The Sergio: The Life and Times of the World’s Largest Carbonado Part 2

Acknowledgements

Lucia Burgio is grateful to the UKRI Arts and Humanities Research Council for the award of a Capability for Collection grant that allowed the purchase of the Bruker Artax XRF spectrometer used in this study (grant reference AH/V012134/1).

Particular thanks are extended to Alex Ball, Head of Division, Imaging and Analysis, Tom Ranson from the Science Innovation Platforms Department, and Thea Mainprize, Science Department, of the Natural History Museum, London, for their contribution to our work by scanning and creating the 3D replicas of the NHM models of The Sergio.

Abstract

The Sergio, named after its finder, was the largest carbonado ever discovered. It passed through the hands of several companies as it was transported from Brazil, to Paris, to London, where it was broken up and sold for its use as an industrial abrasive. The discovery of two models of The Sergio in the collection of the Natural History Museum, London, prompted us to search for more information about the life and times of this amazing carbonado. In Part One of our research journey, we provided a detailed description of those models, enabling a better understanding of The Sergio’s shape and dimensions. Here, in Part Two of our journey, we trace, step by step, the history of The Sergio, from its likely origin to its discovery then its destruction, including details about the lives and motives of those who dealt with it. To our delight, our research led us to the location of another model of The Sergio, one acquired by Henri Moissan, who had donated it to the mineral collections of the Muséum national d’histoire naturelle, Paris in 1896. The finding of this earlier model, describing it, and making detailed comparisons between it and the NHM models enabled us to record, for the first time, and with considerable certainty, the dimensions of original Sergio. We hope that our research journey into the life and times of The Sergio has paid tribute to this remarkable stone.

Introduction

In Part One of this story of The Sergio, the world’s largest carbonado (Hansen et al., 2024), we referred to an article about black diamonds (Hansen and Rennie, 2022) in which we wrote briefly about the history of The Sergio based on the information we had so far uncovered that seemed to be credible. In summary, it was found on 15 July, 1895 by a miner named Sergio Borges de Carvalho, from whom it was first bought and who is recognised by those who refer to the carbonado as “The Sergio”. Usually, however, this stone is referred to in contemporary writings simply as the largest carbon ever found (carbonados at that time were referred to as “carbons”).

The fortuitous discovery of two models in the collections of the Natural History Museum, London (NHM) led us on a journey to uncover the history of this amazing carbonado. The models were acquired by the Museum of Practical Geology (Model BM.1985,MI9558) and the British Museum (Natural History) (Model BM.1911,119) in 1911 from J.R. Gregory & Co. In the first part of our journey, we described the models in detail and ascertained that they were indeed copies of the original Sergio. Composite Figure 1 will remind readers of the nature of the models described in Part One. At the top are reproductions of the two published images that began our story. At left (Figure 1a) is the photograph of The Sergio usually pictured in published references to it (Furniss, 1906), and at right (Figure 1b) is an image of the engraving from the report of The Sergio by Henri Moissan (1895). Below (Figures 1c and 1d) are photographs of the two models in the NHM, shown in the same orientation as the images above, thus showing the ‘front’ and ‘reverse’ sides of this unique carbonado.

During our examination of the models, we found a list of references written on the back of the label for BM.1911,119 (Figure2). Three of these were new to us and opened further avenues for research into what had happened to The Sergio. This article records the second part of our journey to uncover the events between the finding of The Sergio and its ultimate and unfortunate destruction.

It is worth restating the summary given by Furniss (1906), who was the American Consul at Bahia around this time (see Note 1), because his words are frequently repeated by other commentators and they contain details on which we elaborate later in this article. This 1906 report also provides a concise summary of the nature of mining diamonds and carbonados in Brazil up to that time.

The largest carbon ever encountered was found near Lenções in 1895, on the ledge of a mountain which had been worked some time before. It weighed, when found, 3,165 carats, was purchased from the miner for $16,000 and was finally exported to London, where it sold for $31,145, having lost about 50 carats meantime in drying out. In London it was broken into pieces suitable for drills and these pieces sold for about $40,000, while at the present price of carbon they would be worth about $158,000. (Furniss, 1906, p.6).

Because it was a carbonado, we know a considerable amount about The Sergio, even though it is no longer physically present. The chemical, physical and optical properties of carbonado are described by Haggerty (2014). Carbonados are found in only two locations, Bahia, Brazil (first found in 1841) and the Central African Republic (discovered around 1925), and they have similar properties. In both locations, the carbonados are derived from reworked sediments and have not been traced back to a primary rock source. Haggerty reports that analyses of inclusions in carbonados, based on radiogenic lead (Pb), date them to 3.2–3.9Ga, considerably older than diamonds found in kimberlite. Haggerty sets out a compelling argument for an extraterrestrial origin of carbonados that would mean that The Sergio was ‘born’ outside of the Earth. It is likely to be even older than the 3.2–3.9Ga spent during its life on Earth where it was incorporated into, and reworked in and out of, a series of sediments before being picked out of the mud by Sergio, the miner, in 1895.

In stark contrast, we can say the life of The Sergio after that moment was very brief indeed, probably only a few months. What happened in those months?

From Furniss (1906), cited above, we know that The Sergio was found in Brazil, purchased there, and “finally exported” to London. Our subsequent research enabled us to fill in much more detail. In our quest we were aided by the new references found on the back of the label for one of the models (see Figure 2), the vast libraries of the NHM and the records of the Muséum national d’histoire naturelle, Paris.

To provide context we begin with a short overview of the carbonado industry during the period when The Sergio was found, based on Furniss (1906) and other sources including Kunz (1904), who relied extensively on an earlier 1902 report by Furniss (see Note 1). For readers wishing for more detail, we recommend the comprehensive description of the history, geology and mining of Brazilian diamonds and carbonados given by Svisero et al. (2017).

The Diamond and Carbonado Industry in Brazil

In Bahia, diamonds and carbonados are found in sediments and gravels accessible mainly where rivers have cut through the overlying conglomerate and sandstone, leaving concentrations in the riverbeds or still in the sediments. Furniss (1906) explained that the mining was done primarily by ‘garimpeiros’, individuals either working for themselves or in shares with the owner of the claim. The mining methods were primitive, using a few hand tools to remove the subsoil, break up the sediments, and sieve the broken material in the river. In some places, divers could collect the gravel from the riverbeds and bring it to the surface for sieving.

Carbonados were used for abrasives, and as the demand for diamond drills increased from the late 1890s into the 1900s, so did the price for carbonado. The ideal size for carbons was around 3–4 carats, the size used in drills, and prices were highest for this size.Smaller pieces also had use as abrasives, but very large pieces had to be broken up, a costly process due to their hardness, and so the price per carat was somewhat less for large pieces.

Diamonds and carbons were sold to buyers in the field, with the miner providing 20%–25% of the price to the holder of the lease being worked. Stones were taken to Bahia and resold through agents to international buyers, then sent to Paris or London for sorting, with much imported by the United States. An overview of the trade and underlying financial considerations, prepared by Herold (2013), throws some light on this somewhat complicated process!

Finding and Selling The Sergio

Moissan (1895) gave the date of finding The Sergio as 15 July, 1895, and that date is not questioned in contemporary reports although in later reports the year is sometimes given as 1905, for example by Svisero et al. (2017). The Sergio was first sold by the miner for US$16,000, with one fourth going to the lease owner. According to Kunz (1904), it “went through several hands” and was purchased for $25,400 and sent abroad. In an earlier publication, Kunz (1896) noted that this “largest piece of carbonado or bort known, or, indeed, the largest mass of diamond substance of any kind …was taken to Paris, and a strong effort has been made on the part of the Brazilian Government to obtain it for the national museum at Rio Janeiro [sic]” (p.903). Clearly, that effort was unsuccessful, but we know a little more about the chain of events from an article by Yawger, published in The Mining World in 1907. Yawger wrote as follows:

The extraordinary carbon of 3,078 karats [sic]… was discovered about 12 years ago and was purchased by Messrs. Kahn & Co. of Bahia, Brazil. It was first offered to The Yawger-Demmert Company’s predecessors, Victor Bishop & Co. (the pioneers in this carbon business in the United States) through their then resident agent in Brazil, Mr Frank Dennis, who sent a photograph and description of the stone. (This original photograph is still in the possession of The Yawger-Demmert Company, and is here reproduced.) The dimensions of the carbon were approximately 3 in. by 3 in. by 3½ in. Victor Bishop & Co. did not deem it expedient to buy the stone and assume the great risk of breaking it, so it was sent by Kahn & Co. to their house in Paris, where it was again offered to Victor Bishop & Co., and as they still refused to buy it, the carbon was sold by Kahn & Co. for £6,400 ($32,000) to Mr. J. K. Gulland, M. E., of 8 Victoria street, London, E. C., a well known dealer in carbon. Mr Gulland split the carbon into the required sizes for diamond drill work. (pp.7-8).

This extract provides several points of information. First, the photograph included by Yawger in his article is identical to that reproduced by Brazil (1904), Furniss (1906), Herold (2013) and Haggerty (2014), so it appears that the original photograph was taken by Frank Dennis of Victor Bishop & Co. and was still in existence at least a decade later. This is the photograph usually used to depict The Sergio, although when a source is cited for the image it is invariably attributed to Furniss (1906). When Moissan (1895) saw The Sergio in Paris, only two months after it was found, he likely had no access to the original Dennis photograph and therefore he procured an engraving of it, fortuitously giving us a view of its other side.

Second, Yawger indicates that because of its size, The Sergio was difficult to sell. It was purchased by Kahn & Co. in Brazil and hence made its way to their house in Paris. We find details of this in an article by Jacques Baszanger, head of Jacques Baszanger & Co based in Paris, published in 1909 in The Mining Journal. Baszanger stated:

I actually did sell the large carbon in question to Mr. Gulland for account of Messrs. C. Kahn & Co., of Paris, importers of carbon, in September 1895 for the sum of £6,400 – I acting at Mr C. Kahn’s request as a broker, and receiving my brokerage.….At the time, Mr. C. Kahn, who is a personal friend of mine, knew so little what to do with the large stone, and depended so much on my mediation in its disposal, that he consulted me as to the advisability of selling the stone to the Brazilian Government, who wanted it for the State museum. He telegraphed me on July 27, 1895, on the subject, and I beg to enclose herewith a photographic copy of that telegram [not printed in the article]. In the year 1895, it is true, there was practically no other dealer in carbon who would have undertaken to break such a risky stone but Mr. Gulland. (Baszanger, 1909a, pp.333-334).

Indirectly, Baszanger also explains why The Sergio was not sold to the Brazilian Government. Only twelve days after it was found, and by then in the hands of the Brazilian office of Kahn & Co., it seems Baszanger advised Mr. C. Kahn against that sale and then, when The Sergio reached Paris in September, he obtained a brokerage fee for himself by selling it to Mr Gulland in London. Despite careful searches, we have been unable to find any further information about Mr. C. Kahn, or Kahn & Co, to follow up on Baszanger’s statements.

The Destruction of The Sergio

Both Yawger (1907) and Baszanger (1909a), cited above, indicated that The Sergio was broken up by Mr. Gulland, and that is true. On 14 November, 1902, an article was published in the Journal of the Society of the Arts by an unnamed author, but the contents are referenced to the 1902 Consular report by Furniss. This article stated that, after changing hands, the 3,150 carat stone “was finally purchased by a Bahia city exporter for £5,080, and was shipped to Paris, where it was broken up into marketable pieces” (Anonymous, 1902, p.930). A week later, the November 21 issue of the same Journal included a letter written by J. K. Gulland of London (see Note 2), in which he stated:

Regarding the largest carbon ever found, it was not broken up in Paris, I broke it up here myself. The exact weight was 3,078 carats. I bought the stone on the 24th of September, 1895, for £6,464, broke it up in pieces suitable for use in diamond drills and resold the whole at ten percent profit. Had I the stone now it would be worth £26,163. The present price of carbons at the mines for good carbons of one carat and upwards is £8 10s. to £9 per carat, not £5. (Gulland, 1902, p.22).

Evidently, Gulland’s article did not immediately settle the matter of when and where The Sergio was broken up, because five years later Yawger (1907) referred to being in possession of an affidavit by Gulland stating his role in breaking it up. In the October 2, 1909 issue of The Mining Journal, an article appeared under the name of Jacques Baszanger which stated “in 1895 a gigantic stone of 3,078 carats or 615 grammes, which was purchased and broken up by me personally into pieces of suitable size for diamond drills, a special machine being designed for the purpose” (Baszanger, 1909b, p.7). Not surprisingly, Gulland took exception to this claim, responding with a letter to the Editor published in the November 13, 1909, issue of The Mining Journal. “This statement has previously been made by M. Baszanger and completed refuted. M. Baszanger not only was not present when the carbon was broken up by me, but did not even see the carbon after breaking” (Gulland, 1909, p.258). Gulland referred to similar statements made by Baszanger in the United States and Canada and continued, “M. Baszanger called on me and apologised for the mis-statements, saying it was done without his knowledge, and assuring me it would not happen again” (p.258). Two weeks later, Baszanger’s reply appeared in the same journal:

[I]n the interest of truth and honesty I hasten to inform you that Mr Gulland’s statement that I apologised to him for the mis-statement that appeared in the United States and Canadian journals, to the effect that I bought and broke up the big carbon in question, is perfectly correct; and I wish to add that that statement was made erroneously by my New York agent, … at the time that big stone was sold by me as a broker, the New York agent was not connected with my firm, and was therefore not aware of the exact facts. (Baszanger, 1909a, p.333).

Baszanger also explained that this more recent publication came about from his London agent taking information from the old Canadian article in response to a request from The Mining Journal for an article on carbon. This may explain the conversion of 3078 carats (the weight when Baszanger sold The Sergio to Gulland) to 615g, because Canada had adopted the metric carat in 1907. He apologised again to Gulland, and also pointed out that “Today the Brazilian exporters look almost without exception to my firm for the disposal of their large carbons” (Baszanger, 1909a, p.334). It is worth noting that Baszanger routinely advertised his business on the cover page of issues of The Mining Journal, and Figure 3 is an image of his advertisement for his London Office in the 7 October, 1909 issue. It appeared in a page full of advertising for mining equipment, especially drills and rock-breakers.

We found another misconception in Yawger’s article. In his report for 1902 to the United States Geological Survey, Kunz (1904) referred to the three largest carbonados found to that date. He reproduced the original Dennis photographic image on one page (labelled “Largest Piece of Carbon Ever Found – Actual Size”), and he also cited the 1902 US Consular report by Furniss who stated that a carbonado of 975 carats was found in Bahia in 1894, making it the second largest carbonado (however this one, reputedly broken up in Paris, is not referred to elsewhere). The third largest carbonado was found in 1901; it weighed 750½ carats and was taken to the Dusseldorf Exposition in that year. Kunz provided several photographs of this third stone as it was broken up into pieces of three to four carats, the size used in diamond drills. Kunz was very careful to label these photographs as depicting the third largest carbon, however, Yawger (1907) has reproduced the same photographs but referred to them as the breaking up of the largest carbonado, The Sergio. This must be incorrect, and Yawger does not reference Kunz in his piece. We have found no photographic record of The Sergio’s destruction, nor any indication that there was a photographic or any other record.

Stories containing such misstatements pervade the literature around this time; misstatements get repeated and then muddy the waters of what actually happened. Another example is the statement by Herold (2013) that “The Sergio was smashed into small pieces in New York City and used in drills at the Mesabi” (p.12), which is not only incorrect but at odds with a correct statement on the same page referring to its breaking up by Gulland. Herold also stated “In 1930, Bahia’s black diamonds were being used to bore for iron in the Mesabi Range” (p. 24), which may be correct, but is a date too late for The Sergio. Nevertheless, a 1931 article about black carbons in Popular Mechanics written by an anonymous author stated incorrectly that “A 3078-carat carbon was broken up for the diamond drills used to develop the Mesabi Range” (Anonymous, 1931, p.987). Of greater interest to us, though, was the depicted image of The Sergio in this publication of Popular Mechanics. It was not the Dennis photograph usually copied from Furniss (1906) but shows a similar aspect to the engraving published by Moissan (1895), as can be seen by comparing Figure 1b (top right) with the Popular Mechanics image in Figure 4. It appears to be a photograph, but not of Moissan’s engraving which had a hachured texture. Could it be a ‘touched-up’ copy of this engraving, or, more excitingly, could it be a copy of another photograph of the original Sergio taken by Frank Dennis?

He was photographing the specimen for the purpose of showing it to potential buyers, so he would likely have taken photographs from more than one angle. However, Yawger (1907) refers to only one photograph, and we have not found reference to any others. The lighter parts of the Popular Mechanics image could be reflections of light on raised parts of the specimen (the original photograph shows a glossy surface; see Figure 1a, top left) or, given their positions, they could indicate wear on edges of the specimen. This suggested to us that it may be a photograph of a well-handled model of The Sergio. Popular Mechanics was published in Chicago, but models of The Sergio could well have made their way across the Atlantic to the US! As the image is not attributed to a source, we can do no more than guess!

Reminders of The Sergio: Its models

Apart from the written records synthesised above, the only tangible reminders of The Sergio are its models. The NHM models both came from J. R. Gregory & Co. in 1911, but they were likely made much earlier when interest in The Sergio, as the world’s largest carbon, was high. As far as we know, the earliest (and possibly the only) public display of The Sergio was in September 1895, by Moissan (1895) in Paris when The Sergio was in the possession of Kahn & Co. We know that Gulland purchased The Sergio from Kahn, via Baszanger, on 24 September, 1895 (Gulland, 1902). The Comptes-rendus de l’Académie des Sciences (Proceedings of the Academy of Sciences, 1895) and the Journal Officiel de la République Française (Official Journal of the French Republic) both reported that The Sergio was shown by Henri Moissan to the Academicians on 23 September, the day before its official sale to Gulland.

In England, Moissan’s presentation was referenced in a paragraph of the Pharmaceutical Journal, published in London on 28 September, 1895 (Figure 5) and an important point mentioned is that “M. Moissan has made an exact mould of it” (Special Correspondent, 1895, p.270).

Moissan had strong connections with the French Muséum national d’histoire naturelle (MNHN), so we imagined that this Museum was the logical place to deposit any cast of The Sergio.

The first author (RH) contacted François Farges, the scientist in charge of the French national gems collection, and sought his assistance. Initial searches of the historic collection of the MNHN for a ‘carbonado’ model or “The Sergio”, did not reveal a specimen. Then, in a moment of inspiration, François searched for “carbon” or “carbone”. This search revealed specimen number 96.135, a cast donated by Moissan was inventoried early in 1896; the number refers to the 135th specimen catalogued in 1896. The cast was actually categorised in the collection as a “Produit Artificiel”

(PA, Artificial Material), whereas its accompanying historical labels described it as a “Moulage d’un échant[ill]on de Carbone” (translated into English by author FF as “Cast of a sample of carbon)” from “Lançoẽs – Brésil” (Lençóis, Brazil) and donated by “M. Moissan”.

Figure 6 shows the accompanying registration form with the information recorded, and its content (translated by author FF) is shown below:

96.135
Carbon
M. Moissan
C.R. Sept 1895 (publication: Compte-Rendus de l’Académie des Sciences, Septembre 1895)
Cast of a sample of carbon weighing 630 g = 3073 carats
Found in between the Rio de Raucardo and the stream of the Brias, over the territory of the city of Lançoẽs [=Lençóis], prov. [province] de Bahia, Brazil.

The cast was registered by Professor Alfred Lacroix (then mineralogist at the MNHN) in 1896 under “Carbone”, “Moulage” or “Produit Artificiel”, instead of the scientific name for carbon, which is carbonado or, at the time, black diamond. Even more confusingly, he registered the cast like a natural mineral, giving detailed outcrop information (although mis-spelt) in Brazil. It was unusual for such a detailed geological locality to have been recorded in the MNHN at this time, even more so for “artificial materials”, when it would be more logical to record the laboratory and researcher of origin. Other historical casts (of rough or faceted diamonds or sapphires, for example) have never been inventoried as “artificial materials” at the MNHN, such as casts of the French Blue (Farges et al., 2009), the Miroir of Portugal diamonds (Farges, 2014) or the Ruspoli and Devonshire-Branicki sapphires (Farges et al., 2015). This category is reserved, at the MNHN, for synthetic and artificial minerals. Interestingly it includes, among others, the “diamonds” synthesized by Moissan at this period of time (that are, in fact, moissanite syntheses; see Note 3). All these confusing exceptions led to the ‘disappearance’ of the cast of The Sergio given by Moissan.

Finding the cast of The Sergio was great news for us, and with details closely matching those from Moissan’s original article published in 1895: surely this would be our most accurate reproduction of The Sergio! Figure 7 shows the Moissan model of The Sergio in the same orientation as the engraving in Moissan’s article published in 1895. Comparison with the engraving and the models held in the NHM can be made in Figures 1b and 1d, respectively.

An image of the Moissan model in the same orientation as the Furniss photograph (Figure 1a), is shown on the left of Figure 8; the similarity to the two NHM models can be seen by comparing the Figure 8 image with Figure 1c. Shown on the right of Figure 8 is a plastic 3D print of the Moissan model. When author FF discovered that the Museu do Garimpeiro of Lençois in Bahia (where the original Sergio was found) held only a faded photograph of The Sergio, he made this 3D replica and it is on its way to colleagues in Bahia to present to this Museum.

The Moissan model, MNHN 96.135, is very well made: it is a black colour with the surface finish creating a pseudo-adamantine lustre, much like the picture published in Furniss (1906). It appears that the model is painted firstly in a silvery paint which is then covered by a black layer; a material which easily rubs off to blacken the fingers on touching, as with the black coating on the models held in the NHM (Hansen et al., 2024).

To obtain further information about the cast, MNHN 96.135 was scanned at the SURFACUS facility of the MNHN using a Faro Edge Arm, operated with a 450nm, CDRH/IEC class 2M, blue laser beam, resulting in a spatial resolution of 0.03mm.

The model MNHN 96.135 weighs 633.6g, which is 3.6g more than that recorded for The Sergio on the label in Figure 6. If the model is assumed to be solid, then we could determine an approximate density by using data from the three-dimensional scanning to estimate its volume. The volume estimate is 227cm³, giving a density of 2.8g/cm³.

This is on the low side for carbonados as their density ranges between 2.8 and 3.5 (Haggerty, 2014). Nevertheless, these measurements make the cast a good copy of The Sergio, with similar appearance, weight and apparent density.

How similar were the NHM models to the cast donated by Moissan? In Part One, we reported the dimensions of the NHM models as 88 x 82 x 62mm, but these were measured by hand with a ruler, a very imprecise method to measure an irregular object (Hansen et al., 2024). In our searches for information about the Sergio, no dimensions were found, apart from a rough estimate by Yawger, given as “approximately 3in. by 3in. by 3½in.” (1907, p.8). The photograph in Kunz’s (1904) report was labelled “actual size”, so we examined the relevant page of the book and measured the maximum width and height of the figure (in the same orientation as the Furniss image shown in Figure 1a) as 94mm and 87mm, respectively. But this is a two-dimensional representation and told us only that the models we had were in the right range.

We decided to compare the models directly. The NHM models were scanned in 3D using Creaform Go!Scan Spark; this is a handheld scanner which uses white light for structured light scanning. The resulting data was processed in VXElements v11 to create a ‘watertight’ mesh with no holes in the shape of the exterior of the models. From this mesh, 3D prints were made of each model to be carried (by authors RH and LR) to the MNHN, to meet the third author (FF). Visual comparison suggested that the NHM models were similar in size and shape to MNHN 96.135, the Moissan model.

As all three models were now digitally scanned, a more exact comparison could be made of their dimensions. Utilizing the software Meshmixer, author FF used the scan data to compare the three models and found them to be remarkably similar. Figure 9 shows an image of NHM model BM.1911,119 superimposed with MNHN 96.135. The deviations are shown in red and the models are almost identical in dimensions. Model BM1911,119 is 96.0 x 85.3 x 64.0mm. and model MNHN 96.135 is 94.2 x 89.2 x 66.2mm. The variations are roughly ±1mm in width and ±2mm in height and ±4mm in depth (average difference is 2.6mm).

These differences are minimal, given the age of the models and the techniques used to make them at that time. So, in spite of many tiny local differences in the surface details, all three casts are consistent. This implies that all three must have been made from moulds of the original carbonado, as the probability to end up with such similar data if they had not is very low. We can therefore state the dimensions of original The Sergio with considerable certainty. To the nearest mm, they are 96mm x 85mm x 64mm. This is a very comfortable ‘hand full’ as shown in Figure 10.

Having obtained 3D scanned data for all models, it was tempting to calculate the volumes and use the weight of The Sergio reported by Moissan (630g) to calculate its density. However, given the likely differences in casting methods, the irregular surface of The Sergio and the porosity of carbonado, scanned volumes are likely to be over-estimates because the porous surface cannot be perfectly copied in the mould. Further, even though the differences in dimensions averaged only 2.6mm, this translates to a significant difference in volume. Using Meshlab software, and rounding to the nearest cm³, the volumes for the NHM models were found to be 211cm³ and 208 cm³ for BM.1911,119 and BM.1985,MI9558, respectively, an average of 209.5cm³. The volume for MNHM96.135 was larger at 227cm³. Thus, the density estimate based on the NHM models is 3.0 and for the MNHN model it is 2.8. Both are within the low range for the density of carbonado (Haggerty, 2014) indicating that, as expected, the volumes are over-estimates.

The surface of the Moissan model was much more detailed than the surfaces of the NHM models, indicating that a slightly more accurate mould was used to cast the model. During our inspection, we found that a sound could be heard if the Moissan model was inverted, suggesting that it was not solid with a loose piece moving inside. Unlike the NHM models, where each had a small hole clearly indicating hollowness (described in Part One; Hansen et al., 2024), the Moissan model did not, so if it was hollow, such an imperfection had been repaired. This would be the usual procedure if, as seems likely, a lost wax method was used to cast MNHM 96.135.

Portable XRF analyses were carried out on MNHN 96.135 using a Bruker Tracer 5i (2023 model) working with a 50kV-4W Rh target X-ray source, The 8mm spit size diameter, operating in air, is able to detect in these conditions, from sodium (Z=11) to uranium (Z=92). It was internally calibrated by Bruker with a series of either oxides or alloys and further tested on various minerals, especially for sodium and magnesium. The spectrometer automatically selects the correct calibrations according to the data collected for 105 seconds. Two spots were selected on the cast, one black-to-silvery unabraded spot on the surface coating, and an orange-ish spot where the black coating had abraded. Both spectra are similar (Figure 11).

By far the most abundant element is copper (75% by weight), which is consistent with the orange colour that appears when the silver and black layers are degraded. Copper is accompanied by silver (14 wt%) and magnesium (4 wt%). The silver content is likely overestimated as it is present as a thin surface layer which promotes intense X-ray emission, even in the orange areas where it is not completely removed. The Kα peak for magnesium seems tiny compared to silver and copper but is consistent with this element being a minor component. Magnesium could also originate from a surface layer of unknown origin (surface alteration etc). Minor to trace amounts of aluminium and silicon (presumably from the mould), phosphorus, sulphur (probably related to silver chemisorption), calcium, titanium, chromium, manganese, iron, nickel, tin and platinum are also detected. In summary, the cast metal is predominantly copper, possibly alloyed with magnesium, as these materials were often used for industrial casting until the end of the 19th Century. The cast was then ‘painted’ with a layer of silver (to promote a pseudo-adamantine lustre in some areas) and graphite powder for the other, duller blackish areas.

Being able to compare the 3D scan data for all three models and view the Moissan model in person to compare with the 3D prints of the NHM models was invaluable. The extra detail on the surface supported our expectation that MNHN 96.135, the cast donated by Moissan, is a more accurate model. Conversely, the slightly less detailed surface, combined with the younger acquisition date suggest the NHM models are less accurate replicas. We assume that this may be because Gregory’s company had long sold models of diamonds, so it is very likely that he had reusable moulds.

Interestingly, Moissan’s cast of The Sergio was located in the same drawer and stored close to the cast of another significant diamond, the Cullinan. Discovered at the Premier mine in South Africa, on 26 January, 1905, it was named after Thomas Cullinan, the owner of the mine. Weighing 621.20g, the Cullinan was the largest gem quality rough diamond ever found. It was cut into nine large, and almost 100 smaller diamonds in Amsterdam, by Asscher and Co. Figure 12 shows the two models, side by side, and although the difference in the reported weight of the actual specimens is very small, less than 9g, The Sergio model looks considerably larger, probably attributable to carbonado’s porosity creating lower specific gravity (diamond has SG = 3.52, a difference in volume of over 10%), and possibly, the sparkle of the Cullinan!

Another Model of The Sergio

Further investigation into the records of the MNHN revealed more information about The Sergio’s visit to Paris. In a 1913 report, the then French Consul in Brazil, M. Paul Serre (who was also an Associate of the MNHN), wrote “before breaking this enormous piece of carbon that no museum agreed to buy, we took two casts of it. One of them is kept in Paris and the other at the Institut historique et géographique in Sào-Salvador [sic]“ (Serre, 1913, p.135; translation by the authors). Clearly, the cast kept in Paris is the donated Moissan model, and now we have discovered the existence of another model made from the original Sergio, which was probably given to the Instituto Geográfico e Histórico in Bahia de Salvador (Bahia, Brazil). Excited by this new discovery, one author (FF) tried via several methods to communicate with this Institute, but unfortunately, at the time of writing, there had been no response. As an aside, we note from this Institution’s website that after “130 years of uninterrupted operation, free of charge and open to the community, … the institution has been facing difficulties in maintaining its collection…[having had its] resources cut by the Department of Culture [which] represent 85% of the funding for its actions.” (Instituto Geográfico e Histórico da Bahia, 2024). We sincerely hope the model and the collections still exist and are able to be cared for, and future contact is possible.

Additionally, Moissan’s archives at the Académie des Sciences in Paris were searched by one of us (FF) in July 2024, attempting to determine more about how the mould and the casts were made. Alas, the archives do not reveal how, where, or by whom, the casts were made or the conditions of donation in 1896, as there is no exchange document with Alfred Lacroix about this matter. In fact, there is only Moissan’s request to the MNHN to supply kilograms (!!) of diamond-bearing “sands” (cascalho) from Brazil to dissolve in acids to obtain the insoluble fraction for his experiments (see Note 3). Clearly, Lacroix supplied these requested kilograms as the MNHN still keep a cascalho sample of 15 kilograms from Minas Gerais full of diamonds!

We were surprised to discover, just as we completed this manuscript, that the cast of The Sergio was not initially intended for donation to the MNHN. In October 1895, the Journal Officiel de la République Française reported that:

A cast will be made of it and deposited at the Muséum de Minéralogie; this remarkable sample, because of its great value, will be broken up to be used in fragments, as industry is particularly interested in black diamonds, which are harder than transparent diamonds, for drilling machines. (p.5813, translated by author FF).

By the time the cast was made, and Moissan had completed his study of it, instead of the Muséum de Minéralogie, the destination became the Muséum national d’histoire naturelle. Perhaps it was in the context of Lacroix’s assistance in finding material for his research that Moissan donated the cast to the MNHN.

Where Did Gregory’s Models Originate?

Some information comes from James R. Gregory himself. An ardent mineralogist, Gregory was interested in The Sergio and in December 1895, he published an article in The Mining Journal in which he described The Sergio as “Probably one of the largest pieces of diamond ever discovered was recently exhibited at the Academy of Science in Paris” (Gregory, 1895, p.1536). Gregory was a member of the Mineralogical Society of France, so he would have visited France often and likely knew Moissan. His detailed description of The Sergio suggests that he had seen the actual specimen:

It had a dark, brownish black colour, slightly polished on some of its angular surfaces, as is usual with examples of this variety of diamond. On being fresh broken the colour is of an ashen grey or somewhat fawn colour, and very uniform in texture, very nearly compact, or being only very slightly cellular; some specimens have bright glistening specks throughout, though this was not the case with this piece. (p.1536)

Gregory may have seen The Sergio in Paris, but this extract implies that he saw the specimen when it was “fresh broken”, and that must have occurred in London when the specimen belonged to Gulland. Both Gregory and Gulland were members of the Geological Society and being contemporaries, doubtless knew each other. Gregory would know that The Sergio had been sold to Gulland; perhaps he was even there when Gulland broke it up. Given that his company produced models, we would expect Gregory to be very interested in obtaining a mould of it before Gulland broke it up. We cannot know the full story of the making of the NHM models, but the Dennis photograph of The Sergio taken only days after it was found, and Moissan’s engraving, made two months later, are so consistent with the models in the NHM and the MNHN, that a mould of the original Sergio must have been the starting point for the Gregory models. Perhaps Gregory had access to the mould made by Moissan, but it seems more probable to us that he had a mould made when The Sergio was in Gulland’s possession and before it was broken.

An interesting aside is that Gregory “specialised in gems and meteorites and made a superb private collection of examples of the latter from several hundred falls” (Cooper, 2006, p.145). In his description of The Sergio, Gregory (1895) noted that carbonados were found in only a small area of Bahia in Brazil (carbonados had yet to be discovered in Central African Republic) and remarked that “As to its mode of origin, nothing seems to be known; no speculation on that subject has ever appeared” (p.1536). He continued:

This form of diamond is one of the forms of carbon discovered quite recently in certain meteoric irons, where, in addition, are also found white crystallised diamonds, and another strange crystallised form of carbon, to which the name of Cliftonite was given by Mr L Fletcher FRS and was described by him a year or two ago. (p.1536).

It is likely that Gregory was one of the earliest persons to link, at least implicitly, carbonados with meteorites and an extraterrestrial origin.

Final Word

For lovers of minerals and gemstones, the permanent loss of The Sergio is a sad tale. Billions of years old, yet after barely a few months in human hands, The Sergio was gone forever, simply for financial gain. It remains the largest carbonado the world has ever seen, but few people in the world actually saw it! Its tale, as told through historical sources in the English and French languages that were often contradictory or incomplete, has traced what happened to The Sergio after it was found. Thus, the models in the NHM and the MNHN have not only elicited this story of its life, they have enabled us to provide, for the first time, a remarkably accurate statement of The Sergio’s dimensions: 96mm x 85mm x 64mm. The models remain to remind us of what was, in Gregory’s words, “a remarkable diamond”.

Notes

Note 1. Henry Watson Furniss (1868-1955) was born in Brooklyn, New York, to well-educated African-American parents. He became both a doctor and a successful diplomat. In 1897, he was appointed United States Consul to the Brazilian state of Bahia, and in 1905, appointed as US Minister to Haiti until 1913. Later, he moved with his family to Connecticut where he established a medical practice (Erickson, 2022). While in Brazil, Furniss learned Portuguese and took a great interest in health issues, something that is evident in his description of the mining industry in Brazil. He wrote a number of Consular reports to the US, and one of these, No. 1423, dated 20 August, 1902 is referred to extensively by many authors, including Kunz (1904), in writing about diamonds and carbonados in Brazil. Later, after leaving Brazil, he wrote an article published in Popular Science Monthly (Furniss, 1906) summarising the industry. Despite an extensive search, we have not been able to locate a copy of the 1902 Consular report. In this 1902 report, Furniss stated that The Sergio was broken up in Paris and this erroneous information was widely copied. In his 1906 discussion of Brazilian diamonds and carbons, Furniss acknowledged this error, correctly stating that it was broken up in London, as can be seen in the quote earlier.

Note 2. James Ker Gulland (1835-1921) was born in Scotland and received practical training in Glasgow at ironworks. He then turned to underground mining. He set up his own self-named company in 1877 and patented a drill design in 1878. Gulland became a successful consulting engineer; his diamond rock drill was supplied to many of the principal governments (Grace’s Guide, 2023a). He established the Diamond Drill Co, Boring Contractors and Diamond Drill Makers, at 8 Victoria Street, Westminster, London, S.W. By 1914 the company had world-wide connections and claimed to be “the oldest firm in the business and the largest Importers of Black Diamonds for boring in the world” (Grace’s Guide, 2023b). J. K. Gulland became a Member of the Institution of Mechanical Engineers in 1884, and importantly for our story, he was a Fellow of the Geological Society and likely knew James R. Gregory.

Note 3. Henri Moissan (1852-1907) was awarded a Nobel Prize in Chemistry in 1906 for his work on fluorine and its compounds (Fechete, 2016), so one may wonder at his interest in diamonds and carbonados. Moissan’s strong interest in chemistry led to his invention of a powerful electric furnace and the discovery and synthesis of many chemical compounds. He was a student of Edmond Frémy (1814-1894), as was his colleague Auguste Verneuil (1856-1913), both of whom worked toward the synthesis of rubies (as well as sapphires) at the MNHN from the second part of the 19th Century. For them, synthesizing ‘gemstones’ was not a goal for jewellery but for industry: synthetic rubies made possible the manufacture of large quantities of reliable and accurate pocket clocks that every worker could afford (see Evans, 2020, for a history of synthesizing rubies). Similarly, from around 1890, Moissan continued this type of research on diamonds for their use in heavy industry (drilling, tunnels, etc.). To build his knowledge of the nature of diamonds and how they might be formed, he asked Professor Alfred Lacroix (1863-1948), then the scientist-in-charge of the mineralogy collections at the MNHN, about all types of diamonds encountered in nature, particularly in South Africa, Russia and Brazil. At that time, the MNHN collection in Paris contained very minor fragments of carbonados, as well as one significant 329 carat specimen from Bahia (acquired in 1848 and still possessed). This could not be analysed without breaking it, thus Moissan had to search among Paris’ diamond dealers and elsewhere to obtain material for his experiments.

The Musée de Minéralogie (Mineralogy Museum) of the École des Mines de Paris (School of Mines in Paris) has a collection of historic fragments of the Canyon Diablo ferronickel meteorite, and Charles Friedel (1832-1899), scientist-in-charge of the museum at the time, confirmed the presence of micro-diamonds in a slice of this meteorite (Friedel, 1892). Moissan obtained a small piece of this meteorite and was able to observe, in situ, a very small boort diamond and realized they must have formed within metal under very high temperatures (Wisniak, 2002). Moissan developed a high-temperature furnace up to 500°C to calcinate all types of diamonds, including carbonados, to investigate their residual ashes for potential clues about their formation. (See Sella, 2013, for an image and description of the furnace.)

Moissan announced that diamond had been synthesized at the Academy of Sciences on 6 February 1893, and made a demonstration about this on 17 May, 1893 (Moissan, 1893). However, as Wisniak (2002) and Evans (2022) have explained, there has been no proof that the tiny crystals created by Moissan were indeed diamonds; it is more likely they were silicon carbide or carborundum (Sella, 2013). Aptly, the mineral silicon carbide has been named Moissanite. It was his pursuit to synthesise diamonds that made Moissan so interested in The Sergio when it arrived in Paris, and to have a model of it to study and later donate to the MNHN.

References

Anonymous., 1902. Brazilian diamonds and carbons. The Journal of the Society for Arts, 51(2608), pp.928-930. [online] Available at https://www.jstor.org/stable/41335670> [Accessed 21 September 2022].

Anonymous., 1931. The King of the black diamonds. Popular Mechanics Magazine, 56(6) pp.984-987. [online] Available at <https://archive.org/details/PopularMechanics1931/Popular_Mechanics_12_1931/page/n167/mode/2up> [Accessed 24 September 2022].

Baszanger, J., 1909a. Carbons for diamond drills. To the editor of The Mining Journal. The Mining Journal, 87(3875), pp.333-334.

Baszanger, J., 1909b. Carbons (black diamonds) for diamond drills. The Mining Journal, 87(3867), pp.7-8.

Brazil. Commissão, Louisiana Purchase Exposição, 1904, and F. M. De Souza Aguiar. Brazil at the Louisiana purchase exposition. St. Louis, Saml. F. Myerson Ptg. Co. [online] Available at <https://lccn.loc.gov/18006587> [Accessed 16 January 2022].

Comptes-rendus de l’Académie des Sciences (Proceedings of the Academy of Sciences), 1895. Séance du Lundi 25 September 1895, Mémoires et Communications, Chimie Minérale – sur un échantillon de carbon noir du Brésil (Session of Monday September 25, 1895, Memoirs and Communications, Mineral Chemistry – on a sample of black carbon from Brazil). 121(July-December), p.62. [online] Available at https://gallica.bnf.fr/ark:/12148/bpt6k3077p [Accessed 22 October 2024].

Cooper M., 2006. Robbing the sparry garniture. Mineralogical Record, Inc. Tucson, AZ.

Erickson, N., 2022. Fade to White: The racial journey of Henry Watson Furniss. [online] Available at <https://www.imhm.org/blog/12682952> [Accessed 4 April 2023].

Evans, J., 2020. The history of synthetic ruby. London, Lustre Gemmology Limited.

Evans, J., 2022. The early history of synthetic diamond. The Australian Gemmologist, 28(2), pp.90-95.

Farges, F., Sucher, S., Horovitz, H. and Fourcault, J.-M., 2009. The French Blue and the Hope: New data from the discovery of a historical lead cast. Gems & Gemology, 45(1), pp.4-19.

Farges F., 2014. Les grands diamants de la Couronne de François I à Louis XVI (The great diamonds of the Crown of François I to Louis XVI). Versalia, 16, pp.55-79.

Farges, F., Panczer, G., Benbalagh, N. and Riondet, G., 2015. The Grand Sapphire of Louis XIV and The Ruspoli Sapphire. Gems & Gemology, 51(4), pp.392-409.

Fechete, I., 2016. Ferdinand Frédéric Henri Moissan: The first French Nobel Prize winner in chemistry or nec pluribus impar. Comptes Rendus Chimie, 19, pp.1027-1032. https://doi.org/10.1016/j.crci.2016.06.005 [online] Available at <https://www.sciencedirect.com/science/article/pii/S1631074816301631> [Accessed 3 November 2023].

Friedel, C. 1892. Sur l’existence du diamant dans le fer météorique de Canon Diablo. Bulletin de Minéralogie,15(8-9), pp.258-263. [online] Available at <https://www.persee.fr/doc/bulmi_0366-3248_1892_num_15_8_2296> [Accessed 3 November 2023].

Furniss, H.G., 1906. Diamonds and carbons in Brazil. Popular Science Monthly, 69(9), pp.272-280. [online] Available at <https:// https://en.wikisource.org/wiki/Popular_Science_Monthly/Volume_69/September_1906/Diamonds_and_Carbons_in_Brazil> [Accessed 14 January 2022].

Grace’s Guide., 2023a. Grace’s Guide to British Industrial History, James Ker Gulland. [online] Available at <https://www.gracesguide.co.uk/James_Ker_Gulland> [Accessed 11 May 2023].

Grace’s Guide., 2023b. Grace’s Guide to British Industrial History, 1914, Who’s Who in Business. [online] Available at <https://www.gracesguide.co.uk/1914_Who%27s_Who_in_Business:_Company_G> [Accessed 11 May 2023].

Gregory, J., 1895. A remarkable diamond. The Mining Journal, Railway and Commercial Gazette, 65(3147), p.1536.

Gulland, J., 1902. Brazilian carbons. The Journal of the Society for Arts, 51(2609), p.22. [online] Available at <https://www.jstor.org/stable/41335675> [Accessed 21 September 2022].

Gulland, J., 1909. Carbons for diamond drills. To the editor of The Mining Journal. The Mining Journal, 87(3873), p.258.

Haggerty, S., 2014. Carbonado: Physical and chemical properties, a critical evaluation of proposed origins, and a revised genetic model. Earth-Science Reviews, 130, pp.49-72.

Hansen, R. and Rennie, L., 2022. Black diamonds: A reflective overview. The Australian Gemmologist, 28(1), pp.26-36.

Hansen, R., Rennie, L., Burgio, L., Montgomery, W. and Farges, F., 2024. The Sergio: An exploration of the world’s largest carbonado, part 1. The Australian Gemmologist, 28(5), pp.268-276.

Herold, M., 2013. The black diamonds of Bahia (carbonados) and
the building of Euro-America: A half- century supply monopoly
(1880s-1930s). Commodities of Empire Working Paper No.21.
[online] Available at <https://scholars.unh.edu/cgi/viewcontent.cgi?article=1013&context=econ_facpub > [Accessed 8 October 2021].

Instituto Geográfico e Histórico da Bahia. 2024, September 5. Threatened Heritage. [online] Available at <https://www.ighb.org.br/> [Accessed 26 September 2024].

Kunz, G., 1896. Precious stones. In D. T. Day ed., Seventh Annual Report of the United States Part III: Mineral resources of the United States, pp.895-926. Washington: Government Printing Office.

Kunz, G., 1904. Precious stones. In D. T. Day ed., Mineral resources of the United States, Calendar Year 1902, pp.813-865. Washington: Government Printing Office.

Journal Officiel de la République Française (Official Journal of the French Republic), 1895. Académie des Sciences, Séance du 23 septembre (Academy of Sciences, Session of September 23). 27(268 October 3), p.5813. [online] Available at https://gallica.bnf.fr/ark:/12148/bpt6k6267391c?rk=42918;4# [Accessed 21 October 2024].

Moissan, H. 1893. Le diamant: conférence faite à la Société des amis de la science le 17 mai 1893/ par M. Henri Moissan (The diamond: conference given to the Society of Friends of the science on May 17, 1893 by Mr. Henri Moissan. Bibliothèque national de France. [online] Available at <https://gallica.bnf.fr/ark:/12148/bpt6k91874t/> [Accessed 3 November 2023].

Moissan, H., 1895. Carbon noir du Brésil. La Nature Revue des Sciences, 32(1166), p.304.

Sella, A., 2013. Moissan’s furnace: The arc of extreme chemistry. Chemistry World-Opinion 2 June 2013. [online] Available at <https://www.chemistryworld.com/opinion/moissans-furnace/6232.article> [Accessed 3 November 2023].

Serre, P., 1913. Au pays du carbone amorphe (Carbonato) (In the country of amorphous carbon (carbonato). Bulletin du Muséum national d’histoire naturelle pp.133-136. [online] Available at <https://www.biodiversitylibrary.org/page/5051898#page/155/mode/1up> [Accessed 26 September 2024].

Special Correspondent, 1895. Foreign news. Pharmaceutical Journal, September 28, 1895, p. 278. [online] Available at <https://archive.org/details/s1317id1378079/page/278/mode/2up> [Accessed 24 September 2022].

Svisero, D., Shigley, J. and Weldon, R., 2017. Brazilian diamonds: a historical and recent perspective. Gems & Gemology, 53(1), pp.2-33. [online] Available at <http://dx.doi.org/10.5741/GEMS.53.1.2> [Accessed 21 October 2021].

Wisniak, J., 2002. Henri Moissan: The discoverer of fluorine. Educación Química, 13(4), pp.267-274. [online] Available at <https://www.revistas.unam.mx/index.php/req/article/view/66285/58196> [Accessed 3 November 2023].

Yawger, I., 1907. Carbon: a remarkable mineral, from The Mining World [online] Available at <https://babel.hathitrust.org/cgi/pt?id=njp.32101051619235&view=1up&seq=1 [Accessed 19 December 2022].