Our research in the French National Archives (Archives nationales, AN, Minutier MC/ET/XXXIX/619) led us to the will of this little-known crystallographer. This document informed us of the existence of crystal models developed by this pioneer of the science of crystals, which he indicated as constituting the most precious legacy of his collection. Forgotten up until now, this legacy was found, identified and recollected (MIN000-5020) within the mineralogy collections of the Muséum: these hundreds of clay casts of unknown origin and age correspond in fact to the oldest models (in compacted clay of the finest texture) of Romé de l’Isle’s crystals known to date and published in the opus of 1772 (Farges et al. 2015). Among them, No. 71 (Figure 3.2) corresponds, according to Romé, to the crystallographic model of the Grand Saphir, which he considered natural. As for the original, declared “uncut” in 1789 (AN, O1 3362) and therefore not very valuable, it found its “rough” acolytes at the Muséum under the tutelage of Daubenton in 1796, who understood well that he was making a bargain on behalf of his museum by acquiring such an exceptional gem. Even though Romé questioned his conclusion in 1783 in the second edition of his Essai de Cristallographie, he continued to suspect the Grand Saphir of being natural in a 1787 letter to Delamétherie (1743–1817), another French mineralogist of that period who sponsored Romé. Indeed, it was in fact a gem, probably polished by a Mughal hand in 17th-century India. Its most original faceting had fooled the experts of the Enlightenment, but not those of the Grand Siècle1, a hundred years earlier... .
Figure 3.2. Gems and models. © F. Farges/MNHN
COMMENT ON FIGURE 3.2.– (a) Louis XIV’s Grand Saphir (MNHN inv. A.67); (b) ceramic model no. 71 by Romé de l’Isle (circa 1772) representing the Grand Saphir as a natural crystal of “adamantine spar” (corundum); (c) rhinestone model of the Ruspoli sapphire also found in the MNHN collections (inv. 50.167) and given by the same Parisian gem cutter as the model of the blue diamond (see Figure 3.1(a)): a certain M. Achard. The scale is 1 cm.
More recently, it was said that this sapphire should be named the Ruspoli, after the eponymous Roman prince who is said to have owned it before Louis XIV (see references in Farges et al. 2015). The mechanics of storytelling are so effective that this sapphire is actually better known by this nickname than by its original name Grand Saphir. In fact, a misreading of a paragraph by Mr Barbot (see Farges et al. 2015) was at the origin of the confusion between the Grand Saphir and the Ruspoli, which are actually two different gems. This other sapphire made the headlines – from both a society and legal point of view – in 1813 for obscure reasons, due to the mischievous baiting that jewelers liked to throw back and forth. In fact, the Ruspoli was rediscovered at the Muséum (inv. 50.167), at least its historical replica in cobalt blue rhinestones, dating back to the 1820s (Figure 3.2). Its faceting is largely dissimilar to that of the Grand Saphir. The Ruspoli was still traceable until 1952 when it was likely purchased by Cartier in New York (Farges et al. 2015). We do not know what has happened to this gem since then.
For its part, Louis XIV’s Grand Saphir remains strangely absent from the MNHN’s brand image, even though it is certainly the most precious piece in its collections: for example, consider the popularity of the Hope diamond, the “muse” of the Smithsonian Institution in Washington and, more generally, of the Americans who recognize in this blue diamond no less than their own Mona Lisa. Indeed, some semi-confidential gossip in the jewelry world has long suspected the Hope diamond of being Louis XIV’s blue diamond, cut by thieves or their collectors. Without proof, the story will remain as such. France quickly forgot its masterpiece since no painting of this French jewel is known. Until 2007 that is, when we found – in the collections of the Muséum national d’Histoire naturelle – the only cast of this gem stolen in 1792 (Figure 3.1(a)).
3.2. A scientific investigation of color
The first investigation consisted of deciding between the two opponents claiming to be the representation of the stolen French Blue diamond. On the one hand, we had the Russian clan which, through the Terestchenko diamond of 42.92 carats, claimed prestigious royal paternity by its ovoid shape. On the other hand, the American clan had the Hope diamond of 45.52 carats. A laser scan of the lead model was performed to obtain a 3D mesh of thousands of polygons of the lead cast of the French Blue diamond to find the truth of these two hypotheses. Numerical methods such as edge collapse decimation enabled the restoration of the original faceting of the gem. This result, when compared to the two competing representations, was clear: the Terestchenko was too long and not wide enough. The American won hands down (Figure 3.1(d)). The thieves of 1792 thus savagely recut the three main corners of the French triangular diamond to give it the present ovoid shape of the Hope diamond.
Then arose the question of the color of this mythical gem. Confusion reigned: the diamond was described as “violet” in the royal inventories of 1691, as “sapphire blue” in 1787 and then as “light blue” in 1813 (Farges et al. 2012). The study began with the Hope diamond, which was removed from its Cartier necklace for the occasion. An optic oddity – never noted or interpreted before – was revealed as I placed the mythical diamond on a light background: in the center of the gem, where it is thickest, the object became very light in color, which contrasted with the rest of the volume, which was of a navy blue almost black. It became obvious that the gem behaved like a lightly colored parallel prism (Figure 3.1(c)), which no one had noticed. On the other hand, at the periphery, the diamond appeared dark even though it is thinner. To understand this curious anomaly, an optical absorption study was performed on the gem (Figure 3.3).
3.3. The digital decoding of the creative genius of the royal gem cutter
Considerable ab initio calculations of the color of the Hope diamond were undertaken to elucidate the anomaly (Farges et al. 2012). For this purpose, the theoretical dielectric function (𝜓) of a diamond was calculated by considering a doping of its cubic atomic structure with boron atoms, present in trace amounts in the carbon atomic structure of this mineral and supposedly the source of its blue color (Fritsch 1998). We then solved the Bethe–Salpeter equation (Farges et al. 2012) that quantified the interaction between light and this mineral to determine its absorption spectrum between ultraviolet and near-infrared (400–900 nm). In summary, these complex calculations allowed us to reproduce and therefore understand the color anomaly of this diamond. In fact, the Hope diamond was not dark blue as it appeared at first sight, but pale blue! The calculation was able to show that its faceting amplified the pale color of the mineral constituting the Hope diamond to the point where the latter appeared blue (almost black) to us (Figure 3.3).
Figure 3.3. (a) Results of ab initio calculations on the color of the Hope diamond (dashed line) compared to the experiment (solid line) and photo-realistic simulation of this diamond reproducing the color contrast at the center of the gem (compare with Figure 3.1(a)); (b) photo-realistic simulation of this diamond without its cushion-faceting. Images © F. Farges/MNHN
As the faceting of the Hope diamond differed drastically from that of its ancestor, we then had to recalculate