but there is absolutely no protective coating of varnish. Despite this unusual total lack of any protective overcoating, the robe and mantle are as bright and colored as if the paint were newly laid. (p. 11.)

The gold embroidery is the same pigment as the gold stars, but less faded or cracked since it is laid as thin lines. Callahan judges this to be a late addition since this "flat" technique is "not in keeping with the beautiful realism of the face nor the drapery of the cloth." (p. 11)

Virgin of Mercy by Bonanat Zaortiga Callahan considers that the embroidery, the black outlines, the black cross brooch, and fur-lined sleeves were all Spanish International Gothic embellishments. He finds similar features in Bonanat Zaortiga's 15th century Virgin of Mercy. He is not claiming that these features were copied from Zaortiga's work, which is in Barcelona, but only that they exemplify a common style. We can see that Zaortiga's Virgin is markedly different from that of Guadalupe. Instead of having downcast eyes, his Virgin stares squarely at the viewer. His gold trim is much more ornate, the sleeve lining much further from the hands, and the floral embroidery is dark instead of gold, not to mention more obvious dissimilarities. The Virgin's prayerful pose is common in iconography, but the cross brooch is more indicative of a specific style and period.

Callahan claims that the entire lower portion of the painting—the angel and lower fold of the robe—"is a 17th century Gothic addition". (p. 11.) He considers the angel's arms to be clumsy and out of proportion (assuming child or adult proportions?), and its drawing to be mediocre at best. Its face "has none of the beauty or genius of technique" found in that of the Virgin. Its hair overlaps the moon, and the draw-line of the moon shows through in the IR photos, indicating that the angel was painted over it. The red of the angel's robe is laid thick and, unlike the Virgin's robe, is opaque to IR. It is likely red oxide, a permanent pigment. The red is chipping on the robe and wings. The blue in the wings is badly cracked and laid thick.

The angel shows signs of brush strokes which are not evident in the face, hands, mantle or robe of the Virgin. From the way that the cross hatching of the fabric is filled in across the angel's arms, there is evidence that this portion of the cloth tilma was sized before the paint was applied. The same is true of the moon and the rest of the angel including the bottom fold of the robe. (pp. 11-12)

Callahan conjectures that the bottom fold of the robe was designed to resemble the Aztec glyph for "tilmatli", though the similarity is not especially convincing.

The lower left fold of the blue mantle, unlike all the rest of the mantle, shows brush stroke texture and is not the semitransparent blue of the body of the mantle. It was probably added at the same time as the angel to give a hand-hold to the angel. (p. 12)

To recapitulate Callahan's assessment of the sequence of additions, we have: 1) the original face, hands, mantle and tunic; 2) the moon and tassel; 3) the angel and lower fold; 4) the white background and sunburst; 5) the gold trim with black outline; 6) the stars. We know from the Echave copy that all these elements were present by 1606, so they must have been added before then. Yet additions 4 through 6 were supposedly made after the fold lines appeared, which means some considerable time should have elapsed. We may need to seek alternative explanations, such as plaster in the background and retouches in the gold and black, to account for Callahan's evidence.

According to Callahan, "The hands are the most altered part of the painting." (p. 13) This is unsurprising to us, in light of the extreme wear to which this area was exposed for devotional purposes. The fingertips have been shortened, especially on the left hand, where they once extended to the upper left another half inch. The upper left and lower right hands were outlined to emphasize their new shape, and these outlines. The gold bracelet and cuffs at the wrist were late Gothic additions. "The bracelet is of the same transparent gold as the stars and must have been at the same time as the gold stars and the gold border of the mantle." (p.13.) The fur at the neck and cuffs is of an opaque pigment, possibly lime or gypsum. This white overlays the rose of the robe at the cuff edge. "The original hands... show no undersketching whatsoever, and the shadowing between the original fingers is an integral part of the hand pigment." Their original shading, color and pigment are inexplicable for the same reasons as the face, to be described.

The head of the Virgin of Guadalupe is one of the great masterpieces of artistic facial expression. In subtleness of form, simplicity of execution, hue and coloring it has few equals among the masterpieces of the world. Furthermore, there are no portraits that I have ever observed which are executed in a similar manner. (Callahan, p. 14.)

Remarkably, infrared closeups show no underdrawing or sizing of any type. Visible light closeups clearly indicate the lack of sizing by the unfilled spaces in the fabric. The skin tone is definitely Indian, and from a few feet away it appears gray-green.

Callahan is puzzled by the white highlight of the cheek, which appears to be a caked on pigment, semitransparent in IR. Such a pigment should not last for centuries. We know from other evidence that this highlight was likely a twentieth-century retouch.

The gray-hued shadow areas, such as the right side of the face beyond the nose, the mouth area, and the dimple below the mouth area are thinly laid on and the coarse tilma fabric is quite evident in these regions. (p. 14)

The beautiful meditative expression of the face is formed by simple lines, which, up close, seem crude and flat, but from afar there is an "elegant depth of perspective." Remarkably, the painting takes advantage of the tilma's coarseness to achieve its effect. There is one place where a coarse fiber is raised above the rest of the weave, following perfectly along the ridge of the top of the lip. Roughness in the cloth creates similar effects below the highlighted area of the left cheek, and to the lower right of the right eye. "I would consider it impossible that any human painter could select a tilma with imperfections of weave positioned so as to accentuate the shadows and highlights in order to impart such realism." (p. 14)

Infrared photography shows that eyes and shadows around the nose have no underdrawing, but are part of the face pigment. "Close to the painting the highlights of the eyes are subdued to the extent that they appear nonexistent." (p. 15) The face and hands have a phenomenal tonal quality resulting from the diffraction of light off the coarse fabric. He judges it impossible to achieve this marvelous effect by human design. This diffraction accounts for why the color changes with distance. As the pigment and surface blend, at a distance of six or seven feet, the olive-skinned beauty emerges from the gray and white.

Callahan thinks the black pigment for the eyes and hair cannot be iron oxide since they have not cracked or faded. He is apparently unaware of testimonies that these areas had been retouched.

Callahan misinterprets Miguel Cabrera as having found the four techniques of "oil, tempera, watercolor and fresco" in the Image. We have noted that aguazo in baroque Spanish means "gouache." Labrada al temple is not properly fresco, which mixes paint in plaster. It is similar to secco (paint on plaster), using a two-stage process, except the first step uses white egg tempera instead of plaster.

Unlike Cabrera, Callahan believes that the white background used a cold lime plaster. The hardening of the plaster would account for the lack of fold lines in this area. Yet egg tempera might produce a similar effect. Contrary to what Callahan suggests, the absence of fold lines does not help us date the white background as later than the original Image. The plaster (or egg tempera) may have been applied before the fold lines emerged anywhere, and subsequently prevented them from appearing in the more rigid portion of the cloth. The misunderstanding seems to arise from Callahan thinking the cloth was actually folded.

Overall, the Image appears to be tempera, according to Callahan. Yet ancient tempera paintings on canvas will crackle if unvarnished or darken if varnished. As the rose robe is neither crackled nor faded, it is tempera only in appearance, but its technique is inexplicable.

Misunderstanding Cabrera, Callahan does not see how the mantle could be watercolor, as it is laid on too thick, and has not faded. These objections do not hold against gouache, which is the correct translation of Cabrera's aguazo. Callahan regards the blue as an unknown pigment applied with an unknown technique.

Callahan disagrees with Cabrera's claim that the hands and face are of oil. "Oil on unsized cloth and unprotected by varnish would have rotted the tilma within fifty years." (p. 17.) The face and hands have a tempera appearance, but the highlights are of a heavy lime-like substance while the shadows use the texture of the fabric. "There is no glossy appearance as would be true of an oil film." These observations may be the result of twentieth-century retouches. The Virgin's face was said to be much more luminous before these retouches, while the caked-on lime was a late addition. It is possible that the face originally had the appearance of oil paint, though a true oil paint would have destroyed the cloth, save for a miracle.

The conjectural history of the Image that Callahan offers (pp. 18-21) is unsatisfactory on several points. First, he insists that the white background must be a late addition since no one could wear a tilma laden with plaster, though it is by no means proven that the background is plaster, rather than labrada al temple. Second, he speculates that the tilma must have been folded at some point, but never presents evidence of any creases besides those of the stretcher frame. He seems to be ignorant of whether there is currently a wooden frame, mentioning only that, "The portrait is known to be mounted on a large metal sheet." (p. 16)

Least satisfactory of all is his speculation that the Image suffered water damage during the flood of 1629, and that this was covered by the angel and lower fold of the robe. His only evidence is that the Image was brought to the flooded capital by canoe, and the archbishop promised not to return her until he could take her back with "dry feet." This expression simply meant that he would not return the icon until the flood had abated, so she could return "on foot" rather than by boat. At the time, this was a bold declaration of faith, since many considered it possible that the city would remain permanently flooded. At any rate, Callahan's thesis is directly contradicted by Echave's 1606 copy of the Image, which includes the angel and the lower fold.

Callahan is amazed that the original colors of the mantle and cloak survived the first two centuries, when it was exposed to humidity, heat, light and smoke. He measured 600 microwatts of near UV light from a single blessed candle. Hundreds of beeswax candles were present in the old Hermitage. The resultant hydocarbons and ionizations should have destroyed the Image.

Referring to the 15th-century manual Il Libro dell'Arte by Cennino d'Andrea Cennini, Callahan examines the question of whether the Image could be accomplished by Renaissance techniques. First, he notes that Renaissance craftsmanship mandated the use of undersketching in charcoal. His infrared study reveals that there is no such underdrawing in the tilma. "It is extremely unlikely that any 15th or 16th century artist would attempt such a realistic and subtle portrait" without underdrawing, as this would "border on the artistic[ally] impossible" and violate the tenets of Renaissance art. The fur tassel and hands have been outlined over the pink robe, so they are not examples of underdrawing. Further, the lines are in black paint, not charcoal.

The fabric has no sizing, as can be seen especially in the area around the neck and near the fur tassel, where the paint is thin and the space between cross fibers is visible. Renaissance sizing was made of goatskins and sheep parchments, or with lime and cheese. There is no indication in the Virgin's mantle of the special sizing then used for tempering blues, which would form clear crystals. (p. 24)

A typical Renaissance canvas would have multiple layers of sizing. First, a clear glue to impregnate the undercoat fibers. Second, a whitened glue to create a hardened undercoat layer and further fill the interstices. Usually, a clear thin hot size would be added over the paint pigments, which are themselves bound with egg white tempera. Finally a varnish would overcoat the paint. None of these types of preparation are found in the area of the face, hands, and garments.

The surrounding sunburst and clouds, by contrast, are layered and the fabric weave is filled with apparent fresco. As we noted previously, this probably accounts for other claims that the tilma is sized.

Cennini's manual makes clear that adding metallic gold to cloth is not recommended, the wisdom of which is proven by the deterioration of the gold rays. If it must be done, several layers of tempera are required as preparation. (p. 26.)

Jorge Sol Rosales' Examination and Restoration

In 1982, Jorge Sol Rosales, then a young but skillful art restorer, was asked to remount the tilma. He was given only a single night to examine the icon, assess what needed to be done, and accomplish the restoration including remounting. Accordingly, his observations are more cursory than those of Callahan, who had the luxury of a three-hour photo session and prolonged examination of closeup prints.

As Rosales' primary task was to conduct a restoration, not a scientific or technical study, his observations were not published until 1999 in a book by the Guadalupan researcher Leoncio Garza-Valdés, titled La triple imagen. Rosales personally confirmed his participation in a 2002 interview with the leftist political magazine El Proceso. This was sensationalized as "the study that the Vatican tried to suppress," but of course the Vatican has had no effective censorship powers for over a century, especially in secularized Mexico. The abbot of the basilica from 1963 to 1996, Guillermo Schulenburg, was himself a disbeliever in the Image's miraculous origin, so far was the Church from suppressing anti-apparitionists. The real reason for Rosales' silence was fear of public reaction in strongly Guadalupan Mexico.

Curiously, Rosales gives the cloth's dimensions as 172.5 cm x 109 cm, at least 4 cm longer and wider than all previous measurements. It is possible that he is including parts that were folded while mounted. Additionally, he mentions 1 cm "clippings" (recortes) on both borders.

Rosales says there is no protective varnish on the painting, except for an acrylic coating that forms part of the frame.

He identifies the painting techniques as various types of tempera. The technique for the cloak and clothing was known as aguazo in the sixteenth century, "and it derives from the techniques in the painting of the so-called sargas" and requires the cloth to be lightly moistened in order to affix the color. This accounts for the bleeding of the colors through the reverse of the cloth and "the difficulty of finding clear brush strokes".

Rosales does not say that the tilma is a sarga. In fact, this is definitely not the case, as it has a plain weave. Rather, he uses the sarga as an example of how the aguazo technique was used to paint on loosely woven, unmounted cloth. His account of the preparation, which is no different from Cabrera's description, omits the use of any sizing or primer. He expressly agrees that the colors bleed through the reverse, the classical evidence of no sizing in these areas. We have already noted that skilled artists can hide their brush strokes, and Rosales offers that the moisture of the cloth could further obscure them.

The background area, according to Rosales, was done in glue tempera (temple de cola) on top of a preparation also of animal glue (cola) and lime carbonate "slightly heavier than that used in the mantle and tunic." Rosales confirms Callahan's observation that the preparation is plaster-like, containing lime. He insinuates that the cloak and robe area also have a similar, yet lighter, preparation. Yet this is at odds with his claim that these areas were done in aguazo, which, as Cabrera attests, requires an untreated moist cloth. (Cabrera thought the tunic was done in oil and the mantle in ordinary tempera.)

The hands and face are of a third technique; "these are presumably executed with a resin tempera or with oils worked with very thin coats of color" which let the light reflect off of the white base to create a luminosity akin to tempera works. Rosales does not indicate if he finds sizing here, which would be very difficult to assert, given Callahan's presentation of spaces between the weave. Yet the use of oil or resin tempera, as he suggests, would require sizing in order for these painted areas to endure.

Rosales judges that the gold elements were added last, as they fail to match the original contours of the drapery [i.e., of the robe], but instead are a flat [perspectiveless] decoration in the indigenous style. He says they were done in the type of tempera known as labrada. An animal glue is used as a binding agent, followed by gold dust as a pigment, laid in thick films. Unlike Cabrera and Callahan, Rosales fails to distinguish the gold rays from the other gold elements. From IR photography, it is clear that only the former are made of metallic gold. Cabrera claimed that only the rays were made by labrada al temple, while elsewhere the gold was seemingly impressed into the cloth.

Rosales seems to assume that all the elements were made by a single artist, as he remarks on the eclecticism of Indian artists, who incorporated European forms while still retaining some methods of their tradition (e.g., the gold decorations).

Overall, the painting uses a very restricted palette: black, white, blue, green, earth tones, reds, and gold. Rosales makes some confident assertions about the composition of the pigments, though he has made no chemical analysis. "The white is with all assuredness calcium sulfate... The blue green pigments are, with much probability, basic copper oxides known as azurite [blue]... or malachite [green]... The earth tones are iron oxides..." with varying degrees of iron and calcium, these last being commonly used in pre-Hispanic Mexican art. The vermilion was composed of sulfur and mercury and the carmine of Mexican cochineal. He thinks the "mantle" [sic; meaning the robe] was of carmine, while the aura was of vermilion mixed with carmine and possibly some earth red [iron oxide]. He believes the cochineal reds and copper oxides were used on account of these colors bleeding through the cloth. Notwithstanding his confident tone, these claims even from an expert are little more than educated guesses or conjectures, absent a chemical analysis.

More pertinent to his work, Rosales details the state of deterioration of certain parts of the painting. In general, the type of cracking found throughout the painting is arboriform and indicative of age. A different type of cracking is found in the gold rays, affecting 30 percent of their area. Here the cracking takes the form of scaling, revealing the white preparation underneath.

There are also losses of paint along a line 40 cm long and 1 cm wide, at the union of the two cloths. There is abrasion of the paint caused by the crossbeams at distances of 54, 56, 110 and 114 cm from the lower border. Lastly, there is an advanced state of crumbling of the paint and its primer, leaving the support bare in some places, such as the borders of the cuffs, the point of the shoe, and the extreme lower left of the mantle, where there has been a retouching of color.

Rosales' assessment here is consistent with that of Callahan, who found paint loss and abrasion at points of mechanical stress, and in the parts that he judged to be later additions. Note that the areas where Rosales identifies primer—the cuffs, the shoe, and the lower left mantle—are all non-original, according to Callahan's infrared analysis.

There are also stains on the tilma, most notably a humid splotch in the upper right. This is likely what other observers have identified as the nitric acid spill. There are also various small stains, mostly candlewax drippings, including one as large as 1 cm in diameter in the upper right corner. The upper region has a black film that assuredly comes from the candles that were lit before it when it was exposed for the first century of its existence.

Rosales' most controversial claim is his insistence that the painting uses a white primer or sizing. He is aware that other experts have said that the painting is on bare canvas, but he believes this error is due to their observing places where both the painting and the primer have disappeared. He says the primer was applied irregularly, to the point that the areas of color seem to follow it [i.e., to be a continuation of it]. The use of primer or sizing is proven by the presence of drops of this preparation on the reverse of the cloth, having seeped through the looser parts of the weave.

It is unfortunate that Rosales, one of the few modern observers to have seen the reverse of the cloth, was not more specific as to where he saw evidence of primer. The few places he does explicitly mention (the cuffs, shoe, and lower left mantle) are all late additions, and his description of the sizing as "applied irregularly" is suggestive that it was applied only in certain areas, namely the later additions. If the whole cloth were sized, there would be no reason to apply the sizing unevenly, and the unquestionable competence of the original painter precludes the possibility of ineptitude at such a simple task. It seems far more plausible that the sizing was applied irregularly only because it had to be laid on certain areas. When applying a coat in such a manner, there is a tendency to lay it on more thickly at the borders, resulting in the uneven effect.

The work will have to be analyzed for its definitive identification, but microscopic examination (20-80X), its aspect and comportment inclines me to think it is made of linen with some mixture of hemp, I am practically certain that it is not made of hard fibres of the ixtle or henequén types, as is traditionally asserted, nor do I believe that it is cotton.

Rosales does not pretend to have definitely identified the composition of the cloth, which is impossible by mere ocular examination. He used a low-power microscope, without enough magnification to discern, much less measure, cell dimensions on a scale of 20-30 microns. At best he would be able to see flax straw, but would not be able to discern smaller flax cells, nor those of various agave species. Rosales does not explain why he eliminates ixtle. Low-power microscopy could not decide this issue so clearly. It could be because he felt the silky softness of the surface, and judged this to be incompatible with hard fiber. We have learned, however, that light exposure can soften such fibers over time.

The cloth's thread, of medium thickness, is handspun, as shown by its lack of torsion and its irregularity. The cloth is a plain 1/1 weave, with alternation of single threads. The density of the weave varies, as is common in handwoven cloths. It is 10 to 11 threads per centimeter in the weft or woof (crosswise threads), and 12 to 13 threads per centimeter in the warp (lengthwise threads). This is consistent with canvases used in sixteenth century Mexico, ranging from 10 to 14 threads per centimeter (25 to 36 threads per inch, for a "thread count" of 50 to 72).

This is consistent with what baroque testimonies that the cloth was of medium quality. Cabrera himself had painted on canvases with similar thread counts, yet he and others judged it impractical to paint on the tilma. This was owing to the looseness of the weave, that is, the spaces between the irregular threads, as thread count does not tell us everything. The lack of sizing was considered especially marvelous, since the spaces at the interstices were large enough to see through.

Rosales notes that the abrasions were caused by crossbeams of the wood stretcher frame, either the present one or a similar one previously. A defect in the frame's design resulted in a slight lack of tension in the canvas, causing it to sag in the lower center and creating lines of tension in the upper corners and the borders.

Before attempting a restoration, Rosales remarks on previous interventions. First, he notes that the canvas appears to have had its edges cut away on all four sides, perhaps as much as 3 cm. This was likely due to a bad restoration practice, common in colonial times, where the canvas was simply cut off of its old frame when it was remounted, instead of removing the glue. The loss might also be attributable to the practice of taking relics of the tilma.

The present stretcher frame "is of recent manufacture, which almost certainly copies the characteristics of the original stretcher frame," including the design defect mentioned.

Using grazing light (i.e., light parallel to the surface) and ultraviolet light, Rosales made an ocular inspection of the retouches. Notably, the outline of the face was lightly modified. The forehead and cheek were rendered more opaque, while volume was added to the throat, creating a double chin. These observations corroborate claims made by Rodrigo Franyutti.

The Virgin's hands have been repainted, "lightly modifying their original anatomy." There are also retouches at the bottom, the top, and along the union of the two cloths. The retouches in these areas could be done in watercolor or a similar technique, while retouches elsewhere seem to have been done in tempera.

Rosales recommends chemical analyses of the pigments, glues and fibers of the work, in order to determine their composition and assist future restoration work.

For a restoration, Rosales proposed mounting the canvas on a new stretcher frame that provided adequate tension, to prevent further abrasions. Strips of cloth were to be affixed to the edges of the original canvas with a reversible dry glue, to permit mounting without driving nails through the original, as had been done previously. The paint was to be protected using a consolidant called Beva 371, which would not alter the paint's appearance, and would allow the cloth to flex with changes in climate. When remounted, the canvas would have two cloths on its reverse side to serve as dust filters.

Rosales conducted his examination, proposal, and execution of the restoration all in a single night. Accordingly, his judgments are rather perfunctory and hasty in places. Still, given appropriate context, they add significant knowledge regarding the condition and possible composition of the painting. They receive special importance when we consider that his restoration has permanently ended several lines of inquiry. His addition of a varnish makes it impossible to judge if the painting has been preternaturally preserved since 1982. Since the varnish was applied with a brush, there are now visible brush strokes everywhere. The remounting of the canvas means we can no longer examine the reverse side of the cloth. Fortunately, Rosales, himself no believer in the Guadalupan miracle, attests that there was no varnish before then, nor visible brush strokes, and he does not deny the generally good state of preservation of the Image, save for the specific areas he and others have mentioned. He also gives testimony clarifying the probable history of the stretcher frame and the mounting of the canvas, and provides the last testimony, albeit ambiguous, about the bleeding of colors through the reverse.

Dubious Claims of Leoncio Garza-Valdés

Rosales' study came to light through the claims of Leoncio Garza-Valdés, a pediatrician from Texas who moonlighted as a microbiologist with an interest in paranormal phenomena. He presented Rosales' work in the context of his own much more controversial hypotheses, which he discussed with Rodrigo Vera of El Proceso magazine in 1999, and [“La guadalupana: tres imágenes en una.” Por Rodrigo Vera. El Proceso nº 1.334]

Garza-Valdés claims to have studied the Image at least twice, once at around the time of the Rosales restoration, and again in 1999, at the behest of Cardinal Rivera Carrera, using the latest ultraviolet and infrared photography equipment. He thereby "discovered" three superimposed images in the tilma. The first, completely different from the present Image, has the Virgin holding a nude infant Jesus in her arms, and is identical to a wood relief in the monastery of Guadalupe in Extremadura. This original Image has a painted date of 1556 and is signed with the initials M.A. The second is very similar to that of the day, but with a face having more indigenous features. It is supposedly signed "Juan de Arrue Calzonzi," and was done in 1625 The third is that which we know today, done probably in 1632.

No one has ever corroborated Garza-Valdés' extravagant claims. Rosales claims no knowledge of such discoveries, and Guillermo Schulenburg does not mention them in his memoirs, though he had already gone on public record as a non-believer in the apparitions. Other experts who have examined Garza-Valdés' images fail to see any of the features he claims are present. Notably, Dr. Gilberto Aguirre, who accompanied Garza-Valdés at the examination of 1999, told a San Antonio newspaper in 2002:

Dr. Garza-Valdes and I have the same images, but our conclusions are entirely different. I can´t find anyone who agrees with Dr. Garza-Valdes.... Secondly, he claims to not only see two other paintings, but a nude baby Jesus in the arms of the Virgin, as well as the initials M.A. and the date 1556. I have studied these photos, but I do not see these things. [John MacCormack. “Test of faith,” San Antonio Express-News, June 2, 2002.]

It should be noted that Garza-Valdés, unlike Callahan, is not a professional scientist, nor is he an expert in painting or in infrared photographic analysis. His lack of expertise is shown by his employment of ultraviolet photography on the Image while it was still in its glass case, which makes UV imaging practically useless. Nonetheless, his assertions have been uncritically accepted by various anti-apparitionists, which only shows how even so-called "skeptics" are susceptible to confirmation bias.

We already possess enough historiographical information to positively contradict Garza-Valdés on several points. The Image as we know it dates back to at least 1606, where, if anything, she was slightly more European in appearance, not more indigenous. The "M.A." is undoubtedly intended to confirm the investigator's presumption that the artist was Marcos Aquino (a.k.a. Marcos Cipac). Yet the Información of 1556 tends to indicate that the painting (if not its cult) had already existed for some time, or at the very least since 1555, as some sixteenth-century witnesses claim. Further, the early witnesses express perplexity as to why the Image was called Guadalupe, which implies that it did not obviously resemble the Virgin of Extremadura.

This is not even to speak of the technical impossibility of covering or removing such a radically different painting without being detected by Callahan or Rosales. There is nothing magic about today's technology that would make something appear under infrared that was not available to Callahan. His film already had a range of 0.3 to 0.9 micrometers, from UV to infrared, which was more than he needed. The central image had pigments that were mostly transparent to IR, so he would have detected an earlier image if it was there. Notably, the paint on the hands and face is in some places so light as to approach bare canvas, leaving nowhere for an older image to hide. On the whole, then, Garza-Valdés claims are a red herring deserving no credence.

Synthesis of Evidence

The vast majority of scientific and expert analysts of the tilma have drawn conclusions favorable to the thesis that the Image is of miraculous origin, or otherwise inexplicable. The two exceptions are the art restorers José Antonio Flores Gómez and Jorge Sol Rosales, who were performing work for hire. There may be some selection bias here, as those with an interest or devotion toward the Virgin of Guadalupe are more likely to volunteer their time in investigating the phenomenon. Still, this is not sufficient cause to deny the integrity of such investigations, especially those conducted by professional scientists. We find that all of the expert studies are generally consistent with each other in points of fact, even when they are contradictory in matters of interpretation. Accordingly, we can construct a probable account of the Image's composition and condition over time.

The cloth itself is a plain weave of medium coarseness, yet the irregular thickness of its handspun threads causes it to be loose in places, making it practically impossible to paint upon without sizing, at least in standard oil and tempera techniques. If the authenticity of the sample used in the 1946 UNAM study could be verified, we should not hesitate to identify the material as an agave, likely A. popotule, which would account for the silky softness of the light-exposed side. Otherwise, linen or hemp blends would remain possibilities, though cotton is definitely excluded. The thread binding the two cloths together is of uneven thickness, and slender in places. Its material is uncertain.

The original Image consisted at least of the Virgin wearing a blue-green cloak and pink robe, in almost exactly the same form today, though without the gold and Gothic embellishments. The fingers of the left hand originally extended farther and to the left. The face was that of a fourteen to sixteen-year-old girl, very lightly painted, using only white and shadow. The hair was more dark brown than jet black.

All the experts except Rosales (and Bartolache's painters) agree that the area of the original Image is unsized. As evidence, they adduce the seeping of colors through the reverse, which would be impossible if the painting were sized. Also, there are visibly empty spaces between the threads, and there are even bare threads where the brown cloth can be seen. Rosales says the sizing has come off in the bare spots, though this likely does not account for the area of the face, which is very lightly painted. Further, he suggests that the white seeping through the reverse is sizing. He does not address the issue of empty interstices.

The most parsimonious view is that sizing is only present in some areas, likely those that Callahan identified as later additions. Rosales himself acknowledges that other colors besides white have seeped through to the reverse, and he claims that the mantle and tunic were done in aguazo, which does not use sizing. His account of the presence of sizing is confused and self-contradictory. Callahan's IR study resolves the ambiguity, indicating the presence of sizing in the background field, the repainted lower portion, and the Gothic retouches. This accounts for the uneven application observed by Rosales.

There are no signs of any undersketching in the original Image. All outlines are either overlaid or blended into the paint layer. This absence, combined with the Image's marvelous aesthetic use of the cloth's irregular threads and diffractive properties, suggests that the work was beyond human ability.

The original Image had no visible brushstrokes and no varnish. Its pigments have not been positively identified by chemical tests, though various mineral pigments were excluded by Callahan's study. The technique used is likewise mysterious, though the consensus is that most or all of the original Image has the appearance of tempera. The face and hands might be in oil (Cabrera, Rosales) or resin tempera (Rosales), which would account for their luster. Such techniques require sizing, however, yet the facial area is practically bare in places.

Although Cabrera thought the tunic was oil and the mantle was tempera, the restorer José Antonio Flores Gómez claimed that water-based retouches in these areas best matched the original. Similarly, Rosales suggests they were done in aguazo (gouache), which is water-based pigment mixed with a binding agent or white pigment applied to a moist, unsized canvas. Yet water-based paints are notoriously perishable, while the original Image has retained bright and evenly dense colors for over 400 years. Whatever its substance or technique, the tunic pigment is very lightly laid.

The original Image was likely made on loose cloth without a frame. Early in its history, it was cut to size and mounted on a wooden stretcher frame with two crossbeams. Although the original pigments have proven to be preternaturally durable, the painting is not impervious to mechanical stresses on the canvas. The frame's imperfect design caused tension (and eventual paint loss) at the seam, and creasing (with paint abrasion) at the crossbeams.

Some time after the Image was mounted, yet before 1606 (date of the Echave copy), various embellishments in thickly laid coats were added. Callahan's study has determined the probable sequence of these additions. First, the dark moon and the tassel, then the angel and lower fold of the robe, followed by the white background and sunburst. Then came the gold trim with dark outline in the mantle and tunic, and finally, according to Callahan, the stars. The paint in these areas, being laid thick, has cracked and peeled in numerous places.

The white background likely contains lime, hardening the canvas and preventing creasing. The sunburst contains metallic gold (over dirty yellow ochre), while the other gold elements are likely ochre. A gold band once united the upper rays to form a crown. If the stars were added in the late sixteenth century, then the depiction of the Virgin in the Codex Escalada cannot be contemporary with Sahagún's signature. On the other hand, if this part of the codex is authentic, then it follows either that (a) all embellishments were completed by the mid-sixteenth century, or (b) the stars were not late embellishments. In support of the latter consideration, we note that the stars had been retouched in many places, which could account for their overlay of other elements.

The Image was displayed without protection from the smoke and beeswax of votive candles until the mid-seventeenth century, when it received a protective, though not airtight, glass case. During that period, and long afterward, it was repeatedly exposed to be touched and kissed by the devout. Eventually, the hands were repainted, and perhaps the Gothic fur cuffs were added at this time. Despite all this abuse, most of the Image has held up well in proportion to its age. For the first four centuries, there was hardly any noticeable damage except for the golden rays and crown.

The only definitely known early retouches were those in the background, to remove traces of the short-lived cherubim (early 17th century). The gold stars were retouched repeatedly since the 19th century. By the 1940s, there were signs of damage from mold and moisture, as well as peeled paint along the seam. There had already been various light retouches by then, mostly to restore damage from candle smoke in the upper area. Otherwise, the face and cloak were still original.

More noticeable changes were then made to the face. By 1951, the cheeks were rosy. By 1963, the hair was darkened and the outlines emphasized, creating an apparent double chin, while more paint was applied to the cheek. These changes were unnecessary as restorations, since the face was in excellent condition as of 1943. They may have been motivated by a misguided desire to make the subtle features more visible from a great distance.

Amazingly, the tilma was still exposed to ambient air in the old basilica, where it was kept until 1974. Still, there were only minor retouches by the time of Callahan's 1979 study, easily visible due to their lack of patina. In 1982, Rosales found arboriform crackle throughout the painting, but scaling only in the gold rays. There were paint losses along the seam, abrasion at the crossbeams, and crumbling in places that Callahan had identified as late additions (the cuffs and lower portion). Rosales added varnish to prevent further deterioration, operating on the assumption that the entire Image was the work of the same artist.

The original wooden stretcher frame was used until some time in the twentieth century, when it was replaced by another similarly defective frame, until Rosales remounted it in 1982. The canvas was relined in 1947 and again in 1982. Since 1982, the tilma is adequately protected from dust, humidity and mechanical stress for the first time in its long history.

This concludes our study of the icon of Guadalupe, leaving readers with enough information to draw their own conclusions about its miraculous origin and preservation. In the epilogue to follow, I offer my own plausible scenarios that seem most congruent with the bulk of the historiographical evidence.

Continue to Part XIV

© 2013 Daniel J. Castellano. All rights reserved. http://www.arcaneknowledge.org

Back Top