Since the discovery of the Florisbad human skull fragment in 1932 there has been uncertainty about the age of this unique fossil. At first the skull was thought to be somewhere around 40 000 years old, mainly because that age is the effective limit of the conventional radio-carbon dating method. More recently, with the increased awareness of the importance of the skull for understanding modern human evolution, the consensus has been that it must be at least 100 000 years old or possibly more. An important scientific breakthrough was accomplished recently when Australian scientists co-operated with the Florisbad Quaternary Research Department of the National Museum to establish the first reliable dating framework for the Florisbad human skull (Grün et al. 1996). The skull is dated to around 259 000 years ago, while the oldest levels of the site are dated to about 350 000 years ago.

The significance of these dates is that they support a recent African origin for all modern people.

The Florisbad skull is important because it is the only relatively complete example of the late archaic phase of modern human development in Africa. According to one theory, the skull represents an immediate forerunner population of modern people. During the mid 1980s the German physical anthropologist, Günter Bräuer, constructed a framework for the evolution of modern people in Africa, in which there are two recognisable phases of development out of an ancestral Home erectus form. The first phase is seen in early archaic form of Home sapiens, like the skulls from Kabwe in Zambia, Elandsfontein near Saldanha and some other specimens from East Africa. The Florisbad skull is the only relatively complete representative of the next phase in this development.

This late archaic phase is the immediate forerunner of the first anatomically modern humans, from Klasies River Mouth, an archaeological site on the Tsitsikama coast, southern Cape. This interpretation led Bräuer and other, like Chris Stringer from the Natural Museum in London, to propose an African origin for all modern humans. In this scheme, known as the ‘Out of Africa’ hypothesis, it is suggested that archaic forms of people first evolved into fully moderns in sub-Saharan Africa and then spread out of Africa, presumably via the Middle East.

An important key in this hypothesis is the set of fossil remains from Klasies River Mouth. This material dates to around 100 000 years ago and is by far the earliest evidence of anatomically modern forms of people. In spite of being quite fragmentary it clearly shows modern features, such as the absence of heavy ridges above the eyes and the presence of prominent chins. The ‘Out of Africa’ hypothesis further suggests that modern people, having first appeared in sub-Saharan Africa, caused the displacement of existing archaic populations in Europe and Asia. These previously-existing archaic people, such as the well-known Neanderthal people, were the descendants of an initial emigra-tion event from Africa, probably around 1.5 million years ago.

An opposing theory sug-gests that there was continuous development of people from archaic to modern in various parts of the Old World. This theory is based in part on the fact that there was local evolution in some parts of the Old World. In Western Europe the well-known Neanderthal people were descen-dants of an initial emigration of Homo erectus from Africa. In this scenario, known as the ‘Multi-regional Hypothesis’, there would have been no re-placement of archaic forms, but local continuity in the evolution of people in various parts of the Old World.

There are a number of issues in this debate. One of them is whether it can be demonstrated that there was continuous evolution in other parts of the Old World, besides in Africa. For example, in Western Europe the transition from Neanderthal to modern is very abrupt and seems not to support continuous evolution. Another issue, and possibly one of the most important questions in this debate, is the age of the African material. If it should transpire that the African fossils are much older than comparable specimens from elsewhere, it would give support to the idea of an African origin of modern people.

Apart from limitations of the radio-carbon dating method, which is ineffective in dating material older than ca. 40 000 years old, a major obstacle in determining the age of the Florisbad human skull and other fossils was the uncertainty concerning the sedimentary context of this material. During the early and mid 1980s one of the aims of the Florisbad Quaternary Research Development of the National Museum was to reconstruct the context of all previous fossil finds from Florisbad. This was achieved successfully and it could be demonstrated that the human skull and other mammalian remains, which are mineralized in a similar way, derived from ancient spring eyes. These fossils resulted from carnivores hunting and scavenging around the ancient spring pools. As is usual in such situations, parts of carcasses remained intact, became incorporated into the springs and were preserved as fossils (Brink 1987, 1988). These ancient spring eyes are preserved as vertical columns of white sand that interrupted the horizontal layers of the site. This makes it difficult to relate the fossil contents of the spring eyes to the natural stratigraphy, as there is no continuity between these two components of the site. The problem was further complicated by the unconventional excavating techniques of previous workers.

An understanding of the sedimentary context was critical in establishing an age for the human skull, since previously it was erroneously thought that the skull derived from one of the lower horizontal organic units (Peat l). Another important aspect in the planning of the dating exercise was to establish a good sedimentary column which could be used as a basic frame of reference for all other parts of the deposit. For this purpose a test pit known as the third test pit, was excavated.

In 1991 I contacted Dr Rainer Grün of the Quaternary Dating Research Centre at the Australian National University, Canberra, Australia. He is one of the leading figures in applying a new dating technique, known as Electron Spin Resonance (ESR). This method was developed over the last two decades and uses fossil tooth enamel to establish dates. This technique can be applied in conjunction with another technique, Optically Stimulated Luminescence (OSL), which measures the light emis-sion of trapped electrons in quartz crystals, i.e., in sand.

We planned a dating experiment for Florisbad and consequently Dr Grün visited Florisbad in 1993, when most of our sampling was done. Follow-up visits in 1994 and 1995 allowed us to sample the site comprehensively. The OSL/ESR dating exercise on the Florisbad sediments and fossil teeth included 32 sediment for OSL measurement and 33 tooth enamel samples for ESR measurement. The section of the third test pit at Florisbad was used as a control for the site as a whole, as it seems to have had the least disturbance and probably covers the complete time range during which sedimentation occurred at Florisbad. Most of the OSL samples and 13 tooth enamel samples for ESR were taken from this excavation. In com-bination with this, sediment samples and tooth specimens from the extended MSA horizon and two additional test pits were used. We also used 16 enamel samples from hippopotamus and wildebeest teeth from the spring material obtained in previous excavations. The human skull also derives from the springs.

Although the third test pit provided good results for both ESR and OSL measurements, ESR measurements on spring material were disappointingly divers. We found that both hippopotamus and wildebeest teeth from the springs varied in age from about 100 000 years to 350 000 years. These estimates were of course far too imprecise to allow any new insight but agreed with the general age span of the site based on both ESR and OSL from the third test pit. This gave us confidence that the methods were reliable.

However, the imprecise time bracket for the skull, as suggested by the proxy dates for the hippopotamus and wildebeest teeth, was a major impediment. In order to solve this we decided to remove a small sample of enamel from the only tooth found with the human skull. Two fragments of tooth enamel were measured for ESR without destroying the tooth fragments. This was a revolutionary step, since it has never been done before on any other human fossil.

· Grün, R., Brink, J.S., Spooner, N.A., Taylor, L., Stringer, C.B., Franciscus, R.B. & Murray, A. 1996. Direct dating of the Florisbad hominid. Nature 382: 500–501.

· Brink, J.S. 1987. The archaeo-zoology of Florisbad, Orange Free State. Memoirs van die Nasionale Museum 24: 1–151.