Docosane (C22H46, 310 MW, 20.70 at Penn)

Docosane is an alkane hydrocarbon found in high concentrations in flower extracts, notably from the *Tilia* genus, or linden, which according to palynological studies grew on Crete at least during the earlier LBA [35–37]. Its identification is based not only on the solitary presence of the biomarker docosane, but also by careful consideration of the presence or absence of companion diagnostic compounds (e.g., isopulegol, a monoterpene alcohol that when appearing concurrently with docosane would instead suggest *Lilium*, or the lily plant [38]). These holistic phytochemical considerations make *Tilia* the most likely candidate to explain the concentration of docosane, a conclusion verified by the matching chromatographic signature produced from an ethnobotanical *Tilia* sample collected on Crete. ORA evidence for *Tilia* was recovered only from the Mochlos samples and absent from Tourloti.

In addition to its use as a treatment for atherosclerosis, *Tilia* also has antimicrobial properties [39], but the double-flowered blossom from this genus is best known for its aromatic compounds used to make perfumes [19,40]. Its fragrance, known primarily today by the industry as the heart-note infusion "tilleul" and used in perfumes such as Bourjois' "Soir de Paris" and Jesus Del Pozo's "Duende", is often described as reminiscent of citrus or honeysuckle, and as such could have also served as a head note. It is notable that Bourjois' "Soir de Paris" employs both storax and tilleul as complementary base and heart notes, respectively [19], echoing the pairing used in the Mochlos perfumed oil [4].

#### *3.7. Summary of ORA Results*

The collective chemical composition of the residues from the Mochlos and Tourloti objects suggests that both contained (or were used in the production of) a sophisticated olive oil-based perfume infused with *Cistus* labdanum amber (base and heart notes), *Liquidambar* storax balsam (base and heart notes), *Artemisia* camphor (a head note), and *Coriandrum* essential oil (a head note); all are botanicals that would have been locally available in East Crete or nearby regions such as Rhodes. The consistency between the two perfume "recipes"—both in regard to their ingredients and the techniques of their manufacture—can also be said to be a meaningful indicator of cultural continuity over the course of several centuries, where luxury items offered one means by which to negotiate status both during and after the demise of the Cretan Bronze Age palatial centers [1,41].

The results highlight the exciting potential for organic residue studies to illustrate both products and processes, here particularly related to Bronze Age perfume manufacture. The use of *Coriandrum* in both the Mochlos and Tourloti vessels is particularly meaningful in that it likely reflects the practice of stypsis, the treatment of the base oil with an astringent in a two-stage process to make it more receptive to infusion with aromatic compounds [23] (with the added bonus of a complementary head note). This technique was employed in the perfumeries of the Late Bronze Mycenaean palaces, as documented at Pylos in Linear B tablets Un 267 and Vn 130 [23,42], and is now archaeometrically confirmed [1,4]. In this vein, it is significant that the linalool is present in the completed product—the perfumed oil present in both the Mochlos amphora and the Tourloti stirrup jar—but not in the Mochlos vat. Its absence might simply reflect the rapid volatilization of the compound as it was exposed to air in the open vat, but this seems unlikely given the high boiling point of this compound—thus its lower relative volatility—and furthermore given that the Tourloti jar had similarly been exposed to air for over a century, and yet contained a relative abundance (RA) of linalool nearly identical to the Mochlos amphora. Linalool's absence from the vat may instead show that the vat was not employed for stypsis, but instead used in some other aspect of the perfume manufacturing process, perhaps the infusion of

the primary constituent aromatics. The comparable quantities of linalool in the Mochlos and Tourloti jars—in the neighborhood of 15% RA for both—may reflect the absorption potential of olive oil in relation to linalool, or infer something about the proportions of the ancient recipe. While the practice of stypsis needs to be explored more fully, it may also be that biomarkers for coriander could be used to distinguish perfumed oils from other lipid-based organic commodities in the Bronze Age Aegean world and beyond. The comprehensive ORA results, thus, offer unique insight into Bronze Age *chaîne opératoire*, elements of which could have implications for further ORA studies.

Additionally of note is the fact that the earlier Mochlos vessels contained one additional ingredient, tilleul from *Tilia* blossoms, the presence of which was marked by significant quantities of the compound docosane. Additional vessels containing docosane, including cooking pots, were found immediately outside the Vat Room in the same Mochlos workshop and were probably used for maceration, or heated steeping, of the *Tilia* blossoms [4]. The absence of tilleul in the later Tourloti perfume may reflect a shift in the botanical profile of the region from LM IB to IIIC (ca. 1500–1175 B.C.), and in particular a move away from preferential utilization of the plant in the wake of political and social change in East Crete in the intervening centuries. However, it is significant that the LM III-EIA pollen cores from the island echo the same pattern as the perfumes; *Tilia* disappears at the end of the LBA. The disappearance of *Tilia* from the palynological record was, in fact, specifically connected with climatic shifts at the end of the Bronze Age that diminished the tree's natural habitat [36]. It is, therefore, possible that the absence of tilleul from the Tourloti perfume could be a direct reflection of climate change (e.g., aridification) that occurred at the end of the Late Bronze Age [43–48].

The results of our ORA study, thus, shed new light on elite behavior, manufacturing processes, and cultural patterns at the end of the Bronze Age, while also hinting at ORA's exciting and underdeveloped potential as a tool for paleoecological reconstruction.

#### **4. Discussion**

The presence of complex perfumed oils with multiple scent layers, or notes, in the LM IB Mochlos and LM IIIC Tourloti vessels offers a window into the structural and market forces that shaped demand and production of these expensive, value-added commodities during the Bronze Age [1,4]. From an archaeometric perspective, however, the comparative Tourloti and Mochlos results are in many ways even more exciting. They are particularly important in that they demonstrate the value and viability of ORA samples taken from legacy objects, even those for which the "chain of custody" cannot be completely known. They moreover illuminate patterns in degradation processes and sample variation that in turn have potentially significant implications for future extractive and interpretive methodologies, upon which we offer a few preliminary comments.

#### *4.1. Compound Expressions in Legacy and Newly Excavated Objects*

Diagnostic legacy compounds, which are typically lipids, suffer from long-term autoxidation processes and therein degradation, recognition of which is essential for successful interpretation [49,50]. Oleic acid, for example, the major diagnostic and prevalent fatty acid of olive oil, degrades into azelaic acid and nonanoic acid under oxidizing conditions, a process noted in several artifact residue studies [1,20,21]. Our Tourloti case study demonstrates this lipid oxidation in stark terms: The residues from the freshly excavated Mochlos objects preserved substantial quantities of oleic acid and no azelaic acid. By contrast, in the legacy Tourloti jar the degradation is virtually total, and azelaic acid constitutes the largest peak across all chromatograms.

This same phenomenon was detected in early ARCHEM data from a sixth century B.C. olive press installation from Azoria in East Crete. This installation was discovered in the larger west room of a two-room structure (D300) from the Service Building, excavated in 2006 under the direction of D. Haggis and M. Mook [51]. D300 contained overwhelming archaeological evidence for the processing and pressing of olives: a press bench, press beam sockets, numerous weights, and a basin, found together with substantial quantities of olive pits—including seeds from press-cake. The site was

eventually abandoned and destroyed, although apparently not before valuables had been removed [52]. Whereas ORA samples taken from modern olive oil jars and presses on Crete have oleic acid as their predominant compound, the ORA samples taken from the basin, querns, and surrounding soil from the olive press installation at Azoria were dominated by overwhelming evidence for azelaic acid. Indeed, the pressing process during the lifetime of the room would have presumably ensured that the incidental detritus and remnant oils from olives were subjected to constant, unmitigated exposure to air, offering continuous opportunity for natural aerobic processes such as oxidation and degradation.

As a monounsaturated fatty acid, the three-phase overall mechanism (i.e., 1.) initiation, or formation of free radicals; 2.) propagation, or free-radical chain reactions; and 3.) termination, or formation of nonradical products [49,53]) of oleic acid oxidation and therein the rate of degradation, is affected by successive reactions, of which azelaic acid is a dominant terminal nonradical product [1,20]. The OpenARCHEM project is currently conducting a series of fatty acid degradation studies at the Wesleyan University Department of Chemistry in order to better characterize these mechanisms, the nature and timing of which have yet to be comprehensively documented. As we come to understand the processes, which facilitate and catalyze oleic acid degradation, it is possible that the product distribution ratios between oleic and azelaic, and the peroxide value of various intermediary hydroperoxides [53], could eventually serve as meaningful indicators of the length of time for which an object was exposed in antiquity and therein shed light on site use or abandonment processes. It may be possible, for example, that the better preservation of oleic acid in the Mochlos perfume paraphernalia may be due in part to the fact that the space was quickly sealed by the unexpected collapse of the building during the powerful LM IB (ca. 1500 B.C.) earthquake that hit the town [3,4]. These mechanisms clearly demand systematic study, but may eventually also be a useful means by which one could determine something about the history of legacy vessels when the object biography is otherwise opaque.
