SEARCH

SEARCH BY CITATION

Keywords:

  • breech presentation;
  • embryotomy;
  • Hippocrates;
  • micro-computed tomography;
  • obstetrics;
  • palaeopathology;
  • Soranus

ABSTRACT

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. References

A perinatal infant skeleton from the first–fourth century AD Roman villa site at Hambleden, England, shows what appear to be cut marks on the proximal part of the right femur. Gross, microscopic and micro-computed tomography evaluations suggest that they occurred perimortem and were probably caused by a non-serrated blade. The reason for the cuts is uncertain, but their location is consistent with the practice of embryotomy, as described in classical sources for obstructed labour due to a dead or dying foetus in a breech or leg presentation. If this interpretation is correct, this case represents a rare example of embryotomy in the palaeopathological record. Copyright © 2012 John Wiley & Sons, Ltd.


Introduction

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. References

The Yewden Roman villa, at Hambleden, Buckinghamshire, England (Figure 1), was excavated in 1912 (Cocks, 1921; Eyers, 2011). The villa consisted of a complex of buildings surrounded by a boundary wall. Artefactual and numismatic evidence show that the villa was in use between the first and fourth centuries AD, when Britain was a province of the Roman Empire. The original excavator (Cocks, 1921) recorded 103 human burials at the site, of which 97 were infants. Remains of 41 burials, of which 35 are infants, have recently been located in the site archive and have been restudied (Mays et al., 2011). The majority of the infant remains come from a yard area on the north side of the villa complex. In the re-examination of the human remains from the site, one of the infant burials from the central yard area, burial 38, was found to bear cut marks on the right femur. Like the other infant burials from the yard, this interment was recorded at a depth of 0.5 m, but nothing is known of the orientation or the posture of the corpse.

image

Figure 1. Location of Hambleden.

Download figure to PowerPoint

Materials and methods

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. References

The bones of burial 38 are well preserved, with minimal post-depositional erosion. However, the skeleton is only about two-thirds complete (Figure 2). The missing bones likely reflect inadequacy of recovery on-site in 1912. Long-bone lengths (Scheuer et al., 1980) indicate an age at death of approximately 35–37 weeks' gestation. Full term is normally about 38–41 weeks (Tanner, 1989, 43), so burial 38 may have been delivered a few weeks pre-term. A microscopic study of the dentition (Hillier, 1992) failed to reveal a neonatal line. This suggests that the infant was probably either born dead or died within about 14 days of birth (Whittaker & Richard, 1978). The remains were examined using gross inspection, low-power binocular microscopy and micro-computed tomography (μCT). For μCT, a SkyScan 1172 system (SkyScan, Kontich, Belgium) was used at 62 kV/161 μA, with a 0.5-mm aluminium filter and an image pixel resolution of 17.55 µm. Analysis of the resulting reconstructed images was achieved using CTAn (v. 1.9.2.5, SkyScan). Width and depth measurements of cuts were obtained from the μCT scans following the methodology of Thali et al. (2003): 3D μCT volume datasets were used to obtain 2D cross sections that optimally showed the cutting plane of the wound profile, and measurements were taken from these.

image

Figure 2. Schematic depiction of skeletal elements present in burial 38.

Download figure to PowerPoint

Results

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. References

One bone from burial 38 bears what appear to be cut marks. They are five in number and are located on the posterolateral side of the right femur in the subtrochanteric region (Figure 3). The proximal four cut marks run approximately transversely, with the fifth, most distal mark being angled at approximately 30° to the rest. The most proximal mark is about 3 mm long and the remainder about 4 mm long. The surfaces of the marks are weathered, and examination under low-power microscope revealed soil particles in the base of the cuts. These observations indicate that the marks are most likely to have been made in antiquity. There is no sign of new bone formation in or around any of the cut marks. The incisions are wider than they are deep and so may be described as cuts rather than stab wounds, and they are broadly symmetrical in cross section, which suggests the use of a non-serrated blade (Thompson & Inglis, 2009). The CT results reveal that all the cuts penetrate through the full thickness of the cortical bone but do not extend into the cancellous bone. The cut surfaces are non-striated, and the dimensions of the cuts are relatively uniform, with a mean maximum depth of 545 µm (range 478–612 µm) and a mean maximum width of 549 µm (range 497–602 µm). The non-striated cut surfaces also suggest a non-serrated, stiff blade, which, given the archaeological context and the date of the site, is most likely to have been metal (Greenfield, 1999; Thali et al., 2003; Lewis, 2008; Eyers, 2011; Pounder & Sim, 2011). Careful examination of the rest of the skeleton revealed no evidence for further cut marks.

image

Figure 3. Burial 38. Above: posterior surface of the right femur showing cut marks in the subtrochanteric region. Below: cross sections of cuts generated from the micro-computed tomography scan. The micro-computed tomography slices have a pixel resolution of 17.55 µm.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. References

The analysis of the cut marks strongly suggests that they were created around the time of death of the infant (or at least when the bone was still fresh enough to retain its slight elasticity). However, the cut marks are unlikely to be immediately related to the cause of death. Although there is evidence for the practice of infanticide at Hambleden (Mays & Eyers, 2011), the location of the marks, on the posterolateral aspect of the proximal femur, makes it very unlikely that they were inflicted with the intent of killing the infant; there are no major blood vessels in this region. Evidence of the practice of defleshing has been established in other Romano-British skeletal remains, including a second-century AD cranium from Roman Britain (Mays & Steele, 1996). In the present case, however, the lack of any convincing cut marks elsewhere on the skeleton makes it improbable that the marks on the femur were part of a systematic defleshing process. A remaining possibility is that the cut marks were formed during an obstetric operation.

Embryotomy, as a means of managing obstructed labour or removing a dead foetus, could save a woman's life. The practice of embryotomy is documented in the ancient Greek and Roman literature (Jackson, 1988). Possibly as early as the fourth century BC, male physicians were extending the range of their interventions by dealing with births in which there was no possibility of a successful outcome. A detailed description of such an intervention is given in the Hippocratic treatise known as ‘On the Excision of the Fetus’, usually dated to the fifth or fourth century BC and certainly known to Galen, in the second century AD, as a ‘Hippocratic’ work. In fact, only the first chapter of this treatise concerns embryotomy, and it talks about how to proceed with an arm presentation. The arm must be pulled out as far as possible, and then the operator—the male physician—is told to ‘excoriate the upper arm and strip its bone bare: bind a fish skin around two fingers of the hand so that the flesh will not slip away and after that make an incision all around the shoulder and separate it at the joint’. Finally, he makes an incision into the chest cavity to reduce the size and then removes the remainder of the body of the foetus (Potter, 2010, 369–371). Another text on embryotomy, the ‘Gynaecology of Soranus of Ephesus’, from the second century AD, describes using a piece of cloth wrapped around the limb to achieve traction, for both arm and leg presentations, where the foetus is either dead or stuck in the birth canal (Temkin, 1991, 192). Soranus advised the operator to make incisions at the wrists or axillae, with each part being cut as it presents and amputations at the joints always being the preferred method (Temkin, 1991, 194). He described the use of not only a knife but also hooks and crushing devices. Another Hippocratic treatise, ‘On Superfetation’, gives advice on both arm and leg presentation in a living child; in both cases, the limb is to be pushed back in. If both legs present (breech presentation), a vapour bath is recommended to moisten the womb (Potter, 2010, 321). A later chapter describes the use of a ‘claw’—a scalpel blade on a ring—to reduce the bulk of a dead foetus and allow it to be removed in parts (Potter, 2010, 323). In most of these historical descriptions of embryotomy, it is the arms, not the legs, that are seen as causing most difficulty in extracting the body and that need to be removed.

A fourth-century AD foetus from the Poundbury cemetery, Dorchester (Molleson & Cox, 1988; Molleson, 1993), provides independent evidence for the practice of embryotomy by dismemberment in Roman Britain. Like the current case, that example includes a cut at the proximal femur, and the procedures for dismemberment apparently followed in the Poundbury case resemble those described by Soranus (Temkin, 1991, 189–195).

Eighteenth-century accounts by men-midwives of their interventions provide useful comparative material. Although they had an additional instrument to employ—the obstetric forceps—many were in fact reluctant to use it and as a result described similar procedures of wrapping a cloth round a leg and pulling it (e.g. Smellie, 1765, 60; King, 2007). They also used the fillet, a noose that could be placed around parts that included the ankle (Smellie, 1765, 153–154). At this point in history, knowledge of the birth process was such that it was possible to complete a successful delivery from a leg or full breech presentation. It was recognised that, for a large baby, bringing down one leg was unlikely to lead to a successful birth, and William Smellie (1765, 74–75) preferred to push back the leg and take hold of both feet, turning the baby as traction was exerted. If the baby was already thought to be dead, however, there was less need to proceed cautiously. Smellie (1765, 182) recounted a case from 1749 in which the traction was such that the leg was pulled away from the rest of the foetus.

The cut marks on the posterolateral surface of the femur of burial 38 are anatomically consistent with incisions made on the posterior upper thigh of a foetus, perhaps during embryotomy in a breech or leg presentation. This may mean that the cut marks are related to the attempts of a surgeon to remove the foetus to save the life of the mother. The foetus may have already been dead, or dying, in an obstructed labour. Obstructed labour is more common in populations where undernutrition during growth leads to smaller pelves in adult women and is an important cause of maternal death (Neilson et al., 2003). However, there was only one adult female burial from Hambleden (Mays et al., 2011), so no comment can be made regarding female pelvic size at a population level in that group.

In a modern population, breech presentations have been reported to occur in about 6–7% of infants of a similar gestational age to that of burial 38 (Fox & Chapman, 2006). The modern management of breech presentations in developed countries such as the UK is controversial (Hannah et al., 2000), with some criticising the rising rates of planned caesarean section. However, this controversy exists within the context of modern obstetric medicine, where, on the one hand, access to care by clinicians experienced in vaginal breech delivery could make this a relatively safe option, whereas, on the other, modern surgery, with antisepsis, haemostasis and anaesthesia, makes caesarean section a much safer option than it would ever have been in the past. Breech presentation brings with it an increased risk of complications during childbirth, including umbilical cord prolapse and compression and entrapment of the foetal head within the maternal pelvis. Rates of foetal perinatal mortality and neonatal mortality have been reported as around twice to three times higher in pre-term breech compared with cephalic delivery (Malhotra et al., 1994).

Although the cut marks seen in burial 38 do not extend to the proximal extremity of the foetal femur, they could represent adventitious cuts as the surgeon attempted to locate the hip joint of the breech-presenting foetus to disarticulate the legs and ease delivery. Such a procedure may also have left marks on the unmineralised, cartilaginous head of the femur, which does not survive in an archaeological context, or on the pelvis, but these bones are missing from burial 38.

The excavations at the Hambleden Roman villa yielded many high-status finds (Cocks, 1921; Eyers, 2011). Perhaps their wealth enabled the inhabitants, and the wider community, to avail themselves of the best obstetric attention available at the time.

References

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. References
  • Cocks AH. 1921. A Romano-British homestead, in the Hambleden valley, Bucks. Archaeologia 71: 141198.
  • Eyers JE. 2011. Romans in the Hambleden Valley: Yewden Roman Villa. Chiltern Archaeology: High Wycombe.
  • Fox AJ, Chapman MG. 2006. Longitudinal ultrasound assessment of fetal presentation: a review of 1010 consecutive cases. The Australian and New Zealand Journal of Obstetrics and Gynaecology 46: 341344.
  • Greenfield HJ. 1999. The origins of metallurgy: distinguishing stone from metal cut-marks on bones from archaeological sites. Journal of Archaeological Science 26: 797808.
  • Hannah ME, Hannah WJ, Hewson SA, Hodnett ED, Saigal S, Willan AR. 2000. Planned caesarean section versus planned vaginal birth for breech presentation at term: a randomised multicentre trial. The Lancet 356: 13751383.
  • Hillier R. 1992. The striae of Retzius as indicators of disease experience in children. PhD thesis, University of Sheffield.
  • Jackson R. 1988. Doctors and Diseases in the Roman Empire. British Museum: London.
  • King H. 2007. Midwifery, Obstetrics and the Rise of Gynaecology. Ashgate: Aldershot
  • Lewis JE. 2008. Identifying sword marks on bone: criteria for distinguishing between cut marks made by different classes of bladed weapons. Journal of Archaeological Science 35: 20012008.
  • Malhotra D, Gopalan S, Narang A. 1994. Preterm breech delivery in a developing country. International Journal of Gynecology & Obstetrics 45: 2734.
  • Mays S, Eyers J. 2011. Perinatal infant death at the Roman villa site at Hambleden, Buckinghamshire, England. Journal of Archaeological Science 38: 19311938.
  • Mays S, Steele J. 1996. A mutilated skull from Roman St Albans. Antiquity 70: 155161.
  • Mays S, Vincent S, Robson-Brown K, Roberts A. 2011. The human remains. Romans in the Hambleden Valley: Yewden Roman Villa, JE Eyers (ed.). Chiltern Archaeology: High Wycombe; 247258.
  • Molleson T. 1993. The human remains. Excavations at Poundbury 1966–80, Volume 2: The Cemeteries, DE Farwell, TI Molleson (eds.), Dorset Natural History and Archaeological Society Monograph Series 11. Dorset Natural History and Archaeological Society: Dorchester; 142214.
  • Molleson T, Cox M. 1988. A neonate with cut bones from Poundbury Camp. Bulletin de la Société Royale Belge d'Anthropologie et de Préhistoire 99: 5359.
  • Neilson JP, Lavender T, Quenby S, Wray S. 2003. Obstructed labour. British Medical Bulletin 67: 191204.
  • Potter P. 2010. Hippocrates, vol. IX. Loeb Classical Library/Harvard University Press: Cambridge, MA/London.
  • Pounder DJ, Sim LJ. 2011. Virtual casting of stab wounds in cartilage using micro-computed tomography. The American Journal of Forensic Medicine and Pathology 32: 9799.
  • Scheuer JL, Musgrave JH, Evans SP. 1980. Estimation of late foetal and perinatal age from limb bone lengths by linear and logarithmic regression. Annals of Human Biology 7: 257265.
  • Smellie W. 1765. A Collection of Preternatural Cases and Observations in Midwifery, 3rd edn. T. and J. Whitehouse: Dublin.
  • Tanner JM. 1989. Foetus into Man, 2nd edn. Castlemead: Ware.
  • Temkin O. 1991. Soranus' Gynecology. Johns Hopkins University Press: Baltimore.
  • Thali MJ, Taubenreuther U, Karolczak M, Braun M, Brueschweiler W, Kalender WA, Dironhofer R. 2003. Forensic microradiology: micro-computed tomography (micro-CT) and analysis of patterned injuries inside of bone. Journal of Forensic Sciences 48: 13361342.
  • Thompson TJU, Inglis J. 2009. Differentiation of serrated and non-serrated blades from stab marks in bone. International Journal of Legal Medicine 123: 129135.
  • Whittaker DK, Richard D. 1978. Scanning electron microscopy of the neonatal line in human enamel. Archives of Oral Biology 23: 4550.