The accuracy of determination of the sex of skeletal remains varies with the age of the subject, the degree of fragmenta- tion of the bones and biological variability (Table 3.1).
Particularly when studying the skull and pelvis, a subjective impression by the experienced observer defies complete analy- sis, yet objective measurement may be no more accurate: in other words, the processes comprise morphological traits ver- sus morphometry. The determination of sex is statistically the most important criterion, as it immediately excludes approxi- mately half the population whereas age, stature and race each provide points within a wide range of variables. Obvious sex differences do not become apparent until after puberty,
usually in the 15–18-year period, though specialized measure- ments on the pelvis can indicate the sex even in fetal material.
Sex and age are linked, especially where body size and weight are concerned. Similarly, race confuses sexing, for example, the size of the supraorbital ridges in a normal negroid female may exceed those in the average Caucasian male.
The accuracy of sexing is hard to estimate, as various loading factors exist. Krogman comments that he scored 100 per cent accuracy using the whole skeleton, 95 per cent on pelvis, 92 per cent on skull, 98 per cent on pelvis plus skull, 80 per cent on long bones and 98 per cent on long bones plus pelvis. He admitted, however, that, as most anatomy department material has a sex ratio of about 15:1 in favour of men, marked bias could be introduced by assigning all doubtful bones to the male category.
Stewart records that, for the whole skeleton, one can expect a 90–95 per cent success rate and for the skull alone only 80 per cent, but, if the mandible is also present, this rises to 90 per cent (Hrdlicka 1939). In general, adult female skeletal measurements are about 94 per cent that of the male of the same race, but different measurements may vary from 91 to 98 per cent.
The skull
The following features develop after puberty and are modi- fied by senility, so are applicable only between about the 20th and the 55th year. Age as well as race has a profound effect.
■ General appearance.The female skull is rounder and smoother than the rugged male.
■ Size.Male skulls are larger, with an endocranial volume some 200 ml greater.
■ Muscle ridgesare more marked in male skulls, especially in occipital areas where larger muscles are
TABLE3.1 Pfitner’s table of bodily dimensions in the female compared with the male (per cent of male dimensions). The common generalization that the female is 94 per cent of the male size varies in different areas of the body
Stature 93.5 Arm length 91.5
Head breadth 98.0 Sitting height 94.5
Face breadth 94.0 Head circumference 96.0
Face height 90.0 Head height 96.0
Head length 95.5 Leg length 93.0
Anterior nasal spine Nasion Glabella
Bregma
Zygomaxillare Gonion
Gnathion Infradentale
Alveolare
Lambda
FIGURE3.5 Anatomical landmarks of the skull.
The determination of sex
attached to the nuchal crests and in temporal areas for larger masseter and temporalis muscles.
■ Supraorbital ridgesare more marked in male skulls and may be absent in female. The glabella (central forehead eminence above the nose) is small or absent in the female, and prominent in the male, though this is a poor discriminant.
■ Mastoid process.This is larger in male skulls (see Hoshi).
■ Frontal and parietal eminences.These are more prominent in female skulls, which resemble the shape in an infant more than in a male.
■ Palate.This is larger and of a more regular U-shape in men. The smaller female palate tends to be parabolic.
■ Orbitsare set lower on the face in the male skull, with more square and less sharp edges (especially upper edge) than the female.
■ Nasal aperture.This is higher and narrower in the male skull and has sharper edges. The nasal bones are larger and project further forward to meet at a more acute angle than in the female.
■ Forehead.This is high and steep in the female skull, with a more rounded infantile contour than the male.
■ Teethare smaller in the female skull, molars usually having four cusps. The male often has a five-cusp lower first molar.
■ Zygomatic process.The posterior ridge projects back beyond external auditory meatus in the male skull. The zygomatic arches bow outward more than the female, where they remain more medial.
■ Mandible.This is large in the male skull, with a squarer symphysis region. Female jaws are more rounded and
FIGURE3.7 Frontal view of male and female skull. The male (left) is larger, more massive and has heavier eyebrow ridges.
The chin is squarer and the mandible heavier, The female has relatively larger height orbits, a smoother cranium, and fuller frontal and parietal eminences.
FIGURE3.6 Two recovered and reassembled skeletons. After cleaning, the bones should be laid out in correct anatomical position, to gain an approximate direct measure of the stature, allowing as well as possible for lost soft tissue. The upper skeleton, obviously male, shows the frequent loss of small hand and foot bones, but no injuries. The lower remains, from a hidden 40-year-old homicide, reveal that the body was sawn into six sections, there being cuts above the knees and through spine, scapula and humeri.
project less at the anterior point. The vertical height at the symphysis is proportionately greater in the male. The angle formed by the body and the ramus is more upright in the male, being less than 125°. The condyles are larger in the male skull, as is the broader ascending ramus, and there is a more prominent coronoid process.
These sex variations represent the ‘typical’ White and, to a great extent, Asian skull. There is considerable overlap,
especially in subjects from the Indian subcontinent, where osteological sex differences are much less marked. The criteria set out above exclude prepubertal and senile persons, and are less valid for those outside the 20–55 age group. For methods of sexing by discriminant function analysis, the work of Giles and Elliot (1963) should be consulted.
In recent years craniometry has been applied to sexing the skull and multiple accurate measurements between discrete anatomical points have been used to produce discriminant FIGURE3.9 A typical female skull, from a murder victim hidden for 40 years in a cave. The cranium is high, round and smooth with insignificant ridges for muscle attachment. The mastoid is tiny, one of the best discriminants. The post-zygomatic ridge does not continue behind the auditory meatus and there are no supraorbital ridges.
The chin is round and smooth.
FIGURE3.8 A typical male skull, with a sloping forehead and prominent occipital ridges for large muscle attachment.
Probably the best male feature is the large mastoid process; also, the zygomatic process extends well behind the auditory meatus.
function analysis. The papers of numerous authors must be consulted for details of this sophisticated and difficult technique that is accurate in 83–88 per cent of cases. The frequency of correct sexing is no greater than more subject- ive methods, but commands a greater degree of confidence level in respect to its accuracy.
Sex characteristics in the pelvis
The post-pubertal female pelvis is wider and shallower than the rather upright male girdle to allow for the passage of the fetus during childbirth. Though as always there is an overlap of the appearance and measurements in the least characteris- tic individuals, the usual variations are sufficient to allow sex- ing of the adult pelvis to be made with about a 95 per cent confidence rate (see work by Genovés). In the pelvis, unlike the skull, there are differences, albeit subtle, in immature (even fetal) pelves that allow sexual differentiation. The fol- lowing features provide the most useful criteria. One must not be used in isolation, however; as many as possible must be assessed together. Sex features are independent of each other and one may even contradict the other in the same pelvis. It is often a first subjective impression by an experi- enced eye that determines the answer, though where the sex differences are slight, careful measurements may be needed to resolve the problem.
As in the skull, the male pelvis is more rugged because of the attachment of stronger muscles. It stands higher and more erect than the smoother, flatter, female pelvic girdle.
The subpubic angle, measured at the medial intersection of two lines drawn along the best approximation of the lower border of both inferior pubic rami, approaches 90°in the female pelvis, but usually about 70°in the male.
This is often a subjective measurement, however and depends in turn upon the shape of the pubic bone itself.
When the line of the inferior ramus is projected medially and intersected with a horizontal line laid across the upper border of the superior ramus, the reverse size of angle is seen, the male being wider than the female.
The body of the pubic bone, the block lateral to the symphysis, tends to be triangular in shape in the male, whilst the female pubis is more rectangular (Phenice 1969; Iscan and Derrick 1984). Certain sexual variations in the pubis have been used by Phenice. In the female these are:
■ a bony ridge (‘ventral arc’) running down the ventral surface from the pubic crest
■ a concavity of the lower margin of the inferior pubic ramus immediately lateral to the lower border of the symphysis
■ a ridge of elevated bone on the medial aspect of the ischiopubic ramus, immediately lateral to the symphysis; in the male this area is broad and flat.
The ‘ischiopubic index’ devised by Washburn may be helpful, in which the pubic length (100) is divided by the ischial length. The measurements must be carefully made, the pubic length being from the plane of the symphysis to the reference point in the acetabulum and the ischial length being from the same point to the most distal edge of the ischium. The reference point is the site of fusion of the three elements of the immature innominate bone, usually marked by a notch in the articular surface of the acetabulum (Schultz 1937). If the ischiopubic index (in White races) is less than 90, the pelvis is male; if over 95, it is female. The acetabulum is larger in the male, being an average of 52mm in diameter, compared with 46mm in the female. The male joint cup also faces more laterally than that of the female, which tends to look more forward. Naturally, acetabular size is related to that of the femoral head, which will be considered later.
The greater sciatic notch is an important criterion, being deep and narrow in the male, wide and open in the female.
The angle formed by the margins approaches closer to a right angle in the woman than it does in the man. Harrison (1968) and Hrdlicka (1939) both felt that the greater sciatic notch was one of the best discriminants for sex, the latter claiming a 75 per cent success rate using this criterion alone.
The obturator foramen is somewhat ovoid in the male, but triangular in the female. The pre-auricular sulcus, which marks the attachment of the anterior sacroiliac ligament, lies just lateral to the sacroiliac joint and is usually well marked in women but often absent in males. The pelvic inlet, looked at from above, is more circular in the female, the male being heart-shaped as a result of the protrusion of the sacrum into the posterior brim (Greulich and Thomas 1938). A number of other pelvic ‘indices’ have been devised by various authors (see, for example, the work of Turner, Greulich and Thomas, Caldwell and Molloy, Straus, and Derry).
The determination of sex
Female Male
FIGURE3.10 Sex differences in the human pelvis. The most noticeable features are the narrower suprapubic angle and the higher iliac blades in the male pelvis.
Sex characteristics in the sacrum
The sacrum is functionally part of the pelvis and shares in its sex variations. The female sacrum is wide and has a shal- low curve, again related to the larger pelvic canal for child- birth. It is shorter in the female and the curve is limited almost entirely to that distal part below the centre of the third sacral vertebra. The male sacrum may have more than five segments, which is rare in the female. The curve in the male is continuous down the whole bone and there may even be a slight forward projection of the coccyx. Fawcett (1938) compared the transverse diameter of the first sacral vertebra (CW) with that of the base of the sacrum (BW).
The formula CW100/BW averaged 45 in the male and 40 in the female. Kimura (1982) has developed a
‘base wing index’, where the relative widths of the wing and base provide discriminant function coefficients for sex determination.
Sexing from the long bones
The femur is the most useful, its length and massiveness themselves being significant. There is, as usual, consider- able overlap of all long-bone sex characteristics, but Brash’s series showed that the maximum (oblique) length in the male femur was around 459mm, while that of the female was only 426mm. Other figures from Pearson and Bell (1917) suggested mean values of 447mm for men and 409mm for women. Using the trochanteric oblique length, they suggested a range of 390–405mm for women and 430–450mm for men, though there was the usual overlap in the middle. Race and nutrition (which is related to the era and the place in which the samples were obtained) must be allowed for in such measurements.
The size of the femoral heads is claimed to be a better dis- criminant of sex, the vertical diameter being said by Pearson and Bell to be greater than 45mm in the male and less than 41 mm in the female, though again there is an overlap in the distribution curves around the 43mm size (Table 3.2).
Maltby (1917), however, measured 43–56mm in the male and 37–46mm in the female. Femoral head size is part of Pearson’s ‘mathematical sexing of the femur’, which incor- porates several measurements (Table 3.3). Dwight (1904) studied the size of both femoral and humeral heads, claim- ing that they were more useful than bone length. Once again, discriminant function sexing has been carried out, using a number of measurements. Details should be sought in the work of Black and of Iscan and Miller-Shaivitz.
Another sex trait in the femur is the angle that the shaft makes with the vertical. Because the pelvis is relatively wider in the female, the shafts have to slope more to con- verge at knee level, so that the condyles at the lower end of TABLE3.2 Dwight’s table for sexing from humeral and femoral head diameters (in mm)
Vertical Transverse Vertical
humeral humeral femoral
Female 42.67 36.98 43.84
Male 48.76 44.66 49.68
Difference 6.09 5.68 5.84
TABLE3.3 Pearson and Bell’s table for the ‘mathematical sexing of the femur’. The area of uncertainty extends in increasing degrees of confidence outwards from the central column to the more definite limits on each side. Measurements are given in millimetres
Male Male or female Female Vertical diameter of head 45.5 43.5–41.5 41.5
Popliteal length 145 114–132 106
Bicondylar width 78 74–76 72
Oblique trochanteric length 450 405–430 390 FIGURE3.11 The greater sciatic notch is narrower in the male
innominate bone.
Female Male
FIGURE3.12 The variation in shape of the male and female sacrum. The female is broader and more triangular.
the femur sit horizontally on the tibial plateau. Thus when a female femur is seated on its condyles on a flat surface, the angle the shaft makes with that surface is of the order of 76°, while a male bone is more upright, the angle being around 80°. The angle of the neck on the shaft of the femur (the ‘collodiaphyseal angle’) was studied by Godycki (1957), the results suggesting that a bone with an angle of less than 40°had an 85 per cent chance of being male, while if the angle exceeded 50°, there was a 75 per cent chance of it being from a female (Table 3.4).
Most workers have worked with dry bone specimens;
when methods using fresh bone are used, allowance must be made for articular cartilage where relevant. For example, the vertical diameter of a femoral head is 3mm less in the dried specimen.
Sex determination from other bones
The sternum may be helpful in that the length of the manubrium in the female may equal or exceed half the length of the body, while the manubrium of the male is less than half the body length (Table 3.5). This was claimed in the nineteenth century by Hyrtl, but was denied by Krogman and by Dwight. The latter claimed that the ratio of manubrium: body was 52:100 in women and 49:100 in men, a poor discrimination. However, the method has been rehabilitated by Iordanidis (1961), who alleged a suc- cess rate of 80 per cent using the sternum alone. Stewart
and McCormick (1983) used a radiographic technique and claimed total accuracy in stating that a sternal length of less than 121 mm must be female and over 173mm must be male.
The scapula has been studied extensively, but mostly in relation to age. There are relatively poor sex variations in the vertical diameter of the glenoid cavity: the threshold, according to Dwight, is 36mm, those smaller being female.
Iordanidis made an extensive study of scapular measure- ments and concluded that scapular height was the best discriminant, the male usually being greater than 157mm, the female less than 144 mm.
The humerus, radius and ulna yield little helpful sexing evidence, apart from overall size. The presence of a perfor- ated olecranon fossa at the lower end of the humerus occurs more often in females and more commonly on the left side, there being a 3.7:1 ratio compared with males. Godycki studied this and other characteristics of the arm bones as sex determinants, but their value is poor.
The determination of sex
TABLE3.4 Long-bone lengths (mm) in White men and women of different stature
Men Women
Humerus Radius Ulna Stature Femur Tibia Fibula Humerus Radius Ulna Stature Femur Tibia Fibula
295 213 227 1530 392 319 318 263 193 203 1400 363 284 283
298 216 231 1552 398 324 323 266 195 206 1420 368 289 288
302 219 235 1571 404 330 328 270 197 209 1440 373 294 293
306 222 239 1590 410 335 333 273 199 212 1455 378 299 298
309 225 243 1605 416 340 338 276 201 215 1470 383 304 303
313 229 246 1625 422 346 344 279 203 217 1488 388 309 307
316 232 249 1634 428 351 349 282 205 219 1497 393 314 311
320 236 253 1644 434 357 353 285 207 222 1513 398 319 316
324 239 257 1654 440 362 358 289 209 225 1528 403 324 320
328 243 260 1666 446 368 363 292 211 228 1543 408 329 325
332 246 263 1677 453 373 368 297 214 231 1556 415 334 330
336 249 266 1686 460 378 373 302 218 235 1568 422 340 336
340 252 270 1697 467 383 378 307 222 239 1582 429 346 341
344 255 273 1716 475 389 383 313 226 243 1595 436 352 346
348 258 276 1730 482 394 388 318 230 247 1612 443 358 351
352 261 280 1755 490 400 393 324 234 251 1630 450 364 356
356 264 283 1767 497 405 398 329 238 254 1650 457 370 361
360 267 287 1785 504 410 403 334 242 258 1670 464 376 366
364 270 290 1812 512 415 408 339 246 261 1692 471 382 371
368 273 293 1830 519 420 413 344 250 264 1715 478 388 376
Modified by Krogman and Iscan (1986) from Hrdlicka (1939).
TABLE3.5 Sex determination from the sternum. Mean (range) lengths of manubrium and sternal body (mm)
Male Female
Manubrial length 51.7 (41–73) 48.4 (39–61)
Body length 95.4 (74–122) 78.6 (59–95)
Combined length 147 (131–180) 127 (107–140)
Taken from Jit et al.(1980) North Indian population.
There are many reports on sexing from limb and girdle bones and the best approach seems to be a multiple assess- ment using the data for discriminant function analysis.
Evidence of pregnancy from the skeleton
More accurately, parturition causes some changes in the pelvis as a result of the local trauma of childbirth, which is reinforced by multiple pregnancies. These include ‘pubic scars’ from the tearing of tendon insertions and periosteum around the pubic bone. The dorsal pubic surface and the pre-auricular sulcus are best indicators, but most authorities agree that it is not possible to determine the number of births from osteological appearances. The papers of Angel and of Ullrich are most useful in this respect.