16 Joshua Laerm, M. A. Menzel, and J. L. Boone 



larger than the fourth (French 1980a, Junge and Hoffman 1981, Jones 

 et al. 1991). The latter character is frequently considered to be diagnostic 

 (e.g., Hall 1981). However, numerous authors (Miller 1895; Jackson 

 1928; Kellogg 1939; French 1980a, 1980/?, 1980c; Junge and Hoffmann 

 1981) point out that this is not always the case. French (1980a, 19806) 

 reported that 20% of S. longirostris examined in Alabama and Georgia 

 and 12% of those in Indiana were characterized by third and fourth 

 upper unicuspids that were equal or nearly equal in size. Similarly, 

 some populations of S. cinereus exhibit third unicuspids that are smaller 

 than the fourth. For example, Bole and Moulthrop (1942) described 

 S. c. ohioensis, in part, on the basis of the third unicuspid being 

 smaller than the fourth. Elsewhere, Kellogg (1939:251) suggested the 

 synonymy of S. fontinalis (now regarded as a subspecies of S. cinereus; 

 see van Zyll de Jong and Kirkland 1989) with S. longirostris concluding 

 that "...the supposed distinctions between S. longirostris and S. fontinalis 

 are nothing more than individual variation." 



Qualitatively, S. longirostris and S. cinereus are not difficult 

 to distinguish; as Jones et al. (1991:265) point out, "...under visual 

 examination... skulls of the two species differed markedly, S. longirostris 

 has a strongly arched palate and shorter rostrum, and the first two 

 unicuspids are of larger diameter than the third and fourth. S. cinereus 

 has a flat long palate and unicuspids of relatively uniform diameter." 

 Unfortunately, qualitative comparisons are often frustratingly difficult 

 to apply in the absence of a good comparative series. Jones et al. 

 (1991) noted that S. cinereus and S. longirostris were so similar morpho- 

 logically that they were not able to use S. cinereus as an out-group 

 in their study of geographic variation of S. longirostris. 



French (1980c) made quantitative comparisons between the two 

 species using a univariate statistical analysis of cranial measurements 

 of 162 S. cinereus and 110 5. longirostris from Virgo County, Indiana. 

 He concluded that S. cinereus and S. longirostris were morphologically 

 similar and that no single character was 100% diagnostic in distinguishing 

 them. Although 13 standard body and cranial measurements differed 

 significantly between S. cinereus and S. longirostris, none was characterized 

 by non-overlapping ranges. Univariate morphological comparisons in 

 Greer's (1989) study of seven cranial measurements indicated significant 

 differences between the two species for six out of seven characters 

 in Missouri; however, as in the French (1980c) study, there was considerable 

 overlap. 



We are not familiar with a published study of a multivariate 

 morphometric comparison of the two species. The purpose of this paper 

 is to examine the effectiveness of both univariate and multivariate 



