CONCLUSIONS

Estimating concavity descriptively is subjective and may yield quite different assessments depending on the researcher. The method described in this paper gives fairly reliable, objective results. Processing a large number of profiles (over 500 in this paper) would have been almost impossible without computer programs especially written for this task. However, even though we automated the processing, others can reliably calculate comparable concavity parameters with a commercial program, like Corel Draw or Cadcam. We applied this method to glirid molars, but in principle it could be applied to any concave or convex surface with little modification.

Our quantitative indices of concavity may require taxonomic revision of glirids at the supraspecific level because our results were often at odds with the comparative assessment of concavity by previous workers in the field. But it is too early to make a more certain statement because our data for most species were based on only one specimen per dental element. Intraspecific variability in concavity is so far known in one case only.

The relation between concavity and occlusion requires further attention.