Abstract
The fourth-century Greek Bible manuscript Codex Vaticanus B (Vat. gr. 1209) contains pairs of horizontally aligned marginal dots known as distigmai, which correspond to textual variants in other manuscripts. The production of the distigmai has been variously dated to the 4th or 16th centuries. A fourth-century date would prove the early existence of hundreds of textual variants, many of which are otherwise only witnessed by later manuscripts. Near infrared microscope reflectography combined with micro-X-ray fluorescence spectroscopy proved that the distigmai, both those categorized as “apricot” (which are visually similar to the ink of the original main text) and “chocolate” (which are visually similar to the reinking of the main text) were written with ink(s) made from chemically purified vitriol, a process that only became standard in the 16th century. As a result, there is no reasonable chance that the distigmai were written in the fourth century. Horizontal lines that have been hypothesized to function as text-critical obeloi were written in the same ink as the original main text, which differs completely from that of the distigmai. In other words, the distigmai and horizontal lines tested were not produced during the same writing session and are separated by more than 1,000 years, making it impossible for them to have functioned as conjoining text-critical symbols in the 4th century.
Codex Vaticanus B is among the most important manuscripts of the Bible in Greek.1 It was written in majuscule script in the fourth century and mostly laid out in three columns. The codex preserves most of the Old Testament and New Testament. Some pages were replaced in minuscule script in the 15th century.2 A scribe reinked the entire manuscript, systematically tracing over faded letters with new ink. This reinking obscured the original shapes of most of the letters, making it difficult to differentiate between the original hands that wrote the manuscript.3 This is a major reason why there is no consensus as to whether the original text was produced by two, three, or four scribes.4 Nevertheless, the reinker did leave many letters and words unreinked, especially in the case of what he considered to be errors.5 Pietro Versace recently carried out a thorough survey of over 30 different scribal features, arguing that Codex Vaticanus was reinked twice, first in the 10th or 11th century and again in the 16th century.6
1 The Distigme Debate
1.1 Discovery of the Distigmai (“Umlauts”)
In 1995, Philip B. Payne identified pairs of horizontally aligned dots in the margins of Codex Vaticanus that he originally called “umlauts”, but which were later dubbed distigmai (singular: distigme) meaning “two dots”. The number of distigmai has been estimated between 765 and 858.7 According to Payne, the original scribe of Codex Vaticanus had access to other manuscripts in the fourth century, which he systematically compared to Codex Vaticanus, marking the location of textual variants with distigmai (Figure 1).
In 2000, Payne along with Msgr. Paul Canart of the Biblioteca Apostolica Vaticana carried out ground-breaking work by directly examining Codex Vaticanus with the naked eye, a magnifying glass, and a 7× magnification loupe with internal illumination. They proposed a dichotomy of original “apricot” distigmai and “chocolate” reinked distigmai.8 With visual observations as the only available criterion, they correctly noticed that the color of the apricot distigmai matched the original unreinked text whereas the chocolate distigmai corresponded to the dark brown ink of the reinker. They also found two instances of apricot distigmai “protruding” from chocolate distigmai.9 Based on this, they argued that chocolate distigmai were the product of reinking original apricot distigmai. Notable exceptions are distigmai that Payne argued were added by the reinker or possibly by an unspecified later hand and therefore do not have apricot distigmai hidden underneath them.10
Figure 1
A distigme (“two dots”) in the left margin of Codex Vaticanus 1474 A20. According to Philip Payne, this distigme and the adjacent horizontal line are an example of a so-called distigme-obelos pair, indicating that the original scribe of Codex Vaticanus in the 4th century was aware of other manuscripts that omitted the verses in 1 Cor 14:34–35 telling women to be silent in church. He argued that both the distigme and the horizontal line were reinked in the Middle Ages. According to Curt Niccum and Pietro Versace, the distigme was added in the 15th or 16th century and the horizontal line is an unrelated 4th century paragraphos, which divided the text into sense units.
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
The overwhelming statistical correspondence between the distigmai and textual variants known from modern critical editions further suggested to Payne and Canart that the chocolate distigmai were likely to be reinkings of earlier apricot distigmai. As they explained in their 2000 study:
The ink of most of the umlauts in the Vaticanus NT matches the chocolate-brown ink of the reinforced text […] It is implausible if Codex Vaticanus had only eleven original umlauts that a later scribe would have identified their purpose, let alone expanded their use. It is also implausible that a scribe half a millennium later would simply by chance have used the same symbol that the original scribe had used to mark the location of textual variants, especially since it never became conventional after the writing of Vaticanus in the fourth century for scribes to use umlauts for this purpose. Thus, it is far less likely that the reinforcer in the Middle Ages originated these umlauts than that he simply traced over them while reinforcing the rest of the text. It is reasonable to expect that the chocolate-brown umlauts the reinforcer traced, like the text itself and the apricot umlauts that were not reinforced, also date to the original writing of the codex.11
Hence, according to this hypothesis, most of the chocolate distigmai correspond to the locations of original apricot distigmai, which were in turn part of the original fourth-century production of the codex or added soon thereafter. The original distigmai, both those written with apricot ink, and those supposedly reinked in the Middle Ages with chocolate ink, correspond to places where a fourth-century scribe compared Codex Vaticanus with other manuscripts and found textual variants. In their 2009 study, Payne and Canart reiterated this hypothesis:
[…] distigmai matching the dark chocolate-color ink used to reinforce the entire manuscript in the Middle Ages were probably reinforced along with the rest of the manuscript. […] unless there is evidence to the contrary, the presumption is that the dark chocolate color distigmai are reinforcements of original apricot color distigmai and also date to the original writing of Codex Vaticanus, or at least to the early period when it was in the scriptorium with manuscripts whose variants they note.12
If the 700+ distigmai are original, their importance is immense for textual criticism of the New Testament. J. Edward Miller, who accepted the originality of the distigmai, explained:
[H]undreds of variants, emerging only in later manuscripts, can now be dated to the early-fourth century with some measure of confidence. Lastly, for those marked lines failing to yield extant variants, New Testament scholars must acknowledge the likelihood that some variants known to exist in the early-fourth century have been lost.13
In other words, according to this hypothesis, the distigmai serve as a fourth-century witness for numerous variants only known from later manuscripts. Furthermore, when a distigme does not correspond to a significant variant known from other manuscripts, it opens the door for conjectural emendations that can, in a sense, be categorized as textual variants, employing the distigmai as a witness.
1.2 Color and Date of the Distigmai
Payne and Canart initially observed that “eleven umlauts unambiguously match the original ink of Codex Vaticanus,” which is correct based on a strictly visual assessment.14 Payne later identified a further batch of apricot distigmai in the 1999 facsimile edition of Codex Vaticanus.15 In 2009, Canart examined these distigmai, finding some new ones in the process, “first with the naked eye, then with a magnifying glass, and finally with different types of loupes.”16 In 2022, Payne referenced another apricot distigme that he identified in the facsimile edition, although this was not confirmed by a direct examination in the original.17 This made for a total of fifty-four apricot distigmai, fifty-three of which were confirmed through direct examination of Codex Vaticanus.18
1.3 Paragraphoi (Small Horizontal Marginal Lines)
Throughout the New Testament portion of Codex Vaticanus, horizontal lines in the margin mark a new section. These horizontal lines, usually dubbed paragraphoi, are often accompanied by a space in the text where the new section begins.19 In his original 1995 study, Payne argued that some horizontal lines, when accompanied by a distigme, served as an obelos (see Figures 1, 12, 14, and 16). Obeloi are known in Codex Vaticanus in the Prophets where they mark text in the Greek that does not appear in the Hebrew. Indeed, there are a number of explicit notes in the Prophets portion of Codex Vaticanus that explain, “the [lines] marked with obeloi are omitted in the Hebrew”.20 According to this hypothesis, the distigme-obelos combination in the New Testament serve as “conjoining symbols” that mark textual variants of four or more words.21
Payne identified fifteen distigme-obelos pairs that he maintained were written by the original scribe of Codex Vaticanus. However, he only identified the original apricot-colored ink in three of the distigmai and one of the horizontal lines in these fifteen distigme-obelos pairs. Accordingly, the original apricot-colored ink in the remaining twelve distigmai and fourteen obeloi in these fifteen pairs should presumably be hidden beneath the dark ink of the reinker.22
1.4 Reception of Payne’s Hypotheses
In a post-script to their 2009 joint study, Canart distinguished between the phenomena that he and Payne discovered and the interpretation of those phenomena, calling for “further investigation” to “confirm or qualify Payne’s […] conclusions.”23
Christian-Bernard Amphoux accepted the fourth century origin of the distigmai and noted a “consensus” about their text-critical function. However, Amphoux explained them as stemming from a comparison of Codex Vaticanus with Old Latin rather than Greek manuscripts.24
Curt Niccum also accepted the text-critical function of the distigmai but argued that they were added in the 15th century or later. This scholar noticed that there were distigmai on the first page of the last quire of Codex Vaticanus, which was replaced in the 15th century by a scribe writing in minuscule script.25 Although it was possible that the minuscule copyist reproduced these distigmai from a damaged fourth-century page, he found it more likely that all the distigmai post-dated this 15th century replacement quire.26 He also suggested the distigmai might have been added by the Spanish humanist Juan Ginés de Sepúlveda (c. 1490–1573) who served the Curia in Rome under Pope Clement VII.27 In 1533, Sepúlveda sent the Dutch scholar Erasmus a list of 365 textual variants in Codex Vaticanus. Although this list is no longer extant, Erasmus later cited four readings from Codex Vaticanus, each of which is marked in the codex itself with a distigme in the margin.28 According to Niccum, it was probably Sepúlveda who marked over 700 textual variants in Codex Vaticanus with distigmai but sufficed with only 365 of these variants in his letter to Erasmus.29 James Snapp argued that Sepúlveda actually wrote 765 in his now lost original letter to Erasmus but when the letter was published in 1557 it was mistakenly changed to 365.30
For the past thirty years, the strongest argument against Niccum’s late dating has been the correspondence in color between the apricot distigmai and nearby unreinked main text contrasted with the visual correspondence between the chocolate distigmai and the reinking of the main text. In recent years, Pietro Versace had access “on various occasions” to the original during his thorough inventory of the marginalia of Codex Vaticanus.31 Versace dated the distigmai to the 16th century based on their similarity to the Vulgate chapter numbers inscribed in the margins of the NT portion of Codex Vaticanus in Arabic numerals. He noted that the color of the ink of these Arabic numerals is “identical” (identico) to that of the distigmai.32 One of Versace’s examples is an apricot distigme in the left margin of 1276 C31.33 Versace compared this apricot distigme to the Arabic numeral “8” in the right margin of 1276 B34, arguing that they had the “same color” (stesso colore).34 He also noted that the Arabic numerals often have dots on either side of the number, which are the same “form” (forma) as the distigmai.35 Versace concluded that the distigmai were probably added in the 16th century by the same hand that added the chapter numbers in Arabic numerals.36 Hence two competing arguments (Payne and Canart vs. Versace) date the distigmai variously to the 4th century or the 16th century based on a visual appraisal of their color being the same as some other textual element of known date (unreinked text vs. Arabic numerals, respectively).
Overall, the hypothesis about the text-critical function of the distigmai received large acceptance in the scholarly community despite disagreement about their date, whereas the hypothesis that the New Testament portion of Codex Vaticanus contains horizontal lines that function as obeloi has been generally rejected.37
2 Preliminary Methodological Considerations
2.1 Microscopic Reflectography
Our research aimed to determine the similarities and dissimilarities between the inks of different scribal features based on material science instrumental analysis rather than visual appearance and color alone. Our instrumental approach was based on an ink-study protocol routinely employed by the BAM since 2012.38 In short, we start with the identification of the ink type based on the optical properties of three main classes of ink used in Europe and the Middle East since late Antiquity.39 These three ink classes correspond to, a. soot ink, a fine dispersion of carbon pigments in a water-soluble binder, b. extracts from tree bark known as plant or tannin ink, and finally, c. iron-gall ink produced through a chemical reaction between bivalent iron and gallic or tannic acid derived from either gall nuts or tree bark.40 Interaction with near infrared light is indicative for an ink type and can be visualized by a simple camera without an infrared filter. We used a non-invasive non-destructive USB-microscope Dino-Lite (AD4113T-I2V) to compare ink images registered in visible and near infrared light: no change in ink color corresponds to soot ink, complete disappearance of ink might be indicative of a tannin or mainly deteriorated iron-gall ink, and well-preserved iron-gall ink loses some of its opacity at ~940 nm. All the writing inks discussed in this paper are of the iron-gall type. In addition, we used ultraviolet (UV) illumination to enhance the visibility of faded ink, using the fact that tannins in the ink quench the fluorescence of parchment resulting in a better contrast of the ink.
(a)
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
(b)
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
Figure 2
(a) Nehemia Gordon examining a bifolio of Codex Vaticanus with a Dino-Lite (AD4113T-I2V) mounted on a custom light stand. The surface of the parchment is protected from the Dino-Lite with a layer of Melinex. The light table beneath the bifolio was used to shine light through the parchment to verify there was no ink on the other side of the specific spot we wanted to later test with XRF. Each spot was imaged at ~50x magnification in 1. visible, 2. ultraviolet (~395 nm), 3. near infrared (~940 nm), and 4. transmission. (b) Detailed view of the instrument
Before carrying out the second step of our protocol using XRF (see below), we needed to determine whether a specific spot was a good candidate for further testing. With some exceptions, any spot with more than one layer of ink cannot produce meaningful test results. This is because the XRF spectrometer cannot distinguish between the signal it receives from one layer of ink and that which it receives from another. Multiple layers of ink can be a problem from reinking but also from ink on the other side of the page since X-rays pass through the parchment. For each test, we had to examine the exact spot we wanted to test using transmission light, that is, shine a light through the back of the parchment in order to verify that there was nothing on the other side (see Figures 2, 3, and 11).
2.2 Micro-X-ray Fluorescence
The second step of our protocol involved identification of the metals in the ink by means of micro-X-Ray fluorescence spectroscopy (XRF) analysis, which is also a non-invasive, non-destructive process. An object irradiated with kiloelectronvolt X-rays responds with an X-ray emission, which is typical of its chemical elements, delivering therefore the elemental composition of the object. To compare different iron-gall inks, we employ a modified semi-quantitative fingerprint model.41 In short, the XRF spectra are mathematically fitted using the native Artax software, the contribution of the parchment is subtracted, and the net peak intensities of the metallic impurities found in the vitriol precursor are calculated and normalized to that of iron, the main chemical element in iron-gall ink. The resulting characteristic ratios do not depend on the ink thickness and constitute what is referred to as a metallic “fingerprint” for each ink.
(a)
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
(b)
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
(c)
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
(d)
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
Figure 3
An apricot distigme in the left margin of 1332 B10 in visible (a), transmission (b), near infrared (c), and ultraviolet (d) light. To the right of the distigme is a gamma (
(a)
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
(b)
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
Figure 4
(a) A bifolio of Codex Vaticanus mounted beneath the Artax 800 XRF spectrometer. The archival mounting board has a hole beneath the part of the bifolio being tested, and is raised off the table, to avoid picking up any signals from anything other than the manuscript. Ira Rabin operates the Artax using the laptop on the left while the laptop on the right displays ~50x Dino-Lite images taken during an earlier session to help guide her in finding the precise spot to perform the test. The red light at the top warns that X-rays are actively being emitted. (b) Detailed view of the instrument
We used Artax 800, a scanning 30-Watt XRF spectrometer (Bruker Nano GmbH) that features a molybdenum X-Ray tube, focusing polycapillary optics with an interaction spot of 100 µm, and a 30 mm2 SDD detector (see Figure 4). The experimental parameters were set to 50 kv, 600 µA with an acquisition time of 20 seconds per each spot. We used 10-spot line-scans for each ink test so that 10 individual measurements are accumulated into a single spectrum. When analyzing very small scribal features, such as dots, we sometimes performed multiple shorter scans on the scribal feature and combined the results.
2.3 Iron-Gall Ink
Iron-gall inks utilize a chemical reaction between a soluble iron (II) compound (such as iron (II) sulfate) and gallic acid or tannic acid (extracted from gallnuts or tree bark). The reaction produces an ink that turns black upon oxidation in air. Iron-gall ink degrades with time, changing its color to various shades of brown.42 The earliest reference to this chemical reaction, a precursor to iron-gall ink, is thought to be a recipe of Philo of Byzantium (3rd century BCE). Philo’s recipe describes an ink made of gallnut water that is invisible on leather, which turns visible when rubbed with a sponge soaked in a solution of
The earliest inks with tannin and
The earliest known examples of iron-gall ink confirmed through instrumental analysis are the Vercelli Gospels codex on parchment in Latin from the mid-4th century and a number of Coptic papyri from the 4th–5th centuries, with the most notable being the Akhmimic Book of Proverbs (Berlin, Staatsbibliothek, Ms or.oct. 987) and the Roll of Athanasius.51 Our Dino-Lite reflectography examinations (together with XRF tests) found that the inks in Codex Vaticanus currently under discussion were all of the iron-gall type, which partially lose opacity under near infrared light (see Figures 3, 9, and 11).52 The original ink of Codex Vaticanus can now be added to the handful of the earliest examples of iron-gall ink proven through instrumental analysis and the oldest example of iron-gall ink currently confirmed in Greek.
In the first century CE, the Greek physician Pedanius Dioscorides and the Roman naturalist Pliny the Elder described ancient techniques for obtaining
Medieval Latin referred to metallic sulfates as vitriolum from the Latin word for glass (vitrum) because they formed as glassy crystals.55 Naturally occurring vitriol exists as a mixture of hydrated metallic sulfates, which can incorporate the minerals melanterite (FeSO4⋅7H2O), chalcanthite (CuSO4⋅5H2O), and goslarite (ZnSO4⋅7H2O).56 Vitriol’s crystalline lattice can also incorporate metals such as iron, copper, and zinc in different proportions in a single molecule.57 The vitriolic solutions from which vitriol was extracted could also include metals other than iron. The presence of the other metals besides iron did not contribute to an ink’s black color. The proportions of the non-iron metallic impurities relative to iron in a particular ink can serve as an elemental “fingerprint” that reflects the raw vitriol source used to make the ink.58
In the ancient and medieval world, there was no awareness that vitriol contained impurities. By the 16th century, however, it was common to chemically purify the vitriolic solutions from which vitriol was extracted using the “cementation” process.59 This consisted of immersing solid iron in the vitriolic solution, causing the copper to precipitate out.60 The purpose was not to purify the vitriolic solution, it was to extract copper, but the byproduct was an iron-enriched vitriolic solution nearly devoid of copper.
The Chinese began using the cementation process on a commercial scale to extract copper from vitriolic solutions in 1086.61 The earliest confirmed evidence for the cementation process in Europe comes from the town protocol of Smolník (Schmöllnitz) from the year 1439.62 The Swiss alchemist Paracelsus (1493–1541) brought cementation as proof that alchemy could “transmute” iron into copper, even though transmuting base metals into gold eluded him.63 The Italian metallurgist Vannoccio Biringuccio (1480–1537) seems to have understood that the copper came from the vitriolic solution itself.64 It was not until 1675 that Robert Boyle proved the copper came from the vitriolic solution rather than the transmutation of iron.65 Whether or not the mechanism for removing copper was understood, it was standard practice to chemically purify the vitriolic solutions from which vitriol was extracted by the 16th century.
3 Results
In the following chapter we will present the results of our XRF tests. In section 3.1, we will look at the evidence of the raw XRF spectra to compare inks made from unpurified vitriol with those made from purified vitriol. In the rest of the chapter we will draw on semi-quantitative analysis of the spectra to: (3.2) compare the ink of the distigmai to the original main ink and reinking, (3.3) explain how we overcame an important methodological challenge, (3.4) consider whether there is evidence for the apricot distigmai being reinked with chocolate ink, (3.5) discuss the distigme-obelos pairs, and (3.6) compare the distigmai to other late scribal feature.
3.1 XRF Raw Spectra
The raw spectra produced by the Artax XRF spectrometer demonstrate the profound difference between inks made from raw unpurified vitriol (Figure 5) and those made from chemically purified vitriol precursors (Figure 6). To focus on the relevant differences, we present the portions of the spectra between 3 and 9.5 keV, covering the energy range of interest that contain the peaks of potassium (K, at 3.13 keV), calcium (Ca, at 3.69 keV and 4.01 keV), iron (Fe at 6.40 keV and 7.06 keV), copper (Cu at 8.04 and 8.90 keV), and zinc (Zn at 8.64 keV). The height of the individual peaks corresponds to the recorded count rate in the range 0–30,000 counts per second. For comparison, we added a portion of the XRF spectrum of parchment from corresponding pages to each category (dotted curves). The images to the left of the XRF spectra show the text with the yellow lines indicating the exact locations of the tests.
Figure 5a shows the spectra of the original ink of the main text and a supralinear correction by the reinker, both on an original page (1308 C21). Comparison with the parchment spectrum (dotted) reveals a significant presence of iron, followed by the enhancement of copper in both ink spectra. Similarly, the spectra in Figure 5b display the presence of varying levels of impurities in the fifteenth-century Jeromian Canon recorded on flyleaf IV (see below) and the fifteenth-century main text on a minuscule replacement page (1519 A4). The Jeromian Canon has more copper than iron and a pronounced peak of zinc whereas the main ink of the replacement pages has half as much copper as iron and only trace amounts of zinc. We see, therefore, that the spectra of all 4 scribal features demonstrate the presence of significant copper impurities, characteristic of inks before the 16th century, made from vitriol precursors that were not chemically purified. The metallic impurities in the ink of the replacement pages and the Jeromian Canon make them distinct from each other as well as distinct from the original ink and that of the reinker.
In sharp contrast to the spectra in Figure 5, the spectra in Figure 6 display only traces of metallic impurities attributed to vitriol. Figure 6a shows the spectra of apricot (1308 B27) and chocolate (1308 A11) distigmai both on an original page and a distigme on a minuscule replacement page (1519 A12). Figure 6b shows the spectra of Arabic numeral chapter numbers on an original page (1308 C31) and on a minuscule replacement page (1519 B8). Figure 6c shows the spectra of a Roman numeral chapter number on an original page (1308 C30) and the 16th century Table of Contents on flyleaf III (see below). The spectra of all 7 of the scribal features in Figure 6 are nearly devoid of copper and are characteristic of chemically purified vitriol, which appears in Europe in the 16th century. Since there are only traces of impurities, it is difficult to differentiate between these scribal features as far as XRF is concerned. The different intensities of the iron signal may be due to the thickness of each ink and hence cannot be used to distinguish them in any meaningful way.
Figure 5
XRF spectra (right) reflecting the metallic impurities of raw unpurified vitriol. Micrographic images of the scribal features tested (left) display yellow lines indicating the precise location of the measurements. The energy range (3 keV to 9.5 keV) chosen in the presentation covers the distribution of the elements of interest, i.e., from potassium (K) to zinc (Zn). Figure 5a shows original ink (red curve), reinking (green curve), and parchment (grey dotted curve) from an original page (1308). Figure 5b shows the Jeromian Canon from flyleaf IV (line 12) with the corresponding parchment spectrum (green solid and dotted curves, respectively) and the main text from a replacement page (1519) with the corresponding spectrum of parchment (red solid and dotted lines, respectively).
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
3.2 XRF Tests on Distigmai, Original Ink, and Reinking
Semi-quantitative analysis of the XRF spectra can serve as a tool for distinguishing between inks that visually appear to be identical.66 The present section will focus on the distigmai as compared to the original ink and reinking of the main text of Codex Vaticanus. The primary metallic impurity detected throughout our tests of the main text was copper (Cu). The results are expressed below as a value normalized to iron (Fe), meaning that the net intensity of iron is taken as a full unit, while the net intensities of the impurities are determined relative to it. For example, if the copper value given is 0.12, that means that for every 100 units of intensity of iron detected in our tests, 12 units of copper were also detected. Due to the limitations of the XRF spectrometer used, a 10 % margin of error should be assumed. So, a copper value of 0.12 should be understood as 0.108–0.132. Zinc, which is a naturally occurring impurity sometimes found in vitriol, also appeared in significant quantities in one of the later inks.
Throughout the codex we performed 62 tests on 32 distigmai, sometimes testing both the left dot and the right dot or multiple tests on a single dot.67 This included 20 tests on 10 “apricot” distigmai (see Table 1), 35 tests on 20 “chocolate” distigmai, 3 tests on a distigme said to have been “added later”, and 4 tests on a distigme from a 15th century replacement page (Table 2).68 Among these were 9 tests on 4 chocolate distigmai and 5 tests on 3 apricot distigmai that were part of the fifteen distigme-obelos pairs hypothesized to have been penned by the original 4th century scribe (and in the case of the chocolate, later reinked).69 The intensity of copper to iron in all the tests of the distigmai was between 0.00 and 0.02.70
The first major result of our tests was that there was no difference between any of the distigmai, whether apricot or chocolate, as far as the XRF tests could detect. The distigme said to have been “added later” was identical to the apricot and chocolate distigmai. The distigme on the 15th century replacement page was also no different from the apricot and chocolate distigmai. All the distigmai tested had minute amounts of metallic impurities, indicating that they were written with ink, or inks, made from vitriol that was chemically purified. As a result, there is no reasonable chance that the distigmai were written in the fourth century. Indeed, based on the current state of knowledge, ink made from chemically purified vitriol is unlikely to be much earlier than the 16th century.
The presumption that chocolate distigmai have apricot distigmai hidden underneath them ostensibly posed a methodological problem for our technique. If the XRF tests had found significant amounts of metallic impurities in the chocolate distigmai, there would be no way to know which portion of the signal came from the visible chocolate layer and which came from the presumed hidden apricot layer. However, the near lack of metallic impurities in the ink of the chocolate distigmai means that even if there were apricot distigmai hidden beneath them, neither layer had significant amounts of metallic impurities.
The second very important result of our tests is that the distigmai were not written using the ink used to write the original main text in the 4th century. Indeed, the elemental profile of the distigmai contrasts sharply with the elemental fingerprints of the original main text and its reinking. The original main text had copper to iron values ranging from 0.10 all the way up to 0.38, indicating that different parts of the codex were written with different inks or different batches of ink. It is not unusual for a lengthy manuscript to be written with inks that have a range of elemental fingerprints.71 As a scribe runs out of ink, he would replace it with another batch that may have used vitriol with different metallic impurities.
However, what is surprising is that the reinking of the main text had a range of copper values from 0.09 to 0.38, the same range of inks seen in the original text. On some pages the original ink has a distinct elemental fingerprint from that of the reinker and on other pages the elemental fingerprints are indistinguishable. Curiously, on some pages the original ink had a different fingerprint from that of the reinker on that page but a fingerprint that is indistinguishable from that of the reinker on a different page. This is so profoundly unusual that it requires a future study of its own. Whatever the explanation, we carried out 104 tests on the original main ink and 58 tests on the reinking of the main text and none of these tests had an elemental profile nearly devoid of metallic impurities as consistently found in both the apricot and chocolate distigmai. In other words, the distigmai, whether apricot or chocolate, were not written in the same inks as the original text nor were they written in the same inks as the reinking.
Figure 6
XRF spectra (right) of ink made from purified vitriol. Micrographic images of the scribal features tested (left) display yellow lines indicating the precise location of the measurements. The energy range (3 keV to 9.5 keV) chosen in the presentation covers the distribution of the elements of interest, i.e., from potassium (K) to zinc (Zn). Figure 6a shows spectra of “apricot” (1308 B27, green curve) and “chocolate” (1308 A11, red curve) distigmai from an original page and a distigme (1519 A12, blue curve) from a replacement page. The parchment spectra for pages 1308 and 1519 appear in dotted grey and blue curves, respectively. Figure 6b shows spectra of Arabic numerals from an original page (1308 C31, red curve) and a replacement page (1519 B8, green curve) with the corresponding spectra of the parchment as red and green dotted curves, respectively. Figure 6c shows the spectra of a Roman numeral on an original page (1308 C30, green curve) and the Table of Contents (flyleaf III A16, blue curve).
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
Table 1
XRF tests of apricot distigmai
|
Page |
Column |
Line |
Margin |
Dot Tested |
Type |
Cu |
Zn |
Willker No.72 |
|---|---|---|---|---|---|---|---|---|
|
1241 |
B |
9 |
left |
left |
apricot |
0.02 |
0.00 |
20 |
|
1276 |
C |
31 |
left |
left |
apricot |
0.01 |
0.00 |
93 |
|
1276 |
C |
31 |
left |
right |
apricot |
0.01 |
0.00 |
93 |
|
1279 |
B |
1 |
left |
left |
apricot |
0.01 |
0.00 |
102 |
|
1279 |
B |
1 |
left |
right |
apricot |
0.01 |
0.00 |
102 |
|
1279 |
C |
41 |
right |
left |
apricot |
0.01 |
0.00 |
105 |
|
1279 |
C |
41 |
right |
right |
apricot |
0.01 |
0.00 |
105 |
|
1285 |
C |
14 |
right |
left |
apricot |
0.00 |
0.00 |
117 |
|
1285 |
C |
14 |
right |
right |
apricot |
0.00 |
0.00 |
117 |
|
1285 |
C |
14 |
right |
right |
apricot |
0.00 |
0.00 |
117 |
|
1308 |
B |
27 |
left |
left |
apricot |
0.00 |
0.00 |
159 |
|
1308 |
B |
27 |
left |
left |
apricot |
0.00 |
0.01 |
159 |
|
1308 |
B |
27 |
left |
right |
apricot |
0.00 |
0.00 |
159 |
|
1332 |
B |
10 |
left |
left |
apricot |
0.01 |
0.00 |
200 |
|
1332 |
B |
10 |
left |
right |
apricot |
0.01 |
0.00 |
200 |
|
1332 |
B |
15 |
left |
left |
apricot |
0.00 |
0.00 |
201 |
|
1332 |
B |
15 |
left |
right |
apricot |
0.01 |
0.01 |
201 |
|
1332 |
C |
20 |
left |
left |
apricot |
0.01 |
0.00 |
202 |
|
1332 |
C |
20 |
left |
right |
apricot |
0.00 |
0.01 |
202 |
|
1469 |
A |
3 |
left |
right |
apricot |
0.01 |
0.00 |
623 |
3.3 Unreinked and Reinked Pairs
Distinguishing between the original inks and that of the reinker admittedly posed a methodological problem of its own. Our identification of each of these respective categories hinged on our visual appraisal of their color and appearance, which can be at times subjective. Indeed, in some instances it is difficult to distinguish the original ink from that of the reinker based on color alone (see Figure 7).73 To complicate matters, Versace argued that the entire codex was reinked twice, once in the 10th–11th centuries by a reinker using ochre-colored ink and again in the 16th century by a reinker using dark-colored ink,74 a topic that will be explored in a future study.
To be cautious in overcoming this methodological challenge, we can narrow our comparison to the subset of unreinked letters at the end of a line that were replaced with the same letter in the left margin of the following line.75 This may have been because the original scribe and the reinker had different ideas about the syllabification of a word. According to Versace, these letters were replaced by the ochre-reinker in the 10th–11th century but not reinked by the dark-reinker in the 16th century.76 Our examination at ~50x magnification confirms that in each case the letter added by the reinker on the second line does not appear to have two layers of ink (see Figure 8). We tested 8 pairs of this type. In some instances, the original ink had the same elemental fingerprint as the reinker’s replacement on the next line and in others they were distinct. The copper values for the original letters ranged from 0.12 to 0.37 and that of the letters inscribed by the reinker on the next line ranged from 0.17 to 0.36 (see Table 3).
Table 2
XRF tests of chocolate distigmai on the original pages and a distigme on a 15th century replacement page
|
Page |
Col. |
Line |
Margin |
Dot Tested |
Type |
Cu |
Zn |
Willker No. |
|---|---|---|---|---|---|---|---|---|
|
1243 |
A |
12 |
left |
left |
chocolate |
0.00 |
0.00 |
24 |
|
1243 |
A |
12 |
left |
right |
chocolate |
0.01 |
0.00 |
24 |
|
1243 |
C |
40 |
right |
left |
chocolate |
0.01 |
0.00 |
26 |
|
1243 |
C |
40 |
right |
right |
chocolate |
0.01 |
0.01 |
26 |
|
1244 |
C |
40 |
left |
left |
chocolate |
0.01 |
0.00 |
29 |
|
1244 |
C |
40 |
left |
right |
chocolate |
0.01 |
0.00 |
29 |
|
1253 |
A |
38 |
right |
left (light portion) |
chocolate |
0.01 |
0.00 |
— |
|
1253 |
A |
38 |
Right |
left (dark portion) |
chocolate |
0.01 |
0.00 |
— |
|
1253 |
A |
38 |
right |
right (dark portion) |
chocolate |
0.01 |
0.00 |
— |
|
1255 |
A |
39 |
left |
right |
chocolate |
0.01 |
0.00 |
48 |
|
1255 |
B |
3 |
left |
right |
chocolate |
0.01 |
0.00 |
49 |
|
1255 |
B |
3 |
left |
right |
chocolate |
0.01 |
0.00 |
49 |
|
1255 |
C |
1 |
right |
left |
chocolate |
0.01 |
0.00 |
52 |
|
1255 |
C |
1 |
right |
right |
chocolate77 |
0.01 |
0.00 |
52 |
|
1276 |
A |
18 |
left |
left |
chocolate |
0.01 |
0.01 |
92 |
|
1276 |
A |
18 |
left |
right |
chocolate |
0.01 |
0.00 |
92 |
|
1279 |
A |
22 |
left |
left |
chocolate |
0.01 |
0.00 |
100 |
|
1279 |
A |
27 |
left |
left |
chocolate |
0.01 |
0.00 |
101 |
|
1279 |
B |
20 |
left |
right |
chocolate |
0.00 |
0.00 |
103 |
|
1279 |
B |
20 |
left |
left |
chocolate |
0.00 |
0.00 |
103 |
|
1279 |
B |
26 |
left |
right |
chocolate |
0.01 |
0.00 |
104 |
|
1279 |
B |
26 |
left |
right |
chocolate |
0.00 |
0.01 |
104 |
|
1279 |
B |
26 |
left |
right |
chocolate |
0.00 |
0.00 |
104 |
|
1285 |
B |
12 |
left |
right |
chocolate |
0.00 |
0.00 |
116 |
|
1285 |
B |
12 |
left |
right |
chocolate |
0.00 |
0.00 |
116 |
|
1308 |
A |
11 |
left |
left |
chocolate |
0.01 |
0.00 |
158 |
|
1308 |
A |
11 |
left |
right |
chocolate |
0.00 |
0.00 |
158 |
|
1338 |
C |
33 |
left |
left |
chocolate |
0.00 |
0.00 |
218 |
|
1390 |
A |
32 |
left |
left |
“added later” |
0.01 |
0.01 |
309 |
|
1390 |
A |
32 |
left |
right |
“added later” |
0.01 |
0.01 |
309 |
|
1390 |
A |
32 |
left |
right |
“added later” |
0.01 |
0.01 |
309 |
|
1444 |
C |
8 |
left |
right |
chocolate |
0.02 |
0.00 |
506 |
|
1444 |
C |
21 |
left |
right |
chocolate |
0.01 |
0.01 |
507 |
|
1453 |
A |
30 |
left |
left |
chocolate |
0.01 |
0.01 |
548 |
|
1453 |
A |
30 |
left |
right |
chocolate78 |
0.01 |
0.01 |
548 |
|
1469 |
A |
3 |
left |
left |
chocolate |
0.01 |
0.00 |
623 |
|
1469 |
A |
3 |
left |
right |
chocolate |
0.01 |
0.00 |
623 |
|
1498 |
A |
21 |
left |
left |
chocolate |
0.01 |
0.01 |
725 |
|
1519 |
A |
12 |
left |
left |
replacement page distigme |
0.01 |
0.01 |
801 |
|
1519 |
A |
12 |
left |
left |
replacement page distigme |
0.01 |
0.00 |
801 |
|
1519 |
A |
12 |
left |
right |
replacement page distigme |
0.01 |
0.01 |
801 |
|
1519 |
A |
12 |
left |
right |
replacement page distigme |
0.02 |
0.01 |
801 |
621 C41
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
1492 B5
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
Figure 7
The original ink is sometimes difficult to distinguish from that of the reinker based on color alone. Note the reinked zēta (621 C4) and sigma (1492 B5).
1025 B28–29
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
1046 C2–3
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
1257 A30–31
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
Figure 8
Pairs of unreinked letters at the end of a line (top) replaced by the reinker in the left margin of the following line (bottom)
Another subset that we can compare to the distigmai are unreinked letters that the reinker replaced with supralinear orthographic corrections.79 In such cases, it could be argued that either the original letter or the supralinear correction was reinked, even though that does not appear to us to be the case from examining these letters at ~50x magnification (see Figure 9). We tested 11 such pairs. The copper values of the original letters ranged from 0.21 to 0.35 and that of the supralinear corrections from 0.15 to 0.38 (see Table 4). Neither of these subsets differed substantially from the range of elemental fingerprints of the full set of unreinked and reinked letters that we tested. At the same time, these subsets were totally dissimilar from the elemental profile of both the apricot and chocolate distigmai, which had 0.00–0.02 copper. This near lack of metallic impurities indicates that the ink used to write the distigmai was made from vitriol that was chemically purified. Hence, the distigmai were most likely produced no earlier than the sixteenth century when chemically purifying vitriol became standard.
621 C13
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
906 B33
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
1390 A5
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
Figure 9
Pairs of unreinked letters replaced with a supralinear correction by the reinker at ~50x magnification in visible light (top) and near infrared (bottom). Both the original letters and their supralinear corrections partially lose opacity in near infrared (~940 nm), consistent with iron-gall ink. The presence of iron in both inks was confirmed using XRF.
Table 3
Pairs of unreinked letters at the end of a line replaced by the reinker with the same letter in the left margin of the next line
|
Page |
Col. |
Lines |
Letter |
Word |
Verse |
Unreinked (Cu) |
Reinker next line (Cu) |
|---|---|---|---|---|---|---|---|
|
561 |
A |
26–27 |
sigma |
|
2 Chr 30:25 |
0.27 |
0.24 |
|
1025 |
B |
28–29 |
kappa |
|
Isa 26:13 |
0.24 |
0.20 |
|
1028 |
B |
26–27 |
sigma |
|
Isa 29:11 |
0.22 |
0.35 |
|
1046 |
C |
2–3 |
nu |
|
Isa 46:10 |
0.25 |
0.36 |
|
1241 |
A |
24–25 |
sigma |
|
Mat 6:7 |
0.12 |
0.17 |
|
1257 |
A |
30–31 |
sigma |
|
Mat 16:21 |
0.17 |
0.24 |
|
1270 |
B |
12–13 |
sigma |
|
Mat 25:21 |
0.20 |
0.17 |
|
1390 |
B |
1–2 |
nu |
|
Acts 6:12 |
0.37 |
0.19 |
3.4 Chocolate Distigmai as Reinked Apricot?
Canart discovered what appears to be an apricot distigme “protruding” from the chocolate distigme at 1469 A3 (see Figure 10). This example provided an important piece of evidence that chocolate distigmai were actually reinked apricot distigmai, with the original apricot usually hidden beneath the later chocolate ink. We tested the dots of both the chocolate and apricot distigmai at 1469 A3 and measured the intensity of copper to iron in both as 0.01. Despite the difference in their visual appearance, there was no difference as far as the XRF could detect between this apricot distigme and the adjacent chocolate distigme.
1469 A3
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
1501 B42
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
Figure 10
Apricot distigmai “protruding” from chocolate distigmai. These two examples were cited as evidence that chocolate distigmai were usually reinkings of original apricot distigmai. The XRF tests could not detect any difference between the chocolate and apricot distigmai at 1469 A3.
Table 4
Pairs of letters replaced by the reinker with a supralinear orthographic correction
|
Page |
Col. |
Line |
Letters |
Unreinked (Cu) |
Reinker supralinear correction (Cu) |
|---|---|---|---|---|---|
|
238 |
B |
31 |
nu > gamma |
0.29 |
0.25 |
|
569 |
B |
21 |
omicron > ōmega |
0.27 |
0.21 |
|
569 |
C |
8 |
nu > gamma |
0.26 |
0.21 |
|
621 |
C |
13 |
nu > gamma |
0.29 |
0.15 |
|
690 |
A |
28 |
omicron > ōmega |
0.28 |
0.1880 |
|
906 |
B |
33 |
epsilon > iōta |
0.21 |
0.25 |
|
1012 |
A |
12 |
nu > gamma |
0.26 |
0.30 |
|
1308 |
C |
21 |
thēta > tau |
0.35 |
0.38 |
|
1331 |
B |
31 |
alpha > omicron |
0.25 |
0.19 |
|
1390 |
A |
5 |
omicron > ōmega |
0.3581 |
0.2282 |
|
1454 |
B |
40 |
thēta > mu |
0.28 |
0.23 |
A distigme in the left margin of 1453 A30 has a left dot that appears to be “chocolate” whereas the right dot appears to be “apricot”. The difference in color and thickness between the two dots is obvious in both visible light and transmission (see Figure 11).83 Our tests of 1453 A30 found no difference between the dark dot on the left and the light dot on the right, both of which had a copper to iron ratio of 0.01.
(a)
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
(b)
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
(c)
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
(d)
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
Figure 11
A distigme (1453 A30) with a dark-colored left dot and a light-colored right dot. The different thicknesses of ink are noticeable in visible light (a) and transmission (b). In near infrared (c) the inks nearly disappear, characteristic of iron gall inks. Ultraviolet light (d) enhances the contrast so that the differences in the light and dark dots practically disappear. Note the similarity in color in visible light between the right dot and the original ink of the adjacent kappa. The two dots have the same near lack of metallic impurities.
We examined 158 chocolate distigmai at ~50x magnification but did not find convincing evidence of them being the product of a systematic reinking.84 Most distigmai had a uniform appearance as far as the color of the ink was concerned (see Figure 12).
1280 C10
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
1285 B12
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
Figure 12
Chocolate distigmai considered to be reinked original distigmai. However, no trace of apricot ink is visible even at ~50× magnification. In contrast, the reinking of the adjacent ēta at 1280 C10 and the horizontal line and rhō at 1285 B12 are unmistakable.
To be sure, light-colored ink was visible in some chocolate distigmai alongside the dark ink. However, it is not certain that this was caused by reinking. In the case of 1469 A3, which was previously explained as an original apricot distigme “protruding” from the chocolate distigme of the reinker, it is possible that the two sets of dots were the result of a thin and thick application of ink. The thin part of the stroke could be where the pen first touched the parchment. In the case of 1453 A30, which had a dark-colored left dot and light-colored right dot, the light dot might be a thinner application of the same ink used to produce the dark dot.
An instructive example is a chocolate distigme at 1253 A38, which has light-colored ink sticking out from the left side of each dot (see Figure 13).85 We were able to compare this light-colored ink (Cu 0.01) to the dark ink (Cu 0.01) and found them to have the same minute amounts of metallic impurities. The two colors of ink might have been made by a single stroke of the pen with the lighter ink being a thinner application than the dark. Although it seems unlikely to us, we cannot rule out the possibility that this and other distigmai were reinked (in the 17th century?) with dark chemically purified ink covering the earlier sixteenth-century light-colored chemically purified ink.
Figure 13
A distigme at ~50x magnification in the right margin of 1253 A38 with light-colored ink protruding from the left side of each dot. The light-colored ink and the dark ink both had a copper to iron ratio of 0.01. The white horizontal line at the top of the right dot is a scratch on the Melinex sheet used to protect the surface of the codex from the microscope apparatus. On the right is a detail from the same image with digital zoom.
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
3.5 Distigme-Obelos Pairs
As mentioned above, Payne argued that the original scribe recorded fifteen “distigme-obelos” pairs to mark textual variants of four or more words in the New Testament. He identified one of the horizontal lines (“obelos”) as having the same light-colored ink as the original text whereas the other fourteen he presumed to be reinkings of original lines.86 We were able to confirm that, indeed, some of these horizontal lines were reinked. This is evident at ~50× magnification when the line of the reinking does not perfectly cover the original light-colored horizontal line. For example, Payne argued that the horizontal line in the left margin of 1241 B9 was an obelos marking where the doxology is found at the end of the Lord’s Prayer in other manuscripts but omitted from Codex Vaticanus. Under ~50× magnification, two lines are clearly visible with the reinker’s line only partially covering the original line, which it crosses at a 26° angle (see Figure 14). To the left of these partially overlapping horizontal lines is a distigme identified as apricot.87 We compared this distigme-obelos pair with an unreinked sigma on the same page that was entirely rewritten by the reinker on the next line (1241 A24–25).88 The XRF tests indicate that the original horizontal line (Cu 0.12) was written in the same ink as the original main text (Cu 0.12). The horizontal line was then reinked (Cu 0.17) with the same ink used to reink the main text (Cu 0.20).89 In contrast, the ink of the apricot distigme had a copper value of 0.02.90 Hence, there appear to be three distinct inks among these five scribal features (see Table 5):
1. original (main text + horizontal line);2. reinking (main text + horizontal line);3. apricot distigme.
Although the distigme has what has been identified as “apricot ink”, it is completely different from both the original ink used to write the main text and the original horizontal line as well as the ink used to reink both. The hypothesis that the apricot distigme was written in the same ink as the visually similar original ink of the main text was reasonable but has now been disproven by XRF.
Figure 14
An apricot distigme that Payne argued was part of a distigme-obelos pair (1241 B9 LM). At ~50× magnification, the reinking of the horizontal line and adjacent pi (
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
Table 5
Comparison of original ink and reinker of the main text and a horizontal line with an apricot distigme on page 1241
|
Page |
Col. |
Line |
Margin |
Description |
Cu |
Zn |
Ink |
|---|---|---|---|---|---|---|---|
|
1241 |
A |
24 |
— |
main text (unreinked) |
0.12 |
0.02 |
original |
|
1241 |
B |
9 |
left |
horizontal line (original) |
0.12 |
0.03 |
original |
|
1241 |
A |
25 |
left |
main text (reinker) |
0.17 |
0.05 |
reinker |
|
1241 |
B |
9 |
left |
horizontal line (reinker) |
0.20 |
0.05 |
reinker |
|
1241 |
B |
9 |
left |
apricot distigme |
0.02 |
0.00 |
distigme |
Another distigme-obelos pair was identified at 1243 A12.91 At ~50x magnification we were able to identify some of the original ink in this horizontal line alongside that of the reinker (see Figure 15). Adjacent to the horizontal line is a chocolate distigme (see Figure 16).92 We compared this distigme-obelos pair with an unreinked nu on the same page (1243 C20) and a sigma added by the reinker in the left margin of the verso (1244 A24).93 The XRF tests indicate that the original horizontal line (Cu 0.09–0.10) was written in the same ink as the original main text (Cu 0.10). This original ink was distinct from that of the reinker of the main text on the verso (Cu 0.16–0.17).94 Again, the ink of the chocolate distigme (Cu 0.00–0.01) matched neither that of the original nor that of the reinker. In this case, there appear to be three distinct inks among four scribal features (see Table 6):
The chocolate distigme at 1243 A12 had the same elemental profile as the apricot distigme at 1241 B9 and neither matched the original ink of the adjacent horizontal lines (nor that of the reinker). These distigmai and horizontal lines could not have been written as an intentional pair of conjoining symbols in the same act of production because they were produced with two entirely different inks.
Figure 15
A horizontal line in the left margin of 1243 A12. Signs of reinking are visible on the far left and on the righthand side. The original ink juts out to the right past the reinking. Not enough of the reinking was exposed to successfully test it with XRF. Places where the reinking overlaps the original ink would not have yielded a useful test because there would be no way to distinguish which layer of ink contained the varying intensity of metallic impurities.
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
Figure 16
A distigme-obelos pair (1243 A12 LM) with a chocolate distigme supposedly covering an original apricot distigme. However, at ~50× magnification, the distigme exhibits no visible signs of an earlier layer of ink. Note the signs of reinking of the epsilon (
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
Table 6
Comparison of original ink and reinker of the main text and a horizontal line with a chocolate distigme on pages 1243–1244
|
Page |
Col. |
Line |
Margin |
Description |
Cu |
Zn |
Ink |
|---|---|---|---|---|---|---|---|
|
1243 |
C |
20 |
– |
main text (unreinked) |
0.10 |
0.01 |
original |
|
1243 |
A |
12 |
LM |
horizontal line (original) |
0.10 |
0.03 |
original |
|
1243 |
A |
12 |
LM |
horizontal line (original) |
0.09 |
0.02 |
original |
|
1244 |
A |
24 |
LM |
main text (reinker) |
0.16 |
0.03 |
reinker |
|
1244 |
A |
24 |
LM |
main text (reinker) |
0.17 |
0.04 |
reinker |
|
1243 |
A |
12 |
LM |
chocolate distigme |
0.00 |
0.00 |
distigme |
|
1243 |
A |
12 |
LM |
chocolate distigme |
0.01 |
0.00 |
distigme |
3.6 Distigmai Compared to Other Scribal Features
The elemental profile of the distigmai with its near lack of copper impurities was unusual among our tests in Codex Vaticanus, with some notable exceptions. The Arabic numerals of the Vulgate chapters in the New Testament, which according to Versace were added in the 16th century, had the same elemental profile as the distigmai as far as low amounts of metallic impurities. As mentioned above, Versace argued that these Arabic numerals were inscribed by the same hand that recorded the distigmai. He based this on two things: 1. The color of the distigmai and Arabic numerals, which he argued were the same, and 2. dots associated with the Arabic numerals that he argued had the same “form” as the distigmai.96 We carried out 12 tests on 5 of the Arabic numeral chapter notations97 and none of them had any significant metallic impurities. This included one test of an Arabic numeral on a fifteenth-century replacement page, which continued the numbering of the chapters from the original pages. The intensity of copper to iron in all these Arabic numerals and their associated symbols was measured between 0.00 and 0.03 (see Table 7). Like the distigmai, the ink of the Arabic numerals also appears to have been chemically purified.
1447 B8 LM
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
1448 B6 LM
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
1449 B3 LM
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
Figure 17
Reversed pilcrows of the Roman numeral chapter system covering earlier Arabic numeral chapter numbers. The Roman numeral system was added 1571–1600, which means the Arabic numeral chapters could not be later than 1600.
While the results of our XRF tests might seem to confirm Versace’s position that the sixteenth-century Arabic numerals were added by the same hand as the distigmai, we should offer some caution. When two inks have the same elemental fingerprint, with the same proportions of metallic impurities, it suggests they were likely written using the same ink. We have already mentioned how surprising it is that the reinker’s ink had the same elemental fingerprint as the original ink of Codex Vaticanus. In any event, when two inks have only trace amounts of metallic impurities all that can be said is that they have not been proven to be different inks. There is no positive indication from the XRF evidence that they are the same ink. Indeed, this could be said about the distigmai as a group, both apricot and chocolate. Arguments for distigmai being the work of a single scribal project must be based on their form and function. XRF is consistent with this but does not prove it.
Another scribal feature with the same elemental profile as both the apricot and chocolate distigmai is the system of Roman numeral chapter numbers recorded in the margins (Versace’s B32).98 This system includes the word “Cap.[ut]” (chapter) before each Roman numeral and a reversed pilcrow (⁋) marking the line on which the chapter begins. Recently Giacomo Cardinali identified the scribe who wrote the Roman numeral chapter system in Codex Vaticanus as Pedro Chacón of Toledo who arrived in Rome in 1571 or 1572 and died there in 1581. Some of the chapter numbers, however, may have also been added by Fulvio Orsini (1529–1600).99 This gives a narrow timeframe for the Roman numeral chapter numbers as 1571–1600.
The Roman numeral chapter system primarily runs through the Old Testament whereas the Arabic numeral system runs through the New Testament. However, there are some fourteen places in the New Testament where the chapter is marked with Roman numerals.100 In three places, the reversed pilcrow of the Roman numeral system covers the Arabic numeral proving that the Roman numeral system is later than the Arabic (see Figure 17). Since the Roman numerals were recorded 1571–1600, the Arabic numerals could not be later than the 16th century.
An example of the dual numbering system can be found on page 1308 where there is an Arabic numeral chapter number in the left margin and a Roman numeral chapter number in the right, both marking the beginning of Luke chapter 3 (1308 C30–31). Although both numerals lack copper, their visual appearance suggests they were written in different inks using different writing implements. At the same time, the similarity in visual appearance between a chocolate distigme on page 1308 and the dots flanking either side of the Arabic numeral is striking (see Figure 18).
chocolate distigme 1308 A11 LM
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
Arabic numeral chapter number 1308 C31 LM
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
Roman numeral chapter number 1308 C30 RM
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
Figure 18
A chocolate distigme (1308 A11) on the same page as Arabic (1308 C31) and Roman (1308 C30) numeral chapter numbers marking the beginning of Luke chapter 3. The chocolate distigme bears a striking visual resemblance to the dots flanking either side of the Arabic numeral “3”. In contrast, the visual appearance of the Arabic and Roman numerals suggest they were written in different inks using different writing implements. The chocolate distigme (Cu 0.00), Arabic numeral chapter number (Cu 0.01), and Roman numeral chapter number (Cu 0.02) were all written with inks that were made from chemically cleaned vitriol.
Figure 19
An unreinked thēta of the original scribe replaced with a supralinear tau by the reinker (1308 C21) compared to an apricot distigme on the same page (1308 B27). The copper to iron ratios of the unreinked thēta (Cu 0.35) and the reinker’s tau (Cu 0.38) prove they were not written with the same ink as the apricot distigme (Cu 0.00).
An apricot distigme also on page 1308 can be compared to an unreinked thēta on the same page (see Figure 19). The reinker changed the orthography of “Nazareth” in Luke 2:51 by leaving the original scribe’s thēta unreinked and replacing it with a supralinear tau (
Codex Vaticanus contains a Table of Contents on flyleaf III.101 Like the Roman numerals, Cardinali also identified the handwriting of the first column of the Table of Contents as being that of Chacón, which means it was written in Rome sometime 1571–1581.102 The Table of Contents was written in an ink that is visually distinct from that of the distigmai, with what seems to have been a different writing implement (see Figure 20). Despite this, the ink of the Table of Contents has the same near lack of metallic impurities as the distigmai with a copper to iron ratio of 0.01 (see Table 7).
apricot distigme 1276 C31 LM
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
chocolate distigme 1279 B20 LM
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
Table of Contents flyleaf III A16
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
Figure 20
An apricot distigme (1276 C31) and a chocolate distigme (1279 B20) compared to the Table of Contents (flyleaf III A16). All three have the same elemental profile characterized by a near lack of metallic impurities (Cu 0.01, 0.00, and 0.01, respectively).
An interesting contrast to the Table of Contents on flyleaf III is the canon of biblical books according to Jerome on the recto of the same leaf (see Figure 21).103 Daniele Bianconi recently argued convincingly that Cardinal Bessarion (1408–1472) was the scribe who added the Jeromian Canon to Codex Vaticanus.104 The ink of Bessarion’s Jeromian Canon has a copper to iron ratio of 1.00–1.18 meaning that it has as much, or more, copper as iron (see Table 7). Theoretically, Bessarion could have added the Jeromian Canon in Byzantium but it is more likely he did so after the manuscript came into his custody in Italy.105 The ink of the main text on the 15th century minuscule replacement pages has a copper to iron ratio of 0.49–0.58 (see Table 7). Thus, the fifteenth-century Jeromian Canon and the fifteenth-century minuscule replacement pages were written in inks that were not chemically purified. In contrast, the sixteenth-century Arabic numeral chapter numbers, the Table of Contents (1571–1581), and the Roman numeral chapter numbers (1571–1600) were written in inks that were chemically purified.106 The distigmai distinctly belong to the chemically purified inks.
As already mentioned, Niccum pointed to distigmai on one of the minuscule replacement pages (1519) as evidence that all distigmai were added in the 15th century or later.107 Payne’s counterargument was that the distigmai on the replacement pages could have been copied from a deteriorated original page when the minuscule replacement pages were first produced.108 However, the XRF tests prove that the ink of a distigme on one of the minuscule replacement pages (Cu 0.01–0.02) is completely different from the main text on the replacement pages (Cu 0.49–0.58). The replacement pages were written with ink that was not chemically purified whereas all the distigmai, including one on the replacement pages, were written with ink that was chemically purified. These two different types of ink—unpurified vs. purified—might reflect a transition from making inks out of vitriol that was not chemically purified in the 15th century to vitriol that was purified in the 16th century.
Jeromian Canon flyleaf IV 12
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
replacement page main text 1519 A4
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
replacement page distigme 1519 A12 LM
Citation: The Vatican Library Review 3, 2 (2024) ; 10.1163/27728641-00302005
Figure 21
The Jeromian Canon (flyleaf IV 12) compared to the main text on a minuscule replacement page (1519 A4) and a distigme also on a minuscule replacement page (1519 A12 LM). The Jeromian Canon has more copper than iron, the replacement page has more copper than any of the original inks or the reinking but only half that of the Jeromian Canon, and the distigme is devoid of copper (Cu 1.12, 0.58, and 0.00–0.01, respectively). The distigme on the replacement page was not written in the same ink as the main text on the replacement page.
Table 7
Results of XRF tests of 15th and 16th centuries scribal features
|
Page |
Col. |
Line |
Description |
Presumed Date |
Cu |
Zn |
|---|---|---|---|---|---|---|
|
IV |
– |
10 |
Jeromian canon |
15th century |
1.00 |
0.19 |
|
IV |
– |
12 |
Jeromian canon |
15th century |
1.12 |
0.22 |
|
IV |
– |
25 |
Jeromian canon |
15th century |
1.00 |
0.21 |
|
1519 |
A |
6 |
replacement pages main text |
15th century |
0.57 |
0.03 |
|
1519 |
A |
4 |
replacement pages main text |
15th century |
0.58 |
0.03 |
|
1521 |
A |
36 |
replacement pages main text |
15th century |
0.49 |
0.03 |
|
1521 |
A |
36 |
replacement pages main text |
15th century |
0.57 |
0.03 |
|
1255 |
A |
9 |
Arabic numerals |
16th century |
0.01 |
0.00 |
|
1279 |
A |
41 |
Arabic numerals |
16th century |
0.01 |
0.00 |
|
1279 |
A |
41 |
Arabic numerals |
16th century |
0.01 |
0.00 |
|
1279 |
A |
41 |
Arabic numerals |
16th century |
0.03 |
0.00 |
|
1308 |
C |
31 |
Arabic numerals |
16th century |
0.01 |
0.00 |
|
1308 |
C |
30 |
Arabic numerals |
16th century |
0.01 |
0.00 |
|
1308 |
C |
31 |
Arabic numerals |
16th century |
0.01 |
0.00 |
|
1399 |
B |
17 |
Arabic numerals |
16th century |
0.01 |
0.00 |
|
1399 |
B |
17 |
Arabic numerals |
16th century |
0.01 |
0.00 |
|
1399 |
B |
17 |
Arabic numerals |
16th century |
0.01 |
0.00 |
|
1399 |
B |
17 |
Arabic numerals |
16th century |
0.01 |
0.00 |
|
1519 |
B |
8 |
Arabic numerals (replacement pages) |
16th century |
0.02 |
0.01 |
|
III |
A |
1 |
Table of Contents |
1571–1581 |
0.01 |
0.00 |
|
III |
A |
16 |
Table of Contents |
1571–1581 |
0.01 |
0.00 |
|
III |
A |
22 |
Table of Contents |
1571–1581 |
0.01 |
0.00 |
|
1304 |
A |
17 |
Roman numerals |
1571–1600 |
0.00 |
0.00 |
|
1304 |
A |
17 |
Roman numerals |
1571–1600 |
0.01 |
0.00 |
|
1308 |
C |
30 |
Roman numerals |
1571–1600 |
0.02 |
0.01 |
4 Conclusions
This study builds on the pioneering work of Philip Payne who discovered the distigmai in Codex Vaticanus nearly thirty years ago and recognized their text-critical function. It also continues the legacy of Paul Canart who carried out palaeographical observations of Codex Vaticanus at the end of the last century using innovative magnification techniques. Some twenty years ago, Patrick Andrist warned about relying solely on the color of the distigmai to prove their antiquity, as “a visual resemblance is no guarantee of identity. A spectrometric analysis of the ink would perhaps clarify this question.”109 Some fifteen years ago, Payne himself noted a similar caution when he wrote, “confirmation awaits scientific testing, such as multi-spectral imaging or X-ray Fluorescence imaging.”110
Micro-X-ray fluorescence spectroscopy has now proven that all the distigmai tested, both apricot and chocolate, were written in chemically purified ink(s) typical of the 16th century or later. Despite appearances, the apricot distigmai tested were not written in the same ink as the original main text or original horizontal lines (“paragraphoi”). Any similarity in color between the ink of the apricot distigmai and the original ink is purely coincidental and entirely superficial and therefore can no longer be cited as evidence for an early date of the distigmai. Similarly, the chocolate distigmai and the reinking of the main text utilized radically different inks and hence any similarity in appearance is also pure coincidence.
The strongest argument for a late date was always the existence of distigmai on the 15th century minuscule replacement pages. The counterargument was that these distigmai were copied from a damaged original page as part of the original production of the replacement pages. However, the ink of a distigme tested on one of the replacement pages is identical to that of all the other distigmai, apricot and chocolate, and entirely dissimilar from that of the main text on the replacement pages. Hence, the distigmai on the replacement pages are not part of the original production of these pages but rather a paratextual element added later along with all the other distigmai.
Finally, this study should serve as a warning about reaching definitive palaeographical or codicological conclusions based on visual appraisal of the color of ink alone.
Acknowledgments
Our sincere thanks to the Biblioteca Apostolica Vaticana and the generous cooperation of its staff without whom we could not have done this research, especially Don Mauro Mantovani, Msgr. Cesare Pasini, Claudia Montuschi, Irmgard Schuler, and Angela Núñez Gaitan. We also wish to thank the members of the library’s Laboratorio di restauro (Maria Rosaria Castelletti, Marta Grimaccia, Silvia Foschetti, Maruska Di Remigio, Oscar Cocciolo, Mario Tiburzi, Salvatore Giglio, Simone Olini) and Sala Manoscritti (Antonio Schiavi, Andrea Zucchi, Lidia Laudenzi, Lucrezia Dell’Orco) for their tireless assistance and cooperation. Our thanks to the editor of this journal András Németh for his assistance and encouragement in carrying out the research. We also wish to thank Philip Payne and Jesse Grenz for their invaluable assistance. This research was carried out with partial funding from the Institute for Hebrew Bible Manuscript Research (
Authors contributions: Nehemia Gordon defined the corpus of the study and formulated the questions of the study, conducted imaging of the codex using Dino-Lite microscopy, chose the spots to be checked with XRF, and wrote the first draft of the manuscript. Oliver Hahn conducted some of the XRF measurements and reviewed the manuscript. Nelson Calvillo carried out quality control of the data and reviewed the manuscript. Ira Rabin carried out XRF measurements, processed the results of the XRF measurement data, wrote the sections on Microscopic Reflectography, Micro-X-ray Fluorescence, and XRF Raw Spectra, and reviewed the manuscript. Patrick Andrist helped define the corpus of the study, define some research questions, and reviewed the manuscript in depth. Pavlos Vasileiadis helped define the research questions, contributed some philological observations, and reviewed the manuscript. Two lectures with some preliminary findings from this study were presented by Nehemia Gordon and Ira Rabin, respectively, at the Society of Biblical Literature Annual meeting in San Antonio on November 21, 2023.111
Città del Vaticano, BAV, Vat. gr. 1209; diktyon 67840 (
The replacement pages are 1–40 (Gen 1:1–46:28), 695–704 (Ps 105:27–137:6), and 1519–1536 (Heb 9:15–13:25 and Revelation). On the palaeography and date of the replacement pages, see Acerbi & Bianconi 2022, 33–89. Philemon, 1 Timothy, 2 Timothy, and Titus are not extant even among the later pages, but it is probable that they used to be part of the codex, see Andrist 2015, 12; Andrist 2021, 18.
Grenz 2021, 32; Kenyon 1912, 79–80; Metzger 1991, 74; Milne & Skeat 1938, 87.
Cavallo 1967, 52–56; Grenz 2021, 261; Kenyon 1912, 80; Metzger 1991, 74; Milne & Skeat 1938, 89; O’Neill 1989, 220; Ropes 1926, xxxviii; Skeat 1999, 603–604; Tischendorf 1867a, xxi n. 2; Tischendorf 1867b, ix; Traube 1907, 54, 66–67.
Grenz 2021, 32, 119.
Versace 2018, 43–50, 67–68, 259–268; cf. Grenz 2021, 32. Versace referred to the 10th–11th century ochre-reinker as B18 and the 16th century dark-reinker as B37.
Payne counted 765 distigmai, Miller 778, Willker 801, Gravely 812, and Versace 858 (Gravely 2009, 43; Miller 2000, 40 n. 16; Miller 2003, 219; Payne & Canart 2000, 106; Versace 2018, 298–304; Willker 2001). One of the reasons for the range of numbers is that some distigmai might actually be ink transfer from the opposite page of the codex or stray ink.
Payne already mentions this dichotomy in 1995 based on a small sampling of distigmai he examined in the 1965 facsimile edition of Codex Vaticanus but without reference to the terms “apricot” and “chocolate” (Payne 1995, 251).
1469 A3; 1501 B42 (Payne 2010a, 10; Payne & Canart 2009, 214). Originally, Canart observed apricot ink protruding from the chocolate distigme at 1474 A20, but later determined that there was no apricot ink there (Payne & Canart 2000, 110; Payne & Canart 2009, 215 n. 96).
Payne 2010a, 9; Payne 2023, 187; Payne & Canart 2009, 215 n. 97. Also in Payne 2022, 26, 57, 64, an unpublished paper, which we thank the author for sharing with us and giving us permission to cite.
Payne & Canart 2000, 109–110.
Payne & Canart 2009, 215–216.
Miller 2003, 235. Similarly, “What is clear is that the umlauts mark a line of text associated with a variant […] There is also always the possibility that the umlaut is marking a presently unknown variant […]” (Gravely 2009, 56).
Payne & Canart 2000, 113.
Bibliorum sacrorum Graecorum Codex Vaticanus B: Bibliothecae Apostolicae Vaticanae Codex Vaticanus Graecus 1209. Roma: Istituto Poligrafico e Zecca dello Stato, 1999.
Payne & Canart 2009, 203–208, 214–215.
Payne 2022, 63; Payne 2023, 185.
There is some confusion about the precise number of apricot distigmai. Payne and Canart counted the distigmai at 1380 A26 and 1381 C26, one of which may be mirror ink transfer of the other, as a single distigme, so their list of fifty-one actually includes fifty-two locations. The two instances of apricot protruding from chocolate at 1469 A3 and 1501 B42 could be counted as fifty-three and fifty-four. A fifty-fifth example is 1409 B25 where “the left dot appears to be reinforced but the right dot is not reinforced” although this is only classified as “probable” apricot (Payne & Canart 2009, 214). In his 2023 book, Payne writes that “Fifty-one distigmai match the apricot color of original Vaticanus ink” (Payne 2023, 95) but then mentions another apricot distigme (1241 B9) in the same book not included in the fifty-one, which could be counted as a fifty-sixth example (Payne 2023, 185). In total, Payne and Canart (together or separately) identified fifty-six apricot, partial apricot, mirror apricot, or probable apricot distigmai in their published works. However, the apricot distigme at 1243 B21 is actually an unreinked sigma, lowering the number back down to fifty-five. On 1453 A30, not included in these numbers, see note 77 below.
Paragraphoi mark the line on which a section ends, not the line on which the new section begins, although in practice they are usually the same line (Grenz 2018, 10). In the example shown in Figure 1, if the horizontal line is a paragraphos, it marks the end of a section with the final word in 1 Cor 14:33,
Swete 1900, 364 explains the abbreviation
Payne 2017, 605; Payne 2019, 27; Payne 2022, 6; Payne & Canart 2009, 200.
Payne 2017, 605, 614; Payne 2022, 3, 7, 29–31, 36, 45, 58, 63–65; Payne 2023, 89, 95, 185–187.
“tant le phénomène lui-même que les explications qu’on en donne appellent encore un supplément d’enquête: un examen attentif de toutes les distigmai et leur confrontation avec tous les lieux variants attestés par la tradition pourraient confirmer ou nuancer les statistiques et les conclusions de Payne” (Canart 2009, 224–225).
Amphoux 2007.
There is a distigme in the left margin of 1519 A12 and possibly another at 1519 B12. Both of these instances were cited as evidence of the lateness of the distigmai by Head 2009. Our thanks to Peter Head for sharing his unpublished notes and PowerPoint. Head’s lecture is also summarized in Knust & Wasserman 2018, 127 n. 99; Payne 2010b; Wasserman 2009a; Wasserman 2009b.
Niccum’s main argument was that no other paratextual features, such as paragraphoi, were reproduced in the minuscule replacement pages from the original pages so it was unlikely that the distigmai on page 1519 were. For a response, see Payne & Canart 2000, 109 n. 25. See further below.
Bordat 2007, 1343; Hernández 2011, 80–81.
Krans 2020. For distigmai-like marginal symbols in manuscripts used by Erasmus and his team for their 1516 edition of the New Testament, see Andrist 2018, 157–159.
Head 2009; Knust & Wasserman 2018, 127 n. 99; Niccum 1997; Payne 2010b; cf. Wasserman 2009b. Perhaps Sepúlveda chose 365 as a typological number with one variant for each day of the solar year.
Gurry 2022; Snapp, Jr. 2022. Against Snapp’s hypothesis, Andrew J. Brown pointed out in a comment to Gurry 2022 that Erasmus himself referred to 300 readings in Codex Vaticanus, not 700. On the exchange between Sepúlveda and Erasmus concerning Codex Vaticanus, see Yi 2024, 17–25.
“La mia attività in Biblioteca Vaticana è stata seguita dagli scriptores […] mi offrì, in varie occasioni, l’opportunittà di esaminare i fogli originali del Codice B” (Versace 2018, 5).
Versace 2018, 68.
Payne 2010a, 13 n. 32; Payne & Canart 2009, 204.
Versace 2018, 299 n. 32.
Versace 2018, 69.
Versace 2018, 70.
Andrist 2009; Epp 2005, 18–19; Fellows 2019; Kloha 2006, 2:499–516; Krans 2019; Miller 2000; Miller 2003. Payne meticulously answered his critics in Payne 2004; Payne 2010b; Payne 2010a; Payne 2019; Payne 2021.
Rabin et al. 2012.
Hahn et al. 2019; Rabin 2021.
Hahn et al. 2019.
Hahn et al. 2004; Rabin et al. 2012.
Díaz Hidalgo et al. 2023, 2; Hahn et al. 2019; Krekel 1999, 54–55.
Forbes 1993, 236–239; LSJ s.v. “
Philo’s recipe does not explicitly state that the ink turned black and the exact meaning of
LSJ s.v. “
Delange et al. 1990.
Rabin et al. 2019.
Delange et al. 1990; Ghigo et al. 2020; Nehring, Bonnerot, Gerhardt, et al. 2021.
Krekel 1999, 55–56; Zerdoun Bat-Yehouda 1983, 238.
Medlej 2020, 24–25; Rabin 2021, 77; Zerdoun Bat-Yehouda 1983, 356. On a possible difference between
Aceto et al. 2008; Ghigo et al. 2020; Rabin & Krutzsch 2019.
Inks and pigments in Codex Vaticanus that were not iron-gall ink will be discussed elsewhere.
Dioscorides Pedianus of Anazarbus, De materia medica 5.98 (= c. 114 in some editions), ed. Wellmann 1914, 68–69 and trans. Beck 2005, 376; Karpenko & Norris 2002, 998; Pliny, Nat. 34.32.123–127; Stillman 1960, 43–44.
Karpenko & Norris 2002; cf. Golas 1999, 370–372.
Beckmann et al. 1846, 184; Karpenko & Norris 2002, 1001; Rabin 2021, 73; Zerdoun Bat-Yehouda 1983, 356.
Hickel 1963; Kraschewski 2001; Vuillard et al. 2024.
Peterson 2003; Vuillard et al. 2024.
Gordon et al. 2020; Nehring, Bonnerot, Gordon, et al. 2021; Nehring et al. 2022; Rabin 2021; Rabin et al. 2024.
The hydrometallurgical cementation process using solid iron to precipitate copper out of aqueous solutions should not be confused with the brass cementation process. Cf. Bourgarit & Bauchau 2010; Free 2013, 232–235; Habashi 1993, 3–4, 485, 624–639; Habashi 2005, 15; Han 2009, 100–101; Kirnbauer 1966, 53; Lung 1986. Pliny describes a chemical reaction that some interpret to be cementation but it was probably rust, see Jost 2014, 248–249; Karpenko 1995.
Karpenko & Norris 2002, 1001.
Golas 1999, 370–386; Lung 1986; Ts’ao et al. 1959, 144–145.
Lacko & Mayerová 2016, 37, 93–94. On the possible use of the cementation process in Central Europe as early as the 13th century, see Hronček et al. 2021, 181; Kirnbauer 1966, 53; Neumann 1904, 87; Wenzel 1880, 79. On the possible use of the cementation process in medieval Spain, see Bromehead 1956, 11; Hassan & Hill 1986, 249; Jost 2014, 264–267; Read 1937, 250.
Karpenko & Norris 2002, 1001; Waite 1976, 104.
Biringuccio 1990, 95, states about vitriol: “nor is it a thing from which metal is extracted in any manner, although it shows in certain ways that it is not entirely without metal since it produces a certain trace of it. […] it seems to me that there are in it […] the properties of copper and of iron.”
Birch 1772, 4:329–340; Ts’ao et al. 1959, 145.
Quantitative analysis would involve translating the signals into actual quantities, whereas semi-quantitative analysis involves translating them into relative quantities.
These statistics count the apricot distigme and the chocolate distigme from which it supposedly protrudes at 1469 A3 as two separate distigmai, even though they may be thin and thick applications of ink produced by the same stroke of the pen. Overall, we captured 2,612 images on May 3–5, 2022 and 1,118 images on February 13–17, 2023 at ~50x magnification in visible, ultraviolet (~395 nm), near infrared (~940 nm), and transmission using the Dino-Lite USB microscope. We also carried out 258 XRF tests on September 5–16, 2022 and 329 XRF tests on February 13–17, 2023 using the ARTAX spectrometer.
The identification of the distigmai at 1276 C31 LM; 1279 B1 LM; 1279 C41 RM; 1285 C14 RM; 1308 B27 LM; 1332 B10 LM; 1332 B15 LM; 1332 C20 LM as “apricot” was based on Table 10 in Payne & Canart 2009, 204–208. Canart identified 1469 A3 LM as having apricot protruding from a chocolate distigme (Payne & Canart 2009, 214–215). 1241 B9 is identified as apricot in Payne 2022, 30; Payne 2023, 185. We presumed any distigmai not identified as apricot to be chocolate. On 1390 A32 as a later addition, see Payne 2017, 614; Payne 2022, 26; Payne 2023, 187. Miller and Versace both omitted 1241 B9 from their list of distigmai but it was included by Willker and Gravely. Miller omitted 1279 B26, 1338 C33, and 1453 A30 from his list of distigmai, but they were included by Willker, Gravely, and Versace (Gravely 2009, 102; Versace 2018, 298–304; Willker 2001). On 1453 A30, see note 77 below. Versace identified 1244 C40; 1255 A39; 1276 A18 as reinked although he only noted reinking in the distigmai “a titolo di esempio, un certo numero di casi” (Versace 2018, 298 n. 17). On 1253 A38 see note 79 below. Versace also argued that only one of the dots in 1255 C1 was reinked (Versace 2018, 299 n. 23) but we found that the copper to iron ratio of both dots was 0.01. We were not able to identify signs of reinking at ~50x magnification in the distigmai at 1243 A12; 1243 C40; 1255 A39; 1255 B3; 1255 C1; 1276 A18; 1279 A22; 1279 A27; 1279 B20; 1279 B26; 1285 B12; 1308 A11; 1338 C33; 1390 A32; 1444 C8; 1444 C21. It is possible that the distigmai at 1244 C40; 1253 A38; 1453 A30; 1469 A3; 1498 A21; 1519 A12 were reinked but the lighter ink visible at magnification might also be explained as a lighter application of the same ink. The distigme on the fifteenth-century replacement page that we tested is 1519 A12.
Chocolate: 1243 A12, 1243 C40, 1285 B12, 1474 A20; apricot: 1241 B9, 1279 C41, 1332 C20 (Payne 2017, 613, 615; Payne 2022, 63–64; Payne 2023, 89, 185–187).
Our thanks to Philip Payne for suggesting a list of 56 distigmai to test. We managed to test 11 apricot, or possibly apricot, distigmai out of the that list, i.e., 20 % of the corpus (cf. note 19 above). Unfortunately, we were unable to test all the recommended distigmai because of time limitations and access to only a limited number of bifolios. The apricot distigme at 1243 B21 turned out to be a sigma, although we tested it anyway and found it to have the same copper to iron ratio as the second unreinked nu at 1243 C20 (Cu 0.10 and 0.10, respectively). This sigma was omitted from the list of distigmai by Willker, Gravely, and Versace and was typeset as a sigma in Vercellone & Cozza-Luzi 1868.
Geissbühler et al. 2018; Hahn et al. 2007; Rabin et al. 2019; Rabin et al. 2024.
The numbers in this column refer to the “Masterlist of umlauts” published in Willker 2001.
Grenz 2021, 32.
Versace 2018, 43–50, 67–68, 259–268.
Cf. Grenz 2021, 126.
Versace 2018, 45.
On 1255 C1, see note 68 above.
On 1453 A30, see note 77 below.
Cf. Grenz 2021, 138–140.
Average of two scans.
Average of two scans.
Average of two scans.
The right dot of 1453 A30 was not identified as apricot in any published studies. However, it was referenced in Payne 2022, 45. The context was a list of “Seven Reasons to Believe that Scribe B Penned Distigme-Obelos Symbols”: “Fifth, the original-apricot-color ink in three of these distigmai and in part of one obelos supports assigning these four distigme-obelos symbols to Vaticanus’s original production. These associate both distigme-obelos components with Scribe B. Canart further confirmed the association of distigmai with scribe B by identifying fifty-one distigmai that match the apricot color of the original ink of Vaticanus on the same page, two in distigme-obelos symbols, protrusion of apricot-color ink from under some reinked distigmai, and at least one distigme with one apricot-color dot and one matching the color of the reinking of Vaticanus.” Cf. Gravely 2009, 102; Versace 2018, 302; Willker 2001.
We examined the following distigmai, which we presumed to be chocolate (since they were not identified as apricot in any published studies), at ~50× magnification: 1241 A7 RM; 1241 A36 LM; 1241 C7 RM; 1241 C31 RM; 1242 C31 LM; 1243 A12 LM; 1243 C40 RM; 1244 A22 LM; 1244 B40 LM; 1244 C40 LM; 1253 A38 RM; 1253 B13 LM; 1253 B39 LM; 1254 B18 LM; 1255 A31 LM; 1255 A39 LM; 1255 B3 LM; 1255 B23 LM; 1255 B32 LM; 1255 C1 RM; 1256 C31 LM; 1257 C7 RM; 1269 B4 LM; 1269 B18 LM; 1270 A18 LM; 1270 C32 LM; 1273 B4 LM; 1273 B41 RM; 1274 A21 LM; 1275 B10 LM; 1275 B16 LM; 1275 B31 LM; 1276 A18 LM; 1279 A22 LM; 1279 A27 LM; 1279 B20 LM; 1279 B26 LM; 1280 B7 LM; 1280 B20 LM; 1280 C10 LM; 1285 A14 LM; 1285 B12 LM; 1286 A37 LM; 1307 B4 LM; 1307 B9 LM; 1307 C5 RM; 1307 C11 RM; 1307 C24 RM; 1308 A11 LM; 1321 A22 LM; 1322 A9 LM; 1322 B15 RM; 1322 C20 LM; 1331 B27 LM; 1337 A18 RM; 1337 A24 RM; 1337 A39 RM; 1337 C10 RM; 1338 C33 LM; 1389 A20 RM; 1389 B12 LM; 1389 B30 LM; 1389 C28 RM; 1390 A32 LM; 1390 B6 LM; 1390 C21 LM; 1397 A15 LM; 1397 B25 LM; 1397 B39 LM; 1397 C13 RM; 1397 C20 RM; 1397 C23 RM; 1398 B17 LM; 1398 C39 LM; 1399 A30 RM; 1399 B28 LM; 1400 B20 LM; 1400 B30 LM; 1426 C11 LM; 1426 C32 LM; 1443 A20 LM; 1443 C3 RM; 1443 C24 RM; 1443 C28 RM; 1444 B24 RM; 1444 C8 LM; 1444 C21 LM; 1444 C22 LM; 1444 C25 LM; 1453 A15 LM; 1453 A30 LM; 1453 B4 LM; 1453 B17 LM; 1453 B36 LM; 1453 C14 RM; 1453 C39 LM; 1454 A39 LM; 1454 C25 LM; 1454 C27 LM; 1454 C34 LM; 1455 A6 LM; 1455 B8 LM; 1455 B31 LM; 1455 C12 RM; 1455 C18 RM; 1456 A18 LM; 1456 A25 LM; 1456 B24 RM; 1456 C18 LM; 1456 C21 LM; 1456 C32 LM; 1469 B12 LM; 1469 C17 RM; 1470 A1 LM; 1470 A18 LM; 1470 B37 LM; 1470 C8 LM; 1473 A17 LM; 1473 B2 LM; 1473 B24 LM; 1473 B34 LM; 1474 A20 LM; 1474 B23 LM; 1474 C37 RM; 1475 A18 LM; 1475 B3 LM; 1475 C13 RM; 1476 B1 LM; 1476 B31 LM; 1476 C1 LM; 1476 C14 LM; 1476 C35 LM; 1479 A2 LM; 1479 B39 LM; 1480 A36 LM; 1487 A24 LM; 1487 B35 LM; 1488 B27 LM; 1491 A5 LM; 1491 A29 LM; 1491 A33 LM; 1491 B3 LM; 1491 C17 RM; 1492 A5 LM; 1492 C3 LM; 1492 C41 LM; 1497 A28 LM; 1497 B19 LM; 1497 C6 RM; 1498 A3 LM; 1498 A21 LM; 1498 B5 RM; 1498 C3 LM; 1502 A21 LM; 1502 A25 LM; 1502 B42 LM; 1502 C8 LM; 1502 C33 LM. We also examined the following distigmai at ~50× magnification: 1469 A3 LM and 1501 B42 LM, which Canart identified as having apricot distigmai protruding from the chocolate; 1479 A12 LM, which Payne and Canart argued was added by the reinker without an earlier apricot distigme underneath it; 1519 A12 LM, a distigme on one of the fifteenth-century replacement pages; 1276 C31 LM; 1279 B1 LM; 1279 C41 RM; 1285 C14 RM; 1308 B27 LM; 1332 B10 LM; 1332 B15 LM; 1332 C20 LM; 1473 A6 LM; 1475 B11 LM; 1501 A32 LM, which Payne and Canart considered apricot; and 1241 B9 LM, which Payne considered to be apricot, but neither he nor Canart examined directly.
Willker excluded 1253 A38 RM from his list of distigmai but Versace included it, noting that it was reinked. Gravely noted, “it is irregularly placed (to the right of an A column). But the umlaut is distinct and clearly retraced” (Gravely 2009, 107; Versace 2018, 299).
Payne 2017, 605, 614; Payne 2022, 3, 7, 29–31, 36, 45, 58, 63–65; Payne 2023, 89, 95, 185–187.
“Matthew 6:13 1241B Note the original apricot color ink of the distigme” (Payne 2022, 63). “Matthew 6:13 1241B The distigme matches Vaticanus’s original apricot color ink.” (Payne 2023, 185).
This may have been because the original scribe and the reinker had different ideas about the syllabification of the word (
Given the 10 % margin of error, 0.17 (0.153–0.187) is effectively the same elemental fingerprint as 0.20 (0.180–0.220).
Our test of the apricot distigme at 1241 B9 LM involved passing a beam along a line through the left dot and the ink smear that appears to have originated from the two dots. We found this ink mass to be homogenous with no significant amount of metallic impurities in either the dot or the associated smeared ink. Although it is generally problematic to carry out a test on what may be two layers of ink (left dot and ink smear), the results in this case indicate that neither layer had significant metallic impurities. This near lack of metallic impurities in the apricot distigme was in sharp contrast to the original ink (of the main text and the horizontal line) and the reinking (of the main text and the horizontal line).
Payne 2022, 30–31. Similarly, “[t]he left portion of the obelos matches Vaticanus’s original apricot color ink” (Payne 2023, 185).
Versace noted that this distigme was the same color as a nearby Arabic numeral (Versace 2018, 299 n. 20).
The reinker’s sigma was added in place of an unreinked sigma in the word
We attempted to test the reinking of the horizontal line but were not able to isolate that ink successfully from the original ink.
1241 and 1243 are in the same quire but not the same bifolio.
Versace 2018, 68–70.
We carried out 5 tests on the Arabic numerals themselves, 6 on the dots associated with these numerals, and 1 test on a chapter symbol in the text that indicates where the Vulgate chapter begins.
Versace 2018, 63–66 only mentions these Roman numeral chapter numbers for the Old Testament, but they also appear in the New Testament (see below).
Cardinali 2017, 17, 62–63.
Luke 1 (1304 A17 LM); Luke 2 (1306 C25 RM); Luke 3 (1308 C30 RM); Luke 6 (1314 A24 LM); Luke 7 (1316 B7 LM + TM); Romans 2 (1446 A42 LM); Romans 3 (1447 B8 LM + TM); Romans 4 (1448 B6 LM + TM); Romans 5 (1449 B2 LM + TM); Romans 6 (1450 A37 LM); Romans 7 (1451 A22 LM + TM); Romans 8 (1452 A24 LM); Romans 9 (1453 B42 LM + BM); Romans 10 (1454 C18 LM + TM).
Versace 2018, 66, 304–305.
Acerbi & Bianconi 2022, 93; Cardinali 2017, 62–63; Versace 2018, 66–67.
Acerbi & Bianconi 2022, 91–93; Versace 2018, 66–67, 306.
Acerbi & Bianconi 2022, 89–100.
Acerbi & Bianconi 2022, 96–99.
To be precise, the vitriolic solution from which the vitriol used to make the ink was extracted was chemically purified.
Niccum 1997, 245.
Payne 2010b, 17–18; Payne & Canart 2000, 109 n. 25. The Genesis portion of the minuscule replacement pages were apparently copied from the twelfth-century manuscript BAV, Chig. R.VI.38 but the source for the replacement pages containing Psalms and Hebrews has not been identified, see Acerbi & Bianconi 2022, 55–56.
“une ressemblance visuelle n’est pas une garantie d’identité. Une analyse spectrométrique de l’encre permettrait peut-être de tirer au clair cette question” (Andrist 2009, 241).
Payne 2010a, 14 n. 36.
For a summary of these SBL lectures, see Wasserman 2023.
Bibliography
Acerbi, Fabio & Bianconi, Daniele (2022): Il Codex Vaticanus a Bisanzio: vicende e figure di una storia millenaria. Città del Vaticano: Biblioteca Apostolica Vaticana (Studi e testi 556).
Aceto, Maurizio & Agostino, Angelo & Boccaleri, Enrico, & Garlanda, Anna Cerutti (2008): The Vercelli Gospels Laid Open: An Investigation into the Inks Used to Write the Oldest Gospels in Latin, in X-Ray Spectrometry 37/4, 286–292.
Amphoux, Christian-Bernard (2007): Codex Vaticanus B: Les points diacritiques des marges de Marc, in The Journal of Theological Studies 58/2, 440–466.
Andrist, Patrick (2009): Le milieu de production du Vaticanus graecus 1209 et son histoire postérieure: le canon d’Eusèbe, les listes du IVe siècle des livres canoniques, les distigmai et les manuscrits connexes, in Andrist, Patrick (ed.): Le manuscrit B de la Bible: (Vaticanus graecus 1209): Introduction au fac-similé, Actes du Colloque de Genève (11 juin 2001), Contributions supplémentaires. Lausanne: Éditions du Zèbre, 227–256.
Andrist, Patrick (2015): La structure des codex Vaticanus, Alexandrinus et Sinaiticus: questions ouvertes sur le canon, la fabrication et la circulation de ces bibles, in Ruzzier, Chiara & Hermand, Xavier (eds.): Comment le Livre s’est fait livre. La fabrication des manuscrits bibliques (IVe–XVe siècle): Bilan, résultats, perspectives de recherche. Turnhout: Brepols Publishers, 11–37 (Bibliologia 40).
Andrist, Patrick (2018): Érasme 1514–1516 et les étapes de la préparation du texte biblique et des prologues grecs du Novum Instrumentum. Le témoignage des manuscrits, in Dahan, Gilbert & Noblesse-Rocher, Annie (eds.): La Bible de 1500 à 1535. Turnhout: Brepols Publishers (Bibliothèque de l’Ecole des Hautes Etudes, Sciences Religieuses 181), 135–195.
Andrist, Patrick (2021): Au croisement des contenus et de la matière: les structures des sept pandectes bibliques grecques du premier millénaire, in Scrineum Rivista 17/2, 3–106.
Beck, Lily Y. (trans.) (2005): De materia medica by Pedianus Dioscorides. Hildesheim: Olms-Weidmann (Altertumswissenschaftliche Texte und Studien 38).
Beckmann, John & Francis, William, & Griffith, J.W. (1846): A History of Inventions, Discoveries, and Origins, trans. William Johnston. Vol. 1, 4th ed. London: Bohn.
Bibliorum sacrorum Graecorum Codex Vaticanus B: Bibliothecae Apostolicae Vaticanae Codex Vaticanus Graecus 1209 (1999): Roma: Istituto Poligrafico e Zecca dello Stato.
Biringuccio, Vannoccio (1990): The Pirotechnia of Vannoccio Biringuccio: The Classic Sixteenth-Century Treatise on Metals and Metallurgy, trans. Cyril Stanley Smith & Martha Teach Gnudi. New York: Dover Publications.
Bogaert, Pierre-Maurice (2009): Le Vaticanus graecus 1209 témoin du texte grec de l’Ancien Testament, in Andrist, Patrick (ed.): Le manuscrit B de la Bible: (Vaticanus graecus 1209): Introduction au fac-similé, Actes du Colloque de Genève (11 juin 2001), Contributions supplémentaires. Lausanne: Éditions du Zèbre, 47–76.
Bordat, Josef (2007): Sepúlveda, Juan Ginés de, in Bautz, Friedrich Wilhelm (ed.): Biographisch-Bibliographisches Kirchenlexikon. Vol. 27. Nordhausen: Traugott Bautz, 1343–1345.
Bourgarit, David & Bauchau, Fanny (2010): The Ancient Brass Cementation Processes Revisited by Extensive Experimental Simulation, in JOM 62/3, 27–33.
Birch, Thomas (ed.) (1772): The Works of the Honourable Robert Boyle. London: Johnston et al.
Bromehead, C.N. (1956): Mining and Quarrying to the Seventeenth Century, in A History of Technology. Vol. 2. New York: Oxford University Press, 1–40.
Canart, Paul (2009): Post-Scriptum, in Andrist, Patrick (ed.): Le manuscrit B de la Bible: (Vaticanus graecus 1209): Introduction au fac-similé, Actes du Colloque de Genève (11 juin 2001), Contributions supplémentaires. Lausanne: Éditions du Zèbre, 223–225.
Cardinali, Giacomo (2017): “Qui avemo uno spagnolo dottissimo”. Gli anni italiani di Pedro Chacón (1570ca.–1581). Saggio di ricostruzione bio-bibliografica a partire da carteggi coevi. Città del Vaticano: Biblioteca Apostolica (Studi e testi 513).
Cavallo, Guglielmo (1967): Ricerche sulla maiuscola biblica. Vol. 1. Firenze: Le Monnier.
Delange, Elisabeth & Grange, Maurice & Kusko, Bruce, & Menei, Eve (1990): Apparition de l’encre métallogallique en Égypte à partir de la collection de papyrus du Louvre, in Revue d’Égyptologie 41, 213–217.
Díaz Hidalgo & Rafael Javier & Córdoba, Ricardo & Grigoryan, Hermine & Vieira, Márcia & Melo, Maria J. & Nabais, Paula & Otero, Vanessa & Teixeira, Natércia & Fani, Sara, & Al-Abbady, Hossam (2023): The Making of Black Inks in an Arabic Treatise by al-Qalalūsī Dated from the 13th c.: Reproduction and Characterisation of Iron-gall Ink Recipes, in Heritage Science 11/1, 7.
Diels, Hermann & Schramm, Erwin Adelbert (1920): Exzerpte aus Philons Mechanik B. VII und VIII: (Vulgo fünftes Buch); Griechisch und Deutsch, in Abhandlungen der Preußischen Akademie der Wissenschaften. Philosophisch-Historische Klasse 12, 3–84.
Epp, Eldon Jay (2005): Junia: The First Woman Apostle. Minneapolis: Fortress Press.
Fellows, Richard G. (2019): Are There Distigme-Obelos Symbols in Vaticanus?, in New Testament Studies 65, 246–251.
Forbes, R.J. (1993): Studies in Ancient Technology. Vol. 3. 3rd ed. Leiden: Brill.
Free, Michael L. (2013): Hydrometallurgy Fundamentals and Applications. Hoboken: John Wiley & Sons, Inc.
Geissbühler, Mirjam & Dietz, Georg & Hahn, Oliver, & Rabin, Ira (2018): Advanced Codicological Studies of Cod. Germ. 6 (Hamburg, Staats- und Universitätsbibliothek): Part 2, in Manuscript Cultures 11, 133–139.
Ghigo, Tea & Rabin, Ira, & Buzi, Paola (2020): Black Egyptian Inks in Late Antiquity: New Insights on Their Manufacture and Use, in Journal of Archaeological and Anthropological Sciences 12, 70.
Golas, Peter J. (1999): Science and Civilisation in China 5.13 (Chemistry and Chemical Technology: Mining). Cambridge: Cambridge University Press.
Gordon, Nehemia & Bonnerot, Olivier, & Rabin, Ira (2020): Inks Used to Write the Divine Name in a Thirteenth-Century Ashkenazic Torah Scroll: Erfurt 7 (Staatsbibliothek zu Berlin, Ms. or. fol. 1216), in Manuscript Cultures15, 163–184.
Gravely, Edward D. (2009): The Text Critical Sigla in Codex Vaticanus. PhD dissertation. Wake Forest: Southeastern Baptist Theological Seminary.
Grenz, Jesse R. (2018): Textual Divisions in Codex Vaticanus: A Layered Approach to the Delimiters in B(03), in TC: A Journal of Biblical Textual Criticism23, 1–22.
Grenz, Jesse R. (2021): The Scribes and Correctors of Codex Vaticanus: A Study on the Codicology, Paleography, and Text of B(03). PhD dissertation. Cambridge: University of Cambridge.
Gurry, Peter (2022): Snapp on the Distigmai in Vaticanus, Evangelical Textual Criticism. evangelicaltextualcriticism.blogspot.com/2022/08/snapp-on-distigmai-in-vaticanus.html.
Habashi, Fathi (1993): A Textbook of Hydrometallurgy. Québec: Métallurgie Extractive Québec.
Habashi, Fathi (2005): A Short History of Hydrometallurgy, in Hydrometallurgy 79/1–2, 15–22.
Hahn, Oliver & Malzer, Wolfgang & Kanngießer, Birgit, & Beckhoff, Burkhard (2004): Characterization of Iron-gall Inks in Historical Manuscripts and Music Compositions Using X-Ray Fluorescence Spectrometry, in X-Ray Spectrometry 33/4, 234–239.
Hahn, Oliver & Nehring, Grzegorz & Freisitzer, Rudolf, & Rabin, Ira (2019): A Study on Early European Inks from St Paul in Lavanttal, in Gazette du livre médiéval 65/1, 58–81.
Hahn, Oliver & Wolff, Timo & Feistel, Hartmut-Ortwin & Rabin, Ira, & Beit-Arié, Malachi (2007): The Erfurt Hebrew Giant Bible and the Experimental XRF Analysis of Ink and Plummet Composition, in Gazette du livre médiéval 51/1, 16–29.
Han, Kenneth N. (2009): Fundamentals of Aqueous Metallurgy. Littleton: Society for Mining, Metallurgy, and Exploration.
Hassan, Ahmad Y. & Hill, Donald R. (1986): Islamic Technology: An Illustrated History. Cambridge: Cambridge University Press.
Head, Peter M. (2009): The Marginalia of Codex Vaticanus: Putting the Distigmai (Formerly Known as ‘Umlauts’) in Their Place. Lecture presented at the SBL Annual Meeting, New Orleans.
Hernández, Albert (2011): Juan Ginés De Sepúlveda, in De La Torre, Miguel & Floyd-Thomas, Stacey (eds.): Beyond the Pale: Reading Theology from the Margins. Louisville: Westminster John Knox Press, 79–86.
Hickel, Erika (1963): Chemikalien im Arzneischatz deutscher Apotheken des 16. Jahrhunderts, unter besonderer Berücksichtigung der Metalle. Braunschweig: Technische Hochschule Carolo-Wilhelmina zu Braunschweig.
Hronček, Pavel & Gregorová, Bohuslava & Tometzová, Dana & Jesenský, Miloš (2021): Scientific Journeys to One of the Oldest Copper Cementation Sites in Central Europe (Smolník, Slovakia), in History of Geo- and Space Sciences, 12/2, 179–196.
Jost, Alexander Friedrich (2014): From Secret Knowledge to Mass Production: The Wet Copper Industry of Song China (960–1279). PhD dissertation. Tübingen: Eberhard-Karls-Universität Tübingen.
Karpenko, Vladimír (1995): Fe(s) + Cu(II)(aq) -> Fe(II)(aq) + Cu(s): Fifteen Centuries of Search, in Journal of Chemical Education 72/12, 1095.
Karpenko, Vladimír & Norris, John A. (2002): Vitriol in the History of Chemistry, in Chemicke Listy 96, 997–1005.
Kenyon, Frederic G. (1912): Handbook to the Textual Criticism of the New Testament, 2nd ed. London: Macmillan.
Kirnbauer, Franz (1966): Copper Mining and Copper Smelting from the Middle Ages until 1900, in Copper in Nature: Technics, Art and Economy (Translated from the German Kupfer in Natur, Technik, Kunst und Wirtschaft). Hamburg: Norddeutsche Affinerie, 40–53.
Kloha, Jeffrey John (2006): A Textual Commentary on Paul’s First Epistle to the Corinthians. PhD Dissertation. Leeds: University of Leeds.
Knust, Jennifer Wright & Wasserman, Tommy (2018): To Cast the First Stone: The Transmission of a Gospel Story. Princeton: Princeton University Press.
Krans, Jan (2019): Paragraphos, Not Obelos, in Codex Vaticanus, in New Testament Studies 65, 252–257.
Krans, Jan (2020): Erasmus and Codex Vaticanus: An Overview and an Evaluation, in Annali di Storia dell’Esegesi37/2, 447–470.
Kraschewski, Hans-Joachim (2001): Vitriole aus dem Rammelsberg, in Roseneck, Reinhard (ed.): Der Rammelsberg. Tausend Jahre Mensch—Natur—Technik: Bergbau als Kulturträger. Vol. 2. Goslar: Goslarsche Zeitung, 344–356.
Krekel, Christoph (1999): The Chemistry of Historical Iron Gall Inks: Understanding the Chemistry of Writing Inks Used to Prepare Historical Documents, in International Journal of Forensic Document Examiners 5, 54–58.
Lacko, Miroslav & Mayerová, Erika (2016): Das älteste Stadtbuch von Schmöllnitz 1410–1735 Eine Quelle zu den mitteleuropäischen wirtschaftlichen Verflechtungen. Limbach: Slovenská spoločnosť pre sociálne a hospodárske dejiny/ Slowakische Gesellschaft für Sozial- und Wirtschaftsgeschichte.
Lung, T.N. (1986): The History of Copper Cementation on Iron—The World’s First Hydrometallurgical Process from Medieval China, in Hydrometallurgy 17/1, 113–129.
Medlej, Joumana (2020): Inks & Paints of the Middle East: A Handbook of Abbasid Art Technology. London: Majnouna.
Metzger, Bruce M. (1991): Manuscripts of the Greek Bible: An Introduction to Greek Palaeography. New York: Oxford University Press.
Miller, Jeffrey Edward (2000): Scribal Sigla for Variant Readings in Vaticanus, with a Response to Philip Payne’s Conclusions in Fuldensis, Sigla for Variants in Vaticanus, and 1Cor 14.34–35. Master’s Thesis. Dallas: Dallas Theological Seminary.
Miller, Jeffrey Edward (2003): Some observations on the Text-Critical Function of the Umlauts in Vaticanus, with Special Attention to 1Corinthians 14.34–35, in Journal for the Study of the New Testament 26, 217–236.
Milne, Herbert John Mansfield & Skeat, Theodore Cressy (1938): Scribes and Correctors of the Codex Sinaiticus. London: British Museum.
Nehring, Grzegorz & Bonnerot, Olivier & Gerhardt, Marius & Krutzsch, Myriam, & Rabin, Ira (2021): Looking for the Missing Link in the Evolution of Black Inks, in Archaeological and Anthropological Sciences 13/4, 71.
Nehring, Grzegorz & Bonnerot, Olivier & Gordon, Nehemia, & Rabin, Ira (2021): Writing and Correcting a Torah Scroll in Germany of the Thirteenth and Fourteenth Centuries, in Comparative Oriental Manuscript Studies Bulletin 7, 123–147.
Nehring, Grzegorz & Gordon, Nehemia & Rabin, Ira (2022): Distinguishing Between Seemingly Identical Inks Using Scanning µXRF and Heat Maps, in Journal of Cultural Heritage 57, 142–148.
Neumann, Bernhard (1904): Die Metalle Geschichte, Vorkommen und Gewinnung, nebst ausführlicher Produktions- und Preis-Statistik. Vom Verein zur Beförderung des Gewerbefleisses preisgekrönte Arbeit. Halle a.S.: Wilhelm Knapp.
Niccum, Curt (1997): The Voice of the Manuscripts on the Silence of Women: The External Evidence for 1Cor 14.34–35, in New Testament Studies 43, 242–255.
O’Neill, John Cochrane (1989): The Rules Followed by the Editors of the Text Found in the Codex Vaticanus, in New Testament Studies 35, 219–228.
Payne, Philip B. (1995): Fuldensis, Sigla for Variants in Vaticanus, and 1Cor 14.34–35, in New Testament Studies 41, 240–262.
Payne, Philip B. (2004): The Text-Critical Function of the Umlauts in Vaticanus, with Special Attention to 1Corinthians 14.34–35: A Response to J. Edward Miller, in Journal for the Study of the New Testament 27, 105–112.
Payne, Philip B. (2010a): Do the Marginalia of Vaticanus Support or Undermine the Originality of its Distigmai?, www.pbpayne.com/wp-content/uploads/2010/03/Critique-of-Vaticanus-Marginalia-15Apr2010.pdf.
Payne, Philip B. (2010b). Critique of “The Marginalia of Codex Vaticanus: Putting the Distigmai in Their Place” presented by Peter M. Head to the NT Textual Criticism Seminar Nov. 21, 2009 in New Orleans, www.pbpayne.com/wp-content/uploads/2010/01/Critique_of_the_Marginalia_of_Codex_B12.pdf.
Payne, Philip B. (2017): Vaticanus Distigme-obelos Symbols Marking Added Text, Including 1Corinthians 14.34–35, in New Testament Studies 63/4, 604–625.
Payne, Philip B. (2019): Is 1Corinthians 14:34–35 a Marginal Comment or a Quotation? A Response to Kirk MacGregor, in Priscilla Papers 33, 24–30.
Payne, Philip B. (2021): Critique of Vaticanus Distigme-Obelos Denials, www.pbpayne.com/wp-content/uploads/2021/07/Critique-of-Fellows-Krans-Vaticanus-Distigme-Obelos-Denials.pdf.
Payne, Philip B. (2022): The Authenticity and Significance of the Sixteen Vaticanus Distigme-Obelos Symbols that Mark Where Text Was Inserted, Including 1Corinthians 14:34–35, Unpublished Paper.
Payne, Philip B. (2023): The Bible vs. Biblical Womanhood: How God’s Word Consistently Affirms Gender Equality. Grand Rapids: Zondervan.
Payne, Philip B. & Canart, Paul (2000): The Originality of Text-Critical Symbols in Codex Vaticanus, in Novum Testamentum 42, 105–113.
Payne, Philip B. & Canart, Paul (2009): Distigmai Matching the Original Ink of Codex Vaticanus: Do They Mark the Location of Textual Variants?, in Andrist, Patrick (ed.): Le manuscrit B de la Bible: (Vaticanus graecus 1209): Introduction au fac-similé, Actes du Colloque de Genève (11 juin 2001), Contributions supplémentaires. Lausanne: Éditions du Zèbre, 199–223.
Peterson, R.C. (2003): The Relationship between Cu Content and Distortion in the Atomic Structure of Melanterite from the Richmond Mine, Iron Mountain, California, in The Canadian Mineralogist 41/4, 937–949.
Schöne, Richard (ed.) (1893): Philonis Mechanicae syntaxis libri quartus et quintus. Berlin: Reimeri.
Rabin, Ira (2021): Material Studies of Historic Inks: Transition from Carbon to Iron-Gall Inks, in Raggetti, Lucia (ed.): Traces of Ink: Experiences of Philology and Replication. Leiden: Brill (Nuncius Series 7), 70–78.
Rabin, Ira & Cohen, Zina, & Gordon, Nehemia (2024): Inks of the Bologna Scroll: A Preliminary Study, in Manuscript Studies 9, 222–242.
Rabin, Ira & Krutzsch, Myriam (2019): The Writing Surface Papyrus and Its Materials 1. Can the Writing Material Papyrus Tell Us Where It Was Produced? 2. Material Study of the Inks, in Nodar, Alberto & Torallas Tovar, Sofía & Albarrán Martínez, María Jesús & Hernández, Raquel Martin & Pajón Leyra, Irene (eds.): Proceedings of the 28th Congress of Papyrology: Barcelona 1–6 August 2016. Barcelona: Publicacions Abadia de Montserrat (Scripta Orientalia 3), 773–781.
Rabin, Ira & Schütz, Roman & Kohl, Anka & Wolff, Timo & Tagle, Roald & Pentzien, Simone & Hahn, Oliver & Emmel, Stephen (2012): Identification and Classification of Historical Writing Inks in Spectroscopy: A Methodological Overview, in Comparative Oriental Manuscript Studies Newsletter 3, 26–30.
Rabin, Ira & Wintermann, Carsten & Hahn, Oliver (2019): Ink Characterization, Performed in Biblioteca Medicea Laurenziana (September 2018), in Analecta Papyrologica 31, 301–313.
Read, Thomas T. (1937): China’s Civilization Simultaneous, Not Osmotic, in The American Scholar 6/2, 249–251.
Ropes, James Hardy (1926): The Text of Acts. London: Macmillan (The Beginnings of Christianity 3).
Skeat, Theodore Cressy (1999): The Codex Sinaiticus, the Codex Vaticanus, and Constantine, in The Journal of Theological Studies 50/2, 583–625.
Snapp, Jr., James (2022): Distigmai (Umlauts)—Solving the Mystery. www.thetextofthegospels.com/2022/08/distigmai-umlauts-solving-mystery.html.
Stillman, John Maxson (1960): The Story of Alchemy and Early Chemistry. New York: Dover Publications.
Swete, Henry Barclay (1900): An Introduction to the Old Testament in Greek. Cambridge: Cambridge University Press.
Tischendorf, Constantin (1867a): Novum Testamentum Vaticanum. Lipsiae: Giesecke et Devrient.
Tischendorf, Constantin (1867b): Appendix codicum celeberrimorum Sinaitici, Vaticani, Alexandrini. Cum imitatione ipsorum antiqua manu scriptorum nunc primum ed. Lipsiae: Giesecke et Devrient.
Traube, Ludwig (1907): Nomina sacra: Versuch einer Geschichte der Christlichen Kürzung. München: Beck.
Ts’ao, T’ien-Ch’in & Ho, Ping-Yü & Needham, Joseph (1959): An Early Medieval Chinese Alchemical Text on Aqueous Solutions, in Ambix 7/3, 122–157.
Vercellone, Carlo & Cozza-Luzi, Giuseppe (eds.) (1868): Bibliorum Sacrorum Graecus Codex Vaticanus. Vol. 5. Roma: S. Congregationis de propaganda fide.
Versace, Pietro (2018): I marginalia del Codex Vaticanus. Città del Vaticano: Biblioteca Apostolica Vaticana (Studi e testi 528).
Vuillard, Clément & Radepont, Marie & Téreygeol, Florian & Remazeilles, Céline & Rouchon, Véronique (2024): From Experimental Archaeology to Laboratory: Mineralogical Nature and Elemental Composition of Medieval Manufactured Vitriols, in Journal of Archaeological Science: Reports 55, 1–12.
Waite, Arthur Edward (ed.) (1976): The Hermetic and Alchemical Writings of Paracelsus. Vol. 1. Berkeley: Shambhala Publications Inc., 89–113.
Wasserman, Tommy (2009a): SBL New Orleans 2009 I: Peter Head Putting the Distigmai in the Right Place Pt. 1. evangelicaltextualcriticism.blogspot.com/2009/11/sbl-new-orleans-2009-i-peter-head.html.
Wasserman, Tommy (2009b): SBL New Orleans 2009 I: Peter Head Putting the Distigmai in the Right Place Pt. 2. evangelicaltextualcriticism.blogspot.com/2009/11/sbl-new-orleans-2009-i-peter-head_22.html.
Wasserman, Tommy (2023): Distigmai Revisited: When Material Analysis Catches Up with Common Sense. evangelicaltextualcriticism.blogspot.com/2023/11/distigmai-revisited-when-material.html.
Wellmann, Max (ed.) (1914): Pedanii Dioscuridis De materia medica libri quinque. Vol. 3. Berlin: Weidmann.
Wenzel, Gusztáv (1880): Magyarország bányászatának kritikai története. Budapest: M. Tud. Akadémia Könyvkiadó Hivatala.
Whitehead, David (2016): Philo Mechanicus: On Sieges Translated with Introduction and Commentary. Stuttgart: Franz Steiner.
Willker, Wieland (2001): Codex Vaticanus Graece 1209, B/03: The Umlauts. www.willker.de/wie/Vaticanus/umlauts.html.
Yi, An-Ting (2024): From Erasmus to Maius: The History of Codex Vaticanus in New Testament Textual Scholarship. Berlin: De Gruyter.
Zerdoun Bat-Yehouda, Monique (1983): Les Encres Noires au Moyen-Âge (jusqu’à 1600). Paris: Centre National de la Recherche Scientifique.