Abstract: The occipital torus refers to a thickened band of bone usually between the superior and supreme nuchal lines. Since Weidenreich published his study in 1940, the size of the occipital torus has been thought to decrease over the course of human evolution. A strong torus can be seen in crania of Homo erectus, whereas a relatively weak one is presented in Homo sapiens. A well-developed occipital torus therefore may be useful in the diagnosis of Homo erectus.
Recent finds and dating results in China indicate that extensive variation occurred in the development of the occipital torus during the period 190-230 ky B.P. Thus, some authors have suggested the co-existence of Homo erectus and Homo sapiens in China during that time. However, it is also possible that the occipital torus is not a genetic marker, but rather may be an adaptive skeletal response to veriable nuchal muscle action.
In each skull the degree of occipital torus development was recorded and 14 measurements were taken reflecting cranial and occipital forms. The degree of occipital torus development was divided into four categories:
0: No torus. Supreme nuchal line (SMNL) cannot be identified. Superior nuchal line (SNL) may or may not be discernible. The surface between SMNL and SNL is smooth and level with the occipital and nuchal planes.
1:Slight elevation between SMNL and SNL, and SMNL shows a trace only. Alternatively, two small rounded elevations, one on right and left sides, appear between SMNL and SNL, and the SMNL presents its median portion only.
2:Moderately developed elevation between SMNL and SNL. The SMNL and SNL are distinct. The SNL extends laterally toward asterian.
3: Strong elevation with distinct demarcations (SMNL and SNL). The SMNL and SNL are pronounced and elevated. The SNL extends laterally toward asterian.
Results of these observations are given in Table 1. Testing the occurrence of each degree of occipital torus development, x2 tests indicate that there are no significant differences between different populations in either the male or the female samples (P>0.05). There is, however, a significant difference between male and female samples within populations (P<0.01).
To examine the correlation between occipital torus development and cranial or occipital form, statistics for the 14 measurements are given by sex and population. Analysis suggests that only occipital angle and nuchal plane length may be correlated with torus development within male or female samples of some (but not all) populations (Tables 2 and 3). This correlation is not widespread, and the development of the occipital torus is not necessarily affected by overall cranial and occipital form.
To summarize, the incidence of each category of occipital torus development appears to be identical across populations, controlling for sex. Moreover, the incidence is not strongly affected by geographic (environmental) differences in male or female samples of populations having the same racial affinity. Genetic factors thus may be operating in the development of the occipital torus. In this sense, occipital torus size may be considered a genetic marker.Wide variation in the development of the occipital torus can be traced back to Homa erectus. Compared especially with torus shape, the development (size) of the torus does not appear to be an effective taxonomic character, although genetically controlled.
This work was funded by SFPPAS of Academia Sinica., the Office of Fellowships and Grants of the Smithsonian Institution, and a grant from the American Museum of Natural History. The authors wish to express their grateful thanks for helpful suggestions to Prof. M.H. Wolpoff, Prof. E. Delson, and Dr. I. Tattersall, and for valuable assistance to Dr. D.R. Hunt and J.B. Clark.

Key words: Occipital torus; Variation; Homo erectus; Homo sapiens