21 skull and mandibular parameters of Neomys anomalus (five geographic samp
les), N. fodiens (three geographic samples) and N. teres (one sample) were
measured and subjected to statistical analyses. Discriminant Function Analy
sis based on size-out data failed to assort 10.3% of skulls and mandibles a
nd 47.9% of mandibles to their correct geographical sample and led up to 29
.3% of mandibles to be wrongly classified to species. The implications of t
hese findings for palaeoecology and biostratigraphy are discussed.