Skeletal remains from the Kuala Selinsing excavations, Perak, Malay Peninsula
Gordon Harrower, Professor of Anatomy at Singapore, published this osteological analysis of skeletal remains recovered from canoe burials at the Kuala Selinsing site in Perak in 1933. Drawing on comparative craniometry and appendicular measurements, Harrower argued that the buried individuals were of Negrito or Protomalayan origin rather than Malay or Indian, a finding that complicated the interpretation of the site’s associated cultural material.
Summary
Harrower received three groups of skeletal remains (designated Associations 1, 2, and 3) from I. H. N. Evans, Ethnographer at the Perak Museum, for anatomical examination. The material was fragmentary in all three cases, with the skulls particularly badly crushed, requiring careful reconstruction from numerous pieces. The central problem Harrower addressed was racial attribution: the cultural assemblage from the Kuala Selinsing excavations included objects of clear Indian origin (a seal bearing a Pallava inscription, cornelian and agate beads) alongside items of probable Malayan or Bornean affinity, yet the skeletal evidence pointed in a different direction entirely.
The comparative methodology relied heavily on the recently published measurements of three Semang skulls by Schebesta and Lebzelter (1926), which provided the closest available reference point for a Negrito cranium from the region. Harrower systematically compared linear measurements, cranial indices, orbital and nasal indices, and appendicular proportions against reference values for Malay, Semang, Dyak, Tamil, Javanese, Fukien, and Burmese skulls. The Association 2 skull, in particular, showed a constellation of features—low vault, broad flattened nose, rectangular orbits, microcephalic capacity, dolichokerkik and dolichocnemic limb proportions, and pronounced platycnemism—that aligned it closely with the Semang type and decisively excluded it from the brachycephalic Malay or mesorhine Tamil categories.
The Association 1 skull shared the same morphological type but at a significantly larger size, leading Harrower to propose it represented a hybrid individual, possibly a cross between Semang and Dyak (the latter being classified by Keane as Protomalayan). The Association 3 remains, from an adolescent, were too scanty for independent assessment, though the mandible and long bones confirmed affinity with the other two groups. Harrower acknowledged a tension between the skeletal evidence and the cultural assemblage, suggesting the burials might predate or postdate the Indian-influenced material, given the nomadic existence of the peoples involved.
Key Findings
- The Association 2 skull had an estimated cranial capacity of 1,236.5 ccms (via Manouvrier’s formula), a cephalic index of 76.98, and a nasal index of 66 (markedly platyrhine), placing it in the microcephalic, mesaticephalic, platyrhine group most closely comparable to Schebesta’s Menri Semang skull (capacity 1,270 ccms, cephalic index 77.7). (pp. 195–198)
- The Association 1 skull measured 190 mm in maximum length with a cephalic index of 75.0, substantially larger than the Association 2 skull (176 mm) but sharing the same low-vaulted, square-faced, platyrhine morphology; Harrower classified it as a probable Semang–Dyak hybrid. (pp. 198–200)
- Both Association 2 and Association 3 individuals were of pigmy stature, estimated at approximately 1.54 m, while Association 1 was estimated at 1.60 m; the presence of two pigmy individuals among three skeletons was taken as evidence of a dominant Negrito element rather than an adventitious occurrence. (pp. 204–205, 208)
- The Association 2 tibiae showed pronounced platycnemism (indices 70.9 and 62.5), consistent with habitual squatting, and the radio-humeral index of 81.7 placed the individual in the dolichokerkik group—a primitive trait shared with Fuegians, Andamanese, and simians. (pp. 204–205)
- The Association 2 skull exhibited extreme dental wear with complete loss of enamel and exposure of ground-smooth dentine on all occlusal surfaces, a condition Harrower also observed in a modern Hylam Chinese skull but could not attribute to any identifiable dietary cause. (pp. 186–187)
- The Association 2 individual displayed complete osseous sacralisation of the fifth lumbar vertebra, a prenatal condition in which the transverse processes fused to the sacral alae, along with chronic spondylitis affecting the lumbar vertebrae. (p. 187, n. 5)
Conclusion
Harrower’s definitive conclusion was that the Kuala Selinsing canoe burials contained individuals of Negrito or Protomalayan origin, not Malay or Indian, and that the skeletal and cultural evidence likely did not belong to the same temporal horizon. He identified the Association 2 individual as most closely allied to the Semang, the Association 1 individual as a probable Semang–Dyak hybrid, and rejected any suggestion of Indian or pure Malay ancestry on the basis of the cranial and appendicular measurements taken together.
Context
- The primary comparative reference was the 1926 publication by Schebesta and Lebzelter on three Semang skulls from Menri, which at the time represented the only published Negrito craniometric data from the Malay Peninsula region.
- Evans’ appended note (pp. 209–210) contextualises the finding within his broader interpretation of the Kuala Selinsing site, proposing that Proto-Malayan sea-roving peoples may have conducted the bead trade across the Malay Archipelago under Indian influence, with Negritos present as dependants or through intermarriage—a hypothesis that reconciles the Indian cultural material with the Negrito skeletal remains.