BABESCH 92 (2017), 187-208. doi: 10.2143/BAB.92.0.3242694

The Lady of the Simpelveld Sarcophagus: an Osteo- Archaeological Approach

Andrea Waters-Rist & Ruurd Halbertsma & Karen Jeneson

Abstract

The Roman-period burial chest from Simpelveld has captured the interest of archaeologists and the public since its discovery in 1930. Substantial looting has disturbed the original contents, but there remains some cremated human material. Despite the incomplete and fragmented nature of these remains, this paper demonstrates that through detailed and advanced analyses much can be learned about the person buried in the chest. Macroscopic and microscopic analyses are undertaken to reconstruct variables about the cremation process, estimate sex and age-at-death, and document pathological conditions. The results reveal the Simpelveld chest’s occupant was a middle-aged (35-49 years) female, free from pathological lesions or skeletal anomalies, most notably lacking any signs of osteoarthritis. This osteobiographic profile largely matches the representation of the lady on the inside of the chest, suggesting it was made specifically for her and, with the lack of osteoarthritis, that she probably lived a privileged life. This study improves our understanding of this lady and Roman-period villa life in the south of the in general.

INTRODUCTION the decoration on the walls was not damaged and part of the contents escaped the attention of The so-called ‘Sarcophagus of Simpelveld’ is one of the grave robbers. A rough edged hole on one of the most intriguing monuments in the Roman the short sides is a trace of the pillaging. Department of the National Museum of Antiquities On the inside most of the decorated space is in Leiden (figs 1-3).1 Its decoration on the inside of reserved for the representation of the deceased: a the chest, the representation of the reclining woman woman, stretched out on a Roman lectus, holding and the precious gifts have led to a continuous an object in her left hand, probably a handkerchief. stream of publications, which started soon after its The hairdo is reminiscent of portraits of Faustina discovery near the village of Simpelveld (province Minor, the wife of emperor Marcus Aurelius, which of ) in 1930.2 A hitherto rather neglected gives a dating criterion for the years 160-180.3 She aspect of the sarcophagus’ contents were the cre- is clad in a Gallic tunica, with long sleeves, and a mated human remains. This article will try to derive mantle, which is draped over her right shoulder. as much information as possible from the human Earlier interpretations described her as participat- material, and will engage in discourse that brings ing in a funerary banquet, which is a very common to the forefront the value of osteoarchaeological scene on Roman gravestones. But in his 1997 article research in furthering our understanding of life in Frank Zinn astutely pointed out that the funerary the southern part of the Netherlands during the banquet belongs almost exclusively to the realm Roman period. To place this research in context, of men. If women are present, they are sitting and first a brief description of the sarcophagus, its reliefs not reclining. Zinn sees better parallels in Italian and its contents will be given. funerary sculptures, where reclining women are The chest is made from sandstone and meas- represented from Etruscan times onwards.4 The ures 2.40 m x 1.05 m x 0.76 m on the outside. The ‘lady of Simpelveld’ is surrounded by sculpted internal ‘funeral chamber’ has been cut out from furniture and architectural elements. When we the massive block and is decorated with repre- take a clockwise tour of the interior, we encoun- sentations of furniture and architectural elements. ter the following elements. The sarcophagus was originally covered with To the right of the lady a wicker chair is placed, two sandstone slabs, which were fastened to the with a high, curved back and a cushion on its main body by metal clamps. Probably during the seat. In the Germanic and Gallic provinces these Germanic invasions in this part of the Roman ‘cathedra chairs’ appear very often, for example Empire in the 3rd and 4th centuries CE, the sar- on funerary reliefs. Next to the chair stands a cophagus was laid bare and pillaged. Fortunately chest on a pedestal. The key lock and the handle

187 Fig. 1. The Simpelveld Sarcophagus.

Fig. 2. The Simpelveld Sarcophagus. Interior with carved reliefs.

188 Fig. 3. The Simpelveld Sarcophagus. Tableware and furniture.

pod is depicted, with legs decorated with lion’s heads. Such a mensa delphica could be made of mar- ble, silver, bronze or wood.7 The relative thickness of the legs may hint at the latter material. In front of the table an iron knife has attached itself to the floor. The rust stains on the table suggest that the knife originally was placed on the table top. Next to the mensa delphica another sideboard is standing. On the ground stand two tankards. On the first shelf two ‘Hemmoorer Eimer’ (situlas) are depicted. Above these situlas four drinking cups are stand- ing. All these vessels find close parallels in Provin- cial-Roman archaeology.8 Another luxury item is the next piece of furniture, a cabinet with two doors, sculpted with an eye for details. Next to this closet a series of niches of different sizes is represented. These do not seem to represent elements from real architecture, as observed by Zinn.9 His suggestion is that these niches were originally used to display the funerary gifts and the urn with the cremated remains of the deceased. The last architectural ele- ment is the representation of a house, with a smaller annexe. Although attempts have been made to link these buildings with real architecture, a general impression of ‘grand architecture’ is the most likely explanation for this section.10 In conclusion, we can say that the lady of Simpelveld is surrounded by the best tableware and interior design a villa owner living in the 2nd or 3rd century CE could wish for. Fig. 4. Contents of the Simpelveld Sarcophagus. The objects which were found in the chest also give an impression of wealth (fig. 4): luxurious are clearly rendered. In a homely context such an vessels of glass and pottery, golden jewellery, a arca would stand in a prominent place of the stilus (an indication of the literacy of the lady), a house, containing valuable documents and money.5 silver mirror, and domestic items such as a knife Around the corner, facing the lady, there is a low and a pair of scissors. The most impressive of these side table, on which three glass vessels are stand- funerary gifts is a golden dodecagonal ring, made ing, of common types: two cilindrical bottles, and a in the opus interrasile technique, with decorations in rectangular one.6 Next to this table a luxurious tri- three zones. The outer zones show a row of pelta-

189 like elements. The middle one bears an inscription, the condition of the deceased individual prior to which is framed by heart-shaped leaves: IVNONI cremation. Variables such as the temperature, MEAE, ‘to my Iuno’. This inscription places the ring intensity, and duration of the fire, the rapidness and its bearer in the realm of Iuno, the female coun- with which water gets driven out of heated bone, terpart of a man’s Genius. The goddess Iuno is the as well as variation between different bones and protectress of marriage and it is an attractive individuals determine the extent to which bones thought that this piece of jewellery once served as will shrink, deform, warp, crack and fragment.13 the lady’s wedding ring. Almost exact parallels of The chemical composition of bone changes in a this ring (although with different inscriptions) have predictable fashion in response to increasing heat been found in England, and can be dated in the 2nd exposure.14 As such, the temperature of the crema- century CE.11 The Faustina Minor type hairdo of the tion (sometimes referred to as the combustion lady of Simpelveld in combination with the paral- degree) can be inferred by looking at the colour, lels for the opus interrasile ring suggest a date for the texture, and deformation of the bones (fig. 5). For sarcophagus in the third quarter of the 2nd century example, only above about 650°C will bone take on CE, which is in line with the fact that the 2nd and a white ‘calcined’ appearance.15 Another example early 3rd centuries CE were a prosperous period for involves crack patterns with cracks oriented cross- the villa landscapes in the south of the Nether- wise and parallel to the long axis of the bone only lands.12 As such, the sarcophagus of Simpelveld is appearing at temperatures of 750°C or higher.16 As a testimony for the wealth and luxurious taste of well, certain fissures and fracture types (e.g. thumb- the richest inhabitants of the rural areas in this part nail fractures) only appear if bones are burned of Germania Inferior. when they are still covered in tissue.17 The extent with which bone shrinks in size is also correlated SKELETAL MATERIALS with heating temperature. In calcined bone, shrink- age can range from 0 to 30%.18 The average shrink- A total of 386 human bone fragments are present, age of bone heated from 750 to 1000°C is 12 to with 281 identifiable to some degree (table 1). The 14%.19 It should be kept in mind that a single cre- cumulative weight of this material is 273.8 grams. mation can show a range of heating temperatures, No dental remains were recovered. As detailed in depending upon the nature of the fire source (e.g. table 1, both cranial and post-cranial material are size, duration, open air or covered, fuel source, etc.) present. For the cranium, elements from the face and nature of the body (e.g. positioning, time since (maxilla, nasal), vault (parietal, temporal), and base death, fleshed or de-fleshed, number and density of (occipital) are present. There are no mandibular bones/individuals etc.).20 This analysis assesses fragments. For the post-cranium, elements from the every bone fragment for its colour, texture, the pres- axial (rib, vertebra, sacrum) and appendicular (arm, ence or absence of warping, and presence/absence leg, shoulder, pelvic girdle, hand and foot) skeleton and type of cracking. The stage of heating, based are present, although there are few identifiable frag- upon the stages shown in figure 5, that is the best ments from the hands and feet. Thus, with the match to these four variables is then assigned. exception of the mandible, every part of the body is represented at least minimally. The most com- Sex Estimation mon type of material is long bone fragments, sometimes (if large enough or with a diagnostic In cremated remains it is not appropriate to esti- feature) identifiable to a specific bone. There are mate the sex of an individual based solely on the no duplicate elements in the assemblage, nor any size of bones that can shrink due to heat.21 Rather, bones that differ markedly in regards to age or areas of bones that contain features that are more sex, thus, one individual is represented. common in one sex over the other should be relied upon. In general, sex estimation is most METHODS accurate from the pelvis, followed by the cra- nium, followed by the rest of the post-cranial Different methods are used to investigate the sepa- skeleton.22 The overall robusticity of the skeleton rate components of the analysis, as detailed below. and muscularity of the long bones can be used as supplemental indicators of sex.23 As many fea- Cremation Variables tures as possible should be scored and taken into consideration. Heat-induced changes to the bone were examined In this analysis three sexually dimorphic fea- to assess the cremation temperature and to infer tures from the pelvis are scored according to the

190 Fig. 5. Burn Stages, Temperature, Colour, and Hardness of Cremated Bone (from Lemmers 2012: 82). protocol of the cited references: the ventral arc24, accuracy of osteological adult age estimation, and the greater sciatic notch25, and the pre-auricular the fact that a limited number of features can be sulcus26 (these features are visible in fig. 6). No observed in this case, only broad age-categories cranial fragments were recovered with non-size are used, namely: young adult (18-34 years), mid- based sexually dimorphic features. A five-point dle adult (35-49 years), and old adult (50+ years). scale is used to assign a score to sexually dimor- The three macroscopic methods that are applied phic features that gradate over a continuous range: are described first. 5=M (male), 4=PM (probable male), 3=A (ambigu- ous), 2=PF (probable female), 1=F (female). In addi- Macroscopic tion, the robusticity and presence or absence of pro- nounced muscle markings on the humeri and One rough indicator of age-at-death is based femora are noted. Finally, the length of the odontoid upon the status of cranial suture fusion. Cranial process (dens) on the second cervical vertebra (axis) suture fusion can provide a broad age indication is measured and used as a supplementary indicator since it generally does not commence until an of sex based upon research by Schaffler et al. (1992) individual is fully mature and progresses with demonstrating a significant difference between age such that old adults (50+ years) often possess males and females. None of these features will have many fused sutures with the suture line either in been heavily influenced by heat shrinkage, with the the process of remodelling and obliteration or presence vs. absence traits, such as the ventral arc, already obliterated.27 When a cranium is not pre-auricular sulcus, and muscle markings, not intact, as is the case herein, it is necessary to use influenced at all. cranial suture closure to provide only a very broad age estimate (e.g. <50 or >50 years). It also must Age-at-Death Estimation be kept in mind that heat may force sutures which are still in an early stage of fusion, to reopen, erro- Age-at-death estimation is based upon macro- neously indicating a younger individual.28 How- scopic (visible with the naked eye) and micro- ever, examination of multiple suture locations for scopic methods. Given the limited precision and signs of partial or complete fusion, especially in -

191 Fig. 6. Simpelveld Pelvis. Top middle: Ilium fragments; Left: R. auricular surface; Bottom middle: greater sciatic notch and acetabulum fragment; Bottom right: R. pubic symphysis. volving sutures that are heavily inter-digitated (1990) developed a method of age estimation and thus less likely to easily re-open or re-open from the metamorphosis of the symphyseal sur- without causing fracture of the adjoining bone face of the pubis (shown in fig. 6) of the pelvis that can be useful, again, as long as only very broad is specific to females. Age-related changes on the age estimates are used. pubic symphysis continue throughout adulthood. The other two age-at-death methods are based The Suchey-Brooks system uses six stages, with on features of the pelvis, with the first providing mean ages ranging from 19.4 (stage 1) to 60.0 a more precise estimate. First, Brooks and Suchey (stage 6) years. Second, Lovejoy et al. (1985) used

192 Fig. 7a. Pictomicrograph of Simpelveld femur fragment at 40x magnification under polarized light. Periosteal surface on left.

Fig. 7b. Pictomicrograph of Simpelveld femur fragment at 40x magnification under polarized light with struc- tures indicated. Periosteal surface on left. I=intact osteon; P=partial osteon; CL=circumferential lamellae; L=lamellar bone.

193 Fig. 8a. Pictomicrograph of Simpelveld femur fragment at 40x magnification. Periosteal surface on left.

Fig. 8b. Pictomicrograph of Simpelveld femur fragment at 40x magnification with structures indicated. Peri- osteal surface on left. I=intact osteon; I?=probable intact osteon; P=partial osteon; CL=circumferential lamellae; L=lamellar bone; *=primary (type I, non-Haversian) osteon.

194 the auricular surface of the pelvis (shown in fig. can correctly identify and differentiate several 6) to develop an age-at-death estimation system different types of bone cells as well as measure (with males and females combined due to non- the size of the cells and size of the observation significant differences between them). Buckberry field.32 Consistently good visibility of cellular and Chamberlain (2004) modified Lovejoy et al.’s structures is necessary, which is not common in (1985) system whereby the following features are heated bone. Hence, quantitative histological given a numerical score: transverse organization, methods are not applied to estimate age. Rather, surface texture, microporosity, macroporosity, and a qualitative method is used. In qualitative meth- apical changes. The cumulative score then yields ods the cremated bone cross-section is compared a stage (minimum 1, maximum 7). The Simpel- to a series of cross-sections of unheated bone of veld right auricular surface is not complete (miss- known ages. The cross-sections with the most ing most of the superior half and the apical edge) similar structural appearance are then used to so the best that can be offered is a minimum age derive an age range estimate. and a maximum age. The Simpelveld femur sections are analyzed using a qualitative method devised by Maat and Microscopic colleagues (2003). This method is based upon 162 anterior femoral cortex thin sections from 10 to 91 The final method of age estimation is based on year-old individuals of Dutch (western Euro- microscopic examination of bone cells from a pean) descent (86 males; 76 females). Only indi- fragment from the femur. The normal remodel- viduals free from chronic diseases that may have ling (turnover) of bone throughout adult life has affected bone morphology were sampled. A series proved useful for aging. There are many methods of pictomicrographs is provided for every 10-year developed on different bones whereby different age interval, with pictomicrographs given at low histological structures are quantified (called and high magnification and under bright-light histomorphology), usually involving assessment and polarized light exposures. A table notes that of the percentage of remodelled versus unremod- at 30 years of age approximately 50% of the sub- elled (primary) bone.29 The cremation process periosteal (outer surface) area is composed of often causes bone cross-sections to have cracks, non-remodelled bone (circumferential lamellar gaps, and ‘blackened’ or otherwise non-visible bone and non-Haversian or primary vascular areas, and can cause the outer or inner surface canals) while at 40 years of age this percentage is layers to peel or flake off, making observation reduced to 38%.33 difficult.30 Such cracks, gaps, and differential vis- A final point of consideration regarding histo- ibility are indeed visible in the femur pictomicro- logical age estimate is the precision and accuracy graphs (figs 7a to 8b), but fortunately the inner of the approach. Hummel and Schutkowski (1986) and outer surfaces of the sectioned femur fragment found errors in histology-based age estimates of appear intact. Due to the only partial visibility of up to twenty years with the mean difference be - histological structures the resultant age estimate tween estimated and real age to be 8.4 +/- 6.5 must include a wide range. years. While this may seem like a large difference The methods used to section and analyze the it is actually lower than the standard deviations femur fragment are as follows. The femur frag- of many macroscopic age estimation methods. ment was embedded in a 5:1.95 solution of When using broad age categories, as is done in EpoThin resin and hardener to fill cracks and this case, the histological age estimate can make limit fracturing during sectioning and grinding. a useful contribution towards assigning an indi- The sections were cut using an IsoMet® 1000 saw vidual to the age category of best fit. at very slow speed (100 rpm) and ground and polished by hand using fine grit carborundum Pathology paper before fixing on a microscope slide. All sec- tions are examined with a Leica DM 1000 light Finally, careful macroscopic observation of all bone microscope with optional polarizer filter using fragments was undertaken to assess the presence either 40 or 100 magnification. Pictomicrographs of potential pathological alteration or abnormal are made with a Leica EC3 camera using LAS EZ morphology. Especially osteoarthritis was looked Version 2.0 software. for, evident in synovial joints via joint contour There are two main approaches to histology alteration, marginal bony outgrowths, porosity based age estimation: (1) qualitative and (2) quanti- and pitting of the joint surface, and bone-on-bone tative. Quantitative methods31 require that one contact causing a shiny, worn appearance called

195 eburnation. Osteoarthritis is one of the most com- lar sulcus, a feature that is more common and monly observed pathological conditions in better marked in females43, and thus this is scored ancient skeletons, with changes typically begin- as male (a score of 5). However, it is not uncommon ning in young adulthood and progressively for a female skeleton to lack a pre-auricular sulcus. worsening with age.34 In addition to this age- For example, in a large sample of over 3000 indi- based progression, high levels and/or intensities viduals Gülekon and Turgut (2001) found it was of activity can cause osteoarthritis, hence its fre- present in only 23% of females (and completely quent use as a marker of repetitive, strenuous absent in males). activity in past peoples.35 The other features useful for sex estimation are now presented. The size of the dens is within the RESULTS range for females, well below the range for males, and is thus scored as female (a score of 1, (details Cremation Characteristics given in table 1).44 Finally, fragments from both the humerus and femur that can show pronounced The bony characteristics informative about the robusticity and muscle markings usually more pro- temperature of the cremation suggest it reached nounced in males45, are relatively gracile yielding the upper end of the temperature range. Most a score of probable female (a score of 2). These bones exhibited cracks oriented both perpendicu- data, especially the scores for pubic symphysis lar and parallel to the long axis of the bone and greater sciatic notch, result in a final sex esti- suggesting a burning temperature above 750°C mate of female. (stage 5).36 Most bones had white colouration on them indicating burning at temperatures higher Age-at-Death Estimation than approximately 650°C.37 As well, many bones had a smooth and hard, almost brittle texture, Macroscopic which occurs once temperatures are higher than about 750°C (stage 5). There are a few fragments The results of macroscopic age estimation meth- (parts of the pelvis and vertebrae) with colour ods are provided first. The morphology of the and texture changes that occur at lower tempera- Simpelveld right pubic symphysis most closely tures (range from stage 2 to 4), but these areas matches stage 4 of the Suchey-Brooks system: may have been protected from the highest heat because of their position relative to the heat Symphyseal face is generally fine grained source or protection within the body. although remnants of the old ridge and furrow Most fragments had clear ‘thumbnail’ fractures, system will remain. Usually the oval outline is especially those from long bones and the ribs, complete at this stage, but a hiatus can occur suggesting the body was still covered in tissue in upper ventral rim. Pubis tubercle is fully when heated.38 As well, many fragments had separated from the symphyseal face by defini- heat-induced warping, which some authors have tion of the upper extremity. The symphyseal argued is indicative of a fleshed body cremation,39 face may have a distinct rim. Ventrally, bony although others have found warping in dry-bone ligamentous outgrowths may occur on inferior cremations.40 Finally, many Simpelveld fragments portion of pubic bone adjacent to symphyseal appear shrunk but it is not possible to quantify face. If any lipping occurs it will be slight and the extent of shrinkage without a more complete located on the dorsal border.46 skeleton and pieces that were not heated to this high level and hence retained their normal size. Stage 4 corresponds to an age of 38.2 years with a standard deviation of 10.9 years and a 95% Sex Estimation range of 26 to 70 years. This places the Simpel- veld female at the lower end of the middle adult The results of the sex estimation from the pelvic (35-49 years) age category. features are presented first. The right pubic sym- Recall, the auricular surface is incomplete and physis contains a clear ventral arc, a feature that can thus can only be used to provide a minimum occurs predominately in females41, and is thus and maximum age (based on the morphology of scored as female (a score of 1). The greater sciatic the area that is visible the individual must be notch is very wide, as is most common in a female older than or younger than a certain age). Table pelvis42, and is also scored as female (a score of 2 shows the potential minimum and maximum 1). The right auricular surface lacks a pre-auricu- scores and the resultant age range based on the

196 method of Buckberry and Chamberlain (2004). thritic changes. The lack of osteoarthritis can be While a large age range is all that is possible from cautiously suggested to indicate the Simpelveld the auricular surface, between 29 and 67 years of female was not regularly engaged in activities age, it is useful that this range fully covers the that put a lot of stress and strain on these joints. age estimated from the pubic symphysis. Obviously however, many joints were not observ- Finally, several cranial fragments were recov- able (much of the thoracic and lumbar spine, ered and all had completely open cranial sutures. elbow, wrist, hip, most of the knee, and ankle) so Many of these fragments contained sutures that there may have been osteoarthritis in other areas. were highly inter-digitated with no fused pieces evident along the adjoining margins. It is appro- DISCUSSION priate to use this evidence to suggest the Simpel- veld individual was not an old adult (<50 years The discussion begins with a few points regard- of age). ing the age-at-death estimation methods. The age estimate of middle adult (35-49 years) relies most Microscopic heavily on the outcome of the pubic symphysis method.47 Numerous studies have found the Figures 7a-7b and 8a-8b depict the results of the Suchey-Brooks pubic symphysis method to be histological analysis of the subperiosteal femur one of the most accurate for age estimation.48 shaft fragment. Figures ‘a’ are the raw pictomicro- Since the age estimates from the auricular sur- graphs without demarcation of cells or labelling of face, cranial suture closure, and bone histology, the cellular structures; figures ‘b’ are labelled in are more or less in agreement with the pubic order to demonstrate the presence of different symphysis age, the category of middle adult is cells useful in assessing the amount of non- considered quite reliable. remodelled vs. remodelled bone. There have been concerns that histological The subperiosteal areas of the Simpelveld femur analyses of heated bone are problematic because sections contain both non-remodelled and remod- heating may cause alteration of the cellular struc- elled bone. Based on the qualitative method by tures. However, most research has shown that the Maat (2003) the Simpelveld sections are not con- microscopic structure of bone burned at up to sistent with an individual younger than 30 years 700-800°C does not change significantly.49 This (at 20 years of age 88% of the subperiosteal bone research found that the shrinkage that heated is non-remodelled) and unlikely to be from an bone undergoes is experienced uniformly by the individual older than 50 years (at 51 years 28% of different microscopic structures so that any size the bone is non-remodelled, with percentages decrease is linear and thus the basis of age esti- decreasing with increasing age). Thus, the results mation using the comparative quantification of provide an age estimate of greater than 30 to less structures is retained.50 Yet, other scholars have than 50 years of age. argued that bone heated at >600 °C can be altered and thus it is not appropriate to use quantification- Pathology based age estimation methods.51 Regardless of which camp proves correct this analysis avoided No pathological lesions or osseous anomalies quantification-based age estimation methods, were noted in any of the bone fragments. In the relying instead on a qualitative method that will maxillary fragment showing the sockets of three be considerably less affected by potential cellular molar teeth there appeared to be no horizontal or alteration. As well, the histological age estimate vertical bone loss (sockets were deep, alveolar was intentionally kept broad (as were the age margin was even and smooth) suggesting no estimates from the auricular surface and cranial periodontal disease in this area of the mouth. sutures as these were incomplete) in order to Several joint surfaces were observable, including: reflect the fact that a reasonable degree of accu- the left temporomandibular joint, three rib verte- racy is only possible with an imprecise estimate. bral ends, two cervical and one vertebral body, This is a reality faced by most adult age-at- ten superior/inferior vertebral articular facets, death estimation methods as most of the varia- the promontory (joint between lumbar vertebrae bles we assess are based on skeletal degeneration 5 and sacral vertebrae 1), the left and right scapu- and degradation, or in the case of histomorphol- lar glenoid cavities, the left patella, and several ogy, remodelling, which are influenced by many articular surfaces from hand and foot long bones. factors besides age including activity, nutrition, None of these articular surfaces had any osteoar- disease, hormones, and more. Given the possibil-

197 ity that the Simpelveld woman had a life largely The castra, castella and the new towns formed a free from repetitive, strenuous labour, her level of vast market for all sorts of agricultural products, skeletal degeneration may be less than counter- with wheat being the most important commodity. parts engaged in more regular labour. In this The area closest to this northern border of the regard, the pubic symphysis and auricular sur- Empire which was suitable for large-scale arable face may look a bit younger than she actually farming, was the region of Limburg and the Ger- was. This is speculation however, and any differ- man Rhineland. Not surprisingly, a network of ence would likely be slight. Indeed, the Suchey- main roads was constructed as soon as the early Brooks method was developed upon a forensic 1st century CE, connecting the loess-soils region sample of females from the late 20th century, many with the new towns and army camps along the of whom would have led lives without much stren- Rhine. The trade in food and other goods caused, uous labour, so the composition of the method’s amongst other things, an accumulation of wealth sample may fit with the Simpelveld woman rather for large parts of the farming population in the nicely, which can serve to improve the method’s loess region, which is reflected in both the domes- accuracy. Finally, if the Simpelveld sarcophagus tic sphere and in the mortuary evidence.53 inhabitant was a woman of means, as was likely The native building tradition between the Meuse the case, she probably had a good diet (at least most and Rhine rivers consisted of rectangular struc- of the time) that kept her bones strong and healthy, tures made with organic material such as wood, resulting in generally normal turnover rates. This thatch and clay. After the conquest, the Roman bodes well for the accuracy of the histological age style of architecture was introduced to the region estimation, as frequent or prolonged episodes of with buildings made of stone, concrete and mortar, malnutrition, as well as host of diseases (of which and ceramic building materials such as rooftiles, there was no evidence of in the limited fragments and elements such as the porticus, floor-heating that remained), can alter bone remodelling rates and wall-decoration. However, it was not until the causing a discrepancy between an individual’s end of the 1st century CE that this building style chronological age and the age of their skeleton. became more widespread in the area.54 At the same time, the native building style with post- Environment: the Loess-Soils Region of Germania built houses was still in force throughout the Inferior region,55 and recent excavations of entire rural settlements consisting of several contemporaneous The richness of the burial evidence found at Sim- buildings made with organic material only serve pelveld in combination with the results of the to emphasize this fact.56 And on almost every villa study of the cremated bones indicates that the estate in the region the main house was often the person interred here was part of the upper-most only stone-built structure, with the rest of the echelon of provincial-Roman society in this part auxiliary buildings made in the vernacular style.57 of the Empire. Simpelveld lies in the heart of the A stone built Roman-style house did not only loess-soils region that in the Netherlands occurs cost a substantial amount of money, but also only in the southernmost province of Limburg required the expert knowledge of a number of (fig. 9). This region forms part of an extensive craftsmen to produce the various elements of the band of loess-soils that covers most of northern structure. Therefore it was not just a dwelling, but France, southeast Belgium and part of the Ger- first and foremost a status symbol, indicating the man Rhineland. During the Roman period, this wealth and lifestyle of its owner.58 And stone entire region underwent a dramatic transforma- built houses in the countryside were certainly not tion, with farming practices switching from sub- all equal, as has been pointed out recently. The sistence to surplus-producing.52 The establish- large axial villas for example, that represent the ment in the course of the 1st century CE of the most opulent examples of Roman rural building, Roman border or limes along the river Rhine are found frequently in the northeast of France and meant that over 50.000 soldiers settled within less southeast of Belgium, but from the Roman coun- than 100 kms from this region. During the 1st cen- tryside of South Limburg they are completely tury CE four mayor capitals of local civitates were lacking.59 The same study also showed that only founded in the settlement landscape between a small percentage of the stone-built houses in Meuse and Rhine: Atuatuca Tungrorum, Colonia the countryside boasted luxurious elements such Claudia Ara Agrippinensium, Colonia Ulpia Tra- as a hypocaust, private baths, or walls decorated iana and Noviomagus (current-day Tongeren, with frescoes.60 In other words, being able to Cologne, Xanten, and Nijmegen, respectively). afford the construction of a stone-built Roman

198 Fig. 9. Map of Roman South Limburg with the distribution of burial evidence in the villa landscape between and . style house was in itself an accomplishment, but ascribed to the more privileged members of within the group of farmers being able to do so, Roman society. As with the stone-built houses only a minority was so affluent as to be able to however, a subdivision can be made even in the afford truly expensive elements such as private category of monumental mortuary evidence, based baths or luxurious interior decorating.61 This is on the composition of the individual objects and also reflected in the mortuary evidence of the elements in the burial. In one of the four burial region. zones of the vicus of Coriovallum (current-day The number of Roman burials in the country- Heerlen) for example, different types of monu- side of South Limburg is considerable. They are mental burials were found, such as an inscribed especially evident in the quantity and quality of grave pillar, a small ‘sarcophagus’ made of roof- the grave goods. In approximately 75% of all tiles in which several objects were placed, and cases the evidence consists of cremated remains most spectacularly, a group of four sarcophagi, in a container, placed in a pit together with a few with a collection of grave goods of exceptional pieces of pottery. One quarter consist of burial quality, made of gold, glass, and amber (fig. 10). evidence of an altogether much more monumen- The sarcophagi themselves, which were buried tal character, such as stone coffins, tower tombs, alongside the main Roman road from Coriovallum grave pillars, and tumuli.62 In these types of bur- to Colonia Claudia Ara Agrippinensium, were of ials large assemblages of grave goods are often a rather roughly hewn sandstone, with the lid found, of a variety of materials, such as bronze connected to the coffin by means of bronze clamps.63 and glass. Not surprisingly, these burials are In the same area as the sarcophagi, a grave pillar

199 Fig. 10. The discovery of four large stone incinerary urns at the Voskuilenweg in Heerlen in 1920, with local archaeologist Piet Peters on the left. decorated with the image of a man and a woman, although the incompleteness of the skeleton pre- presumably a couple, was found a few years ear- vents this from being concluded definitely. This lier. Because of the hairstyle of the woman, this research demonstrated that even in highly heated grave pillar is referred to as ‘the Ubian couple’. and fragmented human remains histological Arguably this pillar was the above-ground mon- analyses can still provide useful, albeit broad, umental burial marker of two of the four sarcoph- information on age-at-death when assessed using agi.64 The opulence of the entire assemblage not- a qualitative approach. Future human cremation withstanding, this burial pales in comparison to research should consider histological analysis when that of Simpelveld, with its richly decorated cof- identifiable long bone fragments are present and fin in addition to the precious grave goods. The few or no macroscopic age-at-death estimation fact that the coffin showed elements of both the methods can be applied. Thus, despite the fact that interior and exterior of a typical Roman-style stone the past opening and pillaging of the sarcophagus built house proves that it represents that house and resulted in the loss of a significant amount of skel- that it was seen as an integral part of that what etal material, a lot of information was gleaned from the deceased would need in the afterlife. The fact what remained. With this osteobiographical pro- that this is the only coffin in the entire Roman file of the inhabitant of the famous Simpelveld northwest, in the large number of stone coffins sarcophagus we can ponder to what extent the found in the region and beyond, is clear evidence woman’s life was reflected in the decorative fea- of the elevated status and wealthy background of tures of the sarcophagus. Was the sarcophagus the woman for whom it was made. designed specifically for her, and meant to depict her life, or rather, the image of her life she wanted CONCLUSIONS portrayed? Comparison of the Simpelveld sar- cophagus to other monumental burials, including This research has demonstrated that the middle- other stone coffins, from the region, demonstrates aged woman interred in the Simpelveld sarcoph- just how much it stands out. It is the most richly agus was cremated in a fire of high temperature decorated coffin, and probably was the most (>750°C) involving her complete, fleshed, body. richly adorned, although because of looting we The lack of pathological lesions may be evidence cannot know how many precious grave goods for a life free from repetitive strenuous labour, were lost. Likely an important function of the sar-

200 31 cophagus was the idea to showcase in the after- E.g. Ahlqvist/Damsten 1969; Kerley 1965; Maat et al. 2006; Thompson 1979. life the woman’s status and wealth, as a signal as 32 Crowder 2013. well as the material means of providing her with 33 Maat et al. 2003, 8. all that someone of her position should receive. 34 Reviewed in Roberts/Manchester 2007; Waldron 2009. 35 E.g. Molnar et al. 2011; Palmer et al. 2016; Pearson/ NOTES Buikstra 2006. 36 Wahl 1982. 37 Etxeberria 1994; Mays 1998; Shipman et al. 1984; * Andrea Waters would like to thank Simone Lemmers Walker/Miller 2005. for her invaluable advice about the analysis of the cre- 38 Binford 1963; Buikstra/Swegle 1989; Ubelaker 1978. mated material and Menno Hoogland for his photo- 39 Baby 1954; Binford 1963; Etxeberria 1994. graphic expertise. 40 Buikstra/Swegle 1999; Spenneman/Colley 1989; 1 Although this monument is referred to by most schol- Whyte 2001. ars as the ‘Sarcophagus of Simpelveld’ it was not used 41 Phenice 1969. for inhumation: it is a container for a cremation burial, 42 Buikstra/Ubelaker 1994. decorated on the inside as a room, with the personal 43 Buikstra/Ubelaker 1994. belongings of the lady whose ashes were placed here. 44 Schaffler et al. 1992. However, since the word ‘sarcophagus’ has been used 45 Bass 1995. to describe the Simpelveld monument, the term will be 46 Brooks/Suchey 1990. used here in order to prevent confusion. For the role of 47 Brooks/Suchey 1990. mortuary ritual used by priviliged social groups see 48 E.g. Klepinger et al. 1992; Martrille et al. 2007. Crowley 2011, 200. 49 Brandtmiller/Buikstra 1984; Herrmann 1977; Nelson 2 For a complete list of publications, see Zinn 1997. 1992; Shipman et al. 1984; Walker et al. 2008; also see 3 Zinn 1997, 143: ‘6. Porträttyp der Faustina Minor.’ papers in Grupe/Garland 1993. 4 Zinn 1997, 144-146. 50 Hummel/ Schutkowski 1986. 5 Zinn 1997, 136. 51 Holden et al. 1995. 6 These are the types Isings 1957, forms 50-51, common 52 Amongst the many publications referring to this, the in the 1st and 2nd centuries CE. most recent that are relevant for the region at hand are 7 Richter 1966, 111-112; Moss 1988, 37-43; Mols 1999, Habermehl 2013, 1 and Roymans/Derks 2011, 16 and 20. 44-52. 53 Roymans/Derks 2011, 15-16; Crowley 2011, 201. 8 Zinn 1997, 137-138. Produced in Gaul and 54 Habermehl 2013, 60-63. during the 2nd-3rd centuries CE. 55 Jeneson 2011, 262; Jeneson 2013, 139. 9 Zinn 1997, 148-150. 56 See for example Vanderhoeven 2006 and Tichelman 10 ‘… allgemein gehaltene Abbreviatur eines grossen 2009 for examples of recently excavated rural post-built Gebäudekomplexes. […] Dazu wählte der Bildhauer settlements in the region. geradezu chiffrehaft die einfachste Hausform, nämlich 57 See for evidence in Limburg Tichelman 2005, and for ein langrechteckiges Gebäude, und ergänzte es durch evidence in the neighbouring region of the German einen kleineren Anbau.’ Zinn 1997, 139. Rhineland several publications by Gaitzsch, such as the 11 Henig in King/Henig 1981, 129; Deppert-Lippitz ones from 1991, 2008 and 2011. 1985b, 123; Pirzio Biroli Stefanelli 1992, 75; Zinn 1997, 58 Habermehl 2013, 108-113. 140. 59 Roymans/Habermehl 2011, 85. 12 This is evidenced by a stark increase in rural settle- 60 Jeneson 2013, 87-97. ments with stone-built houses in the loess-region as 61 Roymans/Derks 2011, 27. well as an increase in wealth display both in the new 62 Jeneson 2012, 92-97. towns and the countryside. See Zinn 1997, 142-144; 63 Peters 1920, 51. Jeneson 2013, 138. 64 Peters 1920, 51. 13 Lemmers 2012. 14 Devlin 2008. 15 Etxeberria 1994; Mays 1998; Shipman et al. 1984; BIBLIOGRAPHY Walker/Miller 2005. 16 Wahl 1982. Ahlquist, J./O. Damsten 1969, A modification of Kerley’s 17 Binford 1963; Buikstra/Swegle 1989; Ubelaker 1978. 18 method for the microscopic determination of age in Brandtmiller/Buikstra 1984; Grupe/Hummel 1991; human bone, Journal of Forensic Sciences 14, 205-212. Holland 1989; Shipman et al. 1984. Baby, R.S. 1954, Hopewell cremation practices, Ohio His- 19 Wahl 1982. 20 torical Society Papers in Archaeology, 1-7. Lemmers 2012. Bass, W.M. 1995, Human osteology: a laboratory and field 21 Thompson 2002. manual of the human skeleton, Missouri. 22 Buikstra/Ubelaker 1994. Binford, L.R. 1963, An analysis of cremations from three 23 Bass 1995. Michigan sites, Wisconsin Archaeologist 44, 98-110. 24 Phenice 1969. Brandtmiller B./J.E. Buikstra 1984, Effects of burning on 25 Buikstra/Ubelaker 1994. human bone microstructure: a preliminary study, Jour- 26 Buikstra/Ubelaker 1994. nal of Forensic Sciences 29, 535-540. 27 Meindl/Lovejoy 1985; White 2011. Brooks, S.T./J.M. Suchey 1990, Skeletal age determination 28 Lemmers 2012. based on the os pubis: a comparison of the Acsadi- 29 See Robing/Stout 2008 for a review. Nemeskeri and Suchey-Brooks methods, Human Evolu- 30 Hummel/Schutkowski 1993. tion 5, 227-238.

201 Buckberry, J.L./A.T. Chamberlain 2002, Age estimation Herrmann, B. 1988, Behandlung von Leichenbrand. Rekon- from the auricular surface of the ilium: a revised method, struktion, Konservierung und Reproduktion, in R. American Journal of Physical Anthropology 119, 231-239. Knussmann (ed.), Anthropologie. Wesen und Methoden Buikstra, J.E./M. Swegle 1989, Bone modification due to der Anthropologie 1, Stuttgart, 576-585. burning: experimental evidence, in R. Bonnichsen (ed.), Holden, J.L./P.P. Phakey/J.G. Clement 1995, Scanning Bone Modification, Maine, 247-258. electron microscope observations of heat-treated Buikstra, J.E./D.H. Ubelaker (eds) 1994, Standards for data human bone, Forensic Science International 74, 29-45. collection from human skeletal remains (Arkansas Archae- Holland, T. 1989, Use of the cranial base on the identifica- ological Survey Report 44), Fayetteville. tion of fire victims, Journal of Forensic Sciences 34(2), Crowder, C. 2013, Estimation of age at death using cortical 458-460. bone histomorphometry. Final report submitted to the U.S. Hummel, S./H. Schutkowski 1986, Das Verhalten von Department of Justice, Office of Justice Programs, National Knochengewebe unter dem Einfluß höherer Tempera- Institute of Justice, Washington. turen. Bedeutung für die Leichenbranddiagnose, Zeit- Crowley, L. 2011, The role of mortuary ritual in the con- schrift für Morphologie und Anthropologie 71, 1-9. struction of social boundaries by privileged social Hummel, S./H. Schutkowski 1993, Approaches to the his- groups within villa landscapes, in Roymans/Derks tological age determination of cremated human (eds), 2011, 195-210. remains, in Grupe/Garand 1993, 111-123. Cuijpers, S.A.G.F.M. 1997, Possibilities of histological Jeneson, C.F. 2011, Evaluating settlement patterns and set- research on diaphyseal fragments in cremated remains, tlement densities in the villa landscapes between Ton- in A.G. Drusini/E. Iregren/E. Smits (eds), Cremation gres and Cologne, in Roymans/Derks 2011, 259-274. Studies in Archaeology, Saonara, 73-86. Jeneson, C.F. 2013, Exploring the Roman villa world between Deppert-Lippitz, B. 1985, Römischer Goldschmuck – Tongres and Cologne. A landscape archaeological approach Stand der Forschung, in Aufstieg und Niedergang der (unpublished PhD thesis, VU University Amsterdam). römischen Welt II 12. 3, Berlin/New York, 117-126. Kerley, E.R. 1965, The microscopic determination of age in Etxberria, F. 1994, Aspectos macroscópicos del hueso human bone, American Journal of Physical Anthropology sometido al fuego: revisión de las cremaciones descritas 23, 149-163. en el País Vasco desde la arqueología, Munibe 46, 111–116. King, A./M. Henig (eds) 1981, The Roman West in the 3rd Fairgrieve, S. 2008, Forensic cremation: recovery and analysis, Century: Contributions from Archaeology and History Boca Raton. (British Archaeological Reports International Series Gaitzsch, W. 1991, Römische Hof- und Wirtschaftsräume 109), Oxford. im Rheinischen Braunkohleabbau am Beispiel des Ham- Klepinger, L.L./D. Katz/M.S. Micozzi/L. Carroll 1992, bacher Forstes, in B. Gramsch/G. Wetzel (eds), Archäo- Evaluation of cast methods for estimating age from the logische Erkundung und Rettungsarbeit in Tagebaugebieten os pubis, Journal of Forensic Sciences 37, 763-770. Mitteleuropas (Veröffentlichungen des Museums für Ur- Lemmers, S. 2012, Burned culture: osteological research und Frühgeschichte Potsdam 25), Berlin, 125-134. into Urnfield cremation technology and ritual in the Gaitzsch, W. 2008, Von der Via Belgica zur Sophienhöhe. South of the Netherlands, Lunula Archaeologia protohis- Straßenforschung im Tagebau Hambach, in J. Kunow torica, 81-88. (ed.), Erlebnisraum Römerstraße. Via Belgica (Materialien Lovejoy, C.O./R.S. Meindl/T.R. Pryzbeck/R.P. Mensforth zur Bodendenkmalpflege im Rheinland 18/2), , 1985, Chronological metamorphosis of the auricular 31-43. surface of the ilium: a new method for the determina- Gaitzsch, W. 2011, Roman villa landscapes of the lignite tion of adult skeletal age at death, American Journal of mining areas in the hinterland of Cologne, in N. Physical Anthropology 68,15-28. Roymans/T. Derks (eds), Villa landscapes in the Roman Maat, G.J.R./M.J. Aarents/N.J.D. Nagelkere 2005, Age north. Economy, culture and lifestyles (Amsterdam Archaeo- logical Studies 17), Amsterdam, 285-300. prediction from bone replacement: Remodeling of cir- Galestin, M.C. 2001, The Simpelveld sarcophagus: a unique cumferential lamellar bone tissue in the anterior cortex monument in a provincial Roman context, in T.A.S.M. of the femoral shaft of the present Dutch population, Panhuysen (ed.), Die Maastrichter Akten. Typologie, Iko- Barge’s Anthropologica 10. nographie und soziale Hintergründe der provinzialen Grab- Maat, G.J.R./A. Maes/M.J. Aarents/N.J.D. Nagelkerke denkmäler und Wege der ikonographischen Einwirkung, 2006, Histological age prediction from the femur in a Maastricht, 63-76. contemporary Dutch sample, Journal of Forensic Sciences Grupe, G./A.N. Garland (eds) 1993, Histology of ancient 51, 230–237. human bone: methods and diagnosis. Proceedings of the Martrille, L./D.H. Ubelaker/C. Cattaneo/F. Seguret/M. “Palaeohistology Workshop” held from 3–5 October 1990 at Tremblay/E. Baccino 2007, Comparison of four skeletal Göttingen, Berlin/Heidelberg. methods for the estimation of age at death on white and Grupe, G./B. Hummel 1991, Trace element studies on black adults, Journal of Forensic Sciences 52(2), 302-307. experimentally cremated bone. Alteration of the chem- Mays, S. 1998, The archaeology of human bones, New York. ical composition at high temperatures, Journal of McKinley, J. 2000, The analysis of cremated bone, in M. Archaeological Science 18, 177-178. Cox/S. Mays (eds), Human Osteology in Archaeology and Gülekon, I.N./H.B. Turgut HB 2001, The preauricular sul- Forensic Science, Cambridge, 403-422. cus: its radiologic evidence and prevalence, Kaibogaku Meindl, R.S./C.O. Lovejoy 1985, Ectocranial suture clo- Zasshi 76(6), 533-535. sure: A revised method for the determination of skel- Habermehl, D. 2013, Settling in a changing world. Villa etal age at death based on the lateral-anterior sutures, development in the northern provinces of the Roman empire American Journal of Physical Anthropology 68, 57-66. (Amsterdam Archaeological Studies 19), Amsterdam. Molnar, P./T.J. Ahlstrom/I. Leden 2011, Osteoarthritis and Herrmann, B. 1977, On histological investigations of cre- Activity - An Analysis of the Relationship Between mated human remains, Journal of Human Evolution 6, Eburnation, Musculoskeletal Stress Markers (MSM) 101-103. and Age in Two Neolithic Hunter–Gatherer Popula-

202 tions from Gotland, Sweden, International Journal of Uytterschaut, H.T. 1985, Determination of skeletal age by Osteoarchaeology 21, 283-291. histological methods, Zeitschrift für Morphologie und Mols, S.T.A.M. 1999, Wooden Furniture in Herculaneum – Anthropologie 75, 331-340. Form, Technique and Function, Amsterdam. Vanderhoeven, A. (et al.) 2006, Veldwezelt. Vues socio-éco- Moss, C.F. 1988, Roman Marble Tables, Ann Arbor. nomiques nouvelles sur les campagnes, in Dossiers Nelson, R. 1992, A microscopic comparison of fresh and Archéologie et Sciences des Origines 315: La Belgique burned bone, Journal of Forensic Sciences 37(4), 1055-1060. Romaine, 60-63. Palmer, J.L.A./M.H.L. Hoogland/A.L. Waters-Rist 2016, Wahl, J. 2008, Investigations on Pre-Roman and Roman cre- Activity Reconstruction of Post-Medieval Dutch Rural mation remains from southwestern Germany: results, Villagers from Upper Limb Osteoarthritis and Enthe- potentialities and limits, in C.W. Schmidt/S.A. Symes seal Changes, International Journal of Osteoarchaeology (eds), The analysis of burned human remains, London, 145- 26, 78–92. 161. Pearson, O.M./J.E. Buikstra 2006, Behavior and the bones, Waldron, T. 2009, Palaeopathology, Cambridge. in J.E. Buikstra/L.A. Beck (eds), Bioarchaeology: the con- Walker, P./K. Miller 2005, Time, temperature, and oxygen textual analysis of human remains, London, 207-226. availability: an experimental study of the effects of Peters, P. 1920, Romeinse vondsten te Heerlen: vier zand- environmental conditions on the colour and organic stenen askisten, De , 51-57. content of cremated bone, American Journal of Physical Phenice, T.W. 1969, A newly developed method of sexing Anthropology 40, 222. the os pubis, American Journal of Physical Anthropology Walker, P.L./K.W.P Miller/R. Richman 2008, Time, tem- 30 (2), 297-302. perature, and oxygen availability: an experimental Pirzio Biroli Stefanelli, L. (ed.) 1992, L’oro dei Romani, study of the effects of environmental conditions on the Rome. colour and organic content of cremated bone, in Richter, G.M.A. 1966, The Furniture of the Greeks, Etruscans Schmidt/Symes 2008, 110-129. and Romans, London. White, T.D./M. Black/P.A. Folkens 2011, Human Osteology, Roberts, C./K. Manchester 2007, The archaeology of disease, Amsterdam. Cornell. Whyte, T. 2001, Distinguishing remains of human crema- Robling, A.G./S.D. Stout 2007, Histomorphometry of tions from burned animal bones, Journal of Field Archae- human cortical bone: applications to age estimation, in ology 28 (3-4), 437-448. M.A. Katzenberg/S.R. Saunders (eds), Biological Anthro- Zinn, F. 1997, Überlegungen zum Sarkophag von Simpel- pology of the Human Skeleton, Hoboken, 149-182. veld, OMRO 77, 135-158. Roymans, N./T. Derks (eds) 2011, Villa landscapes in the Roman north. Economy, culture and lifestyles (Amsterdam Archaeological Studies 17), Amsterdam. ANDREA WATERS-RIST Roymans, N./T. Derks 2011, Studying Roman villa land- FACULTY OF ARCHAEOLOGY, LEIDEN UNIVERSITY scapes in the 21st century. A multi-dimensional approach, in Roymans/Derks 2011, 1-44. [email protected] Roymans, N./D. Habermehl 2011, On the origin and development of axial villas with double courtyards in RUURD HALBERTSMA the Latin West, in Roymans/Derks 2011, 83-106. CURATOR OF THE CLASSICAL DEPARTMENT, NATIONAL Schaffler, M.B./M.D. Alson/J.G. Heller/S.R. Garfin 1992, MUSEUM OF ANTIQUITIES, LEIDEN Morphology of the dens: a quantitative study, Spine 17(7), 738-743. [email protected] Schmidt, C.W./S.A. Symes (eds) 2008, The analysis of burned human remains, London. KAREN JENESON Spennemann, D./S. Colley 1989, Fire in a pit: the effects CURATOR OF THE ROMAN BATHS MUSEUM, HEERLEN of burning on faunal remains, Archaeozoologia 3, 51-64. Shipman, P./G. Foster/M. Schoeninger 1984, Burnt bone [email protected] and teeth: an experimental study of colour, morphol- ogy, crystal structure and shrinkage, Journal of Archaeo- logical Science 11, 307-325. Thompson, D.D. 1979, The core technique in the determi- nation of age at death in skeletons, Journal of Forensic Sciences 24(4), 902-915. Thompson, T. 2002, Assessment of sex in cremated indi- viduals: some cautionary notes, Canadian Society of Forensic Science Journal 35(2), 49-56. Tichelman, G. 2005, Het villacomplex -Holzkuil (ADC Rapport 155), Amersfoort. Tichelman, G. 2009, Bedrijventerrein Trilandis, gemeente Heerlen. Archeologisch vooronderzoek: een inventariserend veldonderzoek, Weesp. Trautmann, I. 2006, The significance of cremations in Early Neolithic communities in Central Europe, PhD Dissertation, Institut für Ur- und Frühgeschichte und Archäologie des Mittelalters, Eberhard Karls Universität, Tubingen. Ubelaker, D.H. 1978, Human skeletal remains: excavations, analysis, interpretation, Chicago.

203 Table I. Cremated Human Skeletal Material, Simpelveld Sarcophagus.

Item Bone No of Burning Weight Mean Age Sex Pathology/ Notes Pieces Degree (g) Size of Estimate Estimate Anomaly Pieces (cm) Cranial 1 R maxilla alveo- 1 Colour: white 0.7 2.0 adult sized -- No signs lus fragment Texture: chalky sockets of alveolar showing sockets Warping: no loss (peri- of M3, M2, and Cracking: yes odontal distal half of M1 disease) 2 Maxilla 5 Colour: yellow- 1.2 1.0 -- -- no Fragments fragments white show sockets Texture: smooth of root tips. Warping: no Not complete Cracking: yes enough to identify which tooth

3 R. nasal 1 Colour: brown- 0.4 <1.0 -- -- no white Texture: normal Warping: no Cracking: no 4 Occipital 5 Colour: white 10.2 2.5 to all observable no no Can observe fragments Texture: chalky to 4.0 sutures very interdigi- smooth (hard) unfused = tated sutures Warping: yes not old adult Cracking: yes 5 Temporal 3 Colour: white 7.1 2.5 to all observable no no fragments Texture: chalky to 5.0 sutures smooth (hard) unfused = Warping: yes not old adult Cracking: yes 6 L temporal frag- 1 Colour: brown 1.3 1.5 -- -- no (no OA ment of inferior Texture: normal on portion temporo-mandi- Warping: no of TMJ bular joint Cracking: no that was observable) 7 Skull vault 13 Colour: yellow- 14.8 1.0 to all observable no no (no Can observe fragments white 4.0 sutures fragments many open Texture: chalky unfused = with sutures (light) to smooth not old adult porotic (hard) hyperosto- Warping: yes sis) Cracking: yes

8 Sphenoid 1 Colour: white 1.3 3.0 -- -- no fragment Texture: hard Warping: no Cracking: yes

9 Unidentified 39 Colour: brown, 10.5 <1.0 -- -- no No fragments cranial fragments yellow and white with fused Texture: normal to sutures chalky to smooth Warping: on some Cracking: on some

204 Post-Cranial: Axial 10 R rib vertebral 1 Colour: white 2.8 4.0 -- -- no end showing rib Texture: chalky tubercle Warping: no Cracking: yes

11 Rib vertebral 2 Colour: white 1.5 2.0 -- -- no ends Texture: chalky Warping: yes Cracking: yes

12 Rib body 29 Colour: white 16.9 >1.0 -- -- no fragments Texture: chalky to to 6.0 smooth (hard) Warping: yes Cracking: yes 13 axis 2 Colour: yellow- 5.2 4.0 to fully developed - size of no (no OA white 5.0 = adult dens (13.22 on the Texture: chalky mm) closer dens, infe Warping: no to F avg. rior body Cracking: yes (14.0 mm) surface, or than male the inferior avg. articular (14.7mm ): surface) score as F

14 cervical 1 Colour: brown 1.6 1.5 -- -- no (no OA vertebrae body Texture: normal on superior Warping: no or inferior Cracking: yes body)

15 cervical 1 Colour: black- 0.4 <1 cm -- -- no vertebrae white neural arch Texture: charred to chalky Warping: no Cracking: no

16 R side of thoracic 1 Colour: white 3.0 4.0 -- -- no (no OA vertebra, showing Texture: chalky on body edge of superior Warping: no surfaces or and inferior body, Cracking: no inferior the full R neural articular arch and inf. artic- process) ular process

17 vertebrae supe- 8 Colour: white 3.6 <1.0 -- -- no (no OA rior and inferior Texture: normal to to 2.0 or lamellar articular facets chalky to smooth spurs) Warping: yes Cracking: yes 18 vertebra spinous 1 Colour: white 0.6 1.0 -- -- no process Texture: chalky Warping: no Cracking: no 19 vertebra body 1 Colour: brown 0.5 <1.0 -- -- no fragment Texture: normal Warping: no Cracking: no

205 20 Sacrum – Prom- 1 Colour: brown- 8.8 8.0 -- -- no (no OA ontory (S1) white on the Texture: normal promon Warping: yes tory) Cracking: yes Post-Cranial: Appendicular 21 Scapula – 1 Colour: white 3.6 3.0 -- -- no (no OA L glenoid cavity Texture: smooth of the Warping: yes joint) Cracking: yes 22 Scapula – 2 Colour: white 1.8 1.0 to -- -- no (no OA R glenoid cavity Texture: smooth 2.0 of the to chalky joint) Warping: yes Cracking: yes 23 Scapula 4 Colour: white 5.8 2.0 to -- -- no fragments Texture: smooth 4.0 (hard) Warping: yes Cracking: yes

24 Probable 13 Colour: black to 18.2 1.0 to -- - no pro- no humerus shaft white 5.0 nounced fragments Texture: charred muscle to chalky to markings smooth and overall Warping: yes very grac- Cracking: yes ile: score 2 (PF)

25 Ulna shaft 4 Colour: yellow to 6.5 2.0 to -- -- no - 2 fragments fragments white 3.0 mend Texture: chalky to smooth (hard) Warping: no Cracking: yes

26 Radius shaft 4 Colour: yellow to 5.0 1.0 to -- -- no - 3 fragments fragments white 2.0 mend Texture: chalky to smooth (hard) Warping: no Cracking: yes

27 Probable lower 6 Colour: yellow- 2.6 1.0- -- -- no - no mends arm bone frag- white 2.0 ments (radius or Texture: chalky ulna) Warping: no Cracking: yes 28 Pelvis R pubic s 1 Colour: yellow- 3.5 5.0 Brooks-Suchey - clear no ymphysis white (1990): stage 4 ventral arc: Texture: chalky = 38.2+/-10.9 score 1 (F) Warping: no years Cracking: yes

206 29 Pelvis R 2 Colour: dark 4.8 6.0 Buckberry & - no pre- no auricular surface (mend) brown-grey Chamberlain auricular Texture: normal (2004): sulcus: Warping: no composite score 5 (M) Cracking: yes minimum score of 8 (stage 2): min- imum age 29.33+/-6.71; composite maximum score of 15 (stage 6): max- imum age 66.7+/-11.88 years

30 Pelvis iliac blade 2 Colour: dark 8.7 2.0 adult - epiph- -- no and portion of brown-white to ysis fused and greater sciatic Texture: normal 10.0 line obliterated notch Warping: no Cracking: yes 31 Pelvis fragments 3 Colour: dark 6.6 3.0 -- - very wide no (one piece shows brown-white to sciatic part of the greater Texture: chalky 8.0 notch: sciatic notch) Warping: yes score 1 (F) Cracking: yes 32 Femur shaft 12 Colour: white 33.3 2.0 -- - no pro- no - 8 pieces fragments Texture: chalky to nounced mend and smooth (hard) 8.0 muscle Warping: yes markings Cracking: yes and not robust: score as a 2 (PF)

33 L. patella 1 Colour: dark 4.5 5.0 -- -- no posterior sur- (~70% complete) brown face intact; Texture: normal anterior sur- Warping: no face damaged Cracking: yes 34 Tibia shaft 6 Colour: yellow- 9.5 2.0 -- -- no no mends fragments white to Texture: chalky to 4.0 smooth (hard) Warping: no Cracking: yes

35 Fibula shaft 5 Colour: white 8.5 2.0 -- -- no 2 mends fragments Texture: chalky to to. smooth (hard) 5.0 Warping: yes Cracking: yes 36 Long bone 7 Colour: brown to 11.3 2.0 -- -- no no mends articular ends black to Texture: normal to 6.0 charred Warping: no Cracking: yes

207 37 Unidentified 86 Colour: white 33.8 <1.0 -- -- no not identifiable long bone Texture: chalky to to 2.0 to a specific fragments smooth (hard) long bone Warping: some yes Cracking: some yes 38 metacarpal head 1 Colour: white 0.7 2.0 -- -- no Texture: chalky Warping: no Cracking: yes 39 carpal distal 1 Colour: white 0.1 1.0 -- -- no phalange Texture: chalky Warping: yes Cracking: yes 40 carpal or tarsal 1 Colour: white 0.2 1.0 -- -- no proximal Texture: chalky phalange Warping: yes Cracking: yes 41 calcaneus 1 Colour: white 0.7 2.0 -- -- no fragment Texture: chalky Warping: yes Cracking: yes

42 probable 1 Colour: brown 0.1 <1.0 -- -- no can see a metatarsal base Texture: normal small bit of fragment Warping: no the articular Cracking: yes surface 43 unidentified 105 Colour: white 11.6 <1.0 -- -- no fragments Texture: chalky to smooth (hard) Warping: some yes Cracking: some yes

Abbreviations: L = left; R = right; MC = metacarpal; MT = metatarsal; OA = osteoarthritis * No dental material recovered.

Table 2. Simpelveld Auricular Surface Scores following Buckberry and Chamberlain (2004).

Minimum score Maximum Score Comments Transverse Organization 3* 4 ½ observable Surface Texture 2* 4 ½ observable Microporosity 1* 2 ½ observable Macroporosity 1* 2 ½ observable Apical Changes 1 3 not observable Composite Score 8 15 Older than 29 Stage 2 6 and younger than 67 years Age and Standard Deviation 29.33 +/- 6.71 66.71 +/- 11.88

*Actually observed. The minimum score represents what was actually observed on the skeleton (except for the apical changes where the lowest possible value is listed); the maximum score is what could theoretically be obtained if the unobservable areas displayed the highest amount of change possible.

208