It was recognized early on that the teeth identified here as non-hominoid had an unusual morphology for primate canines, but their hominoid status never seems to have been seriously questioned. The first such specimen, GSI D. 192, was described by Pilgrim (1915, plate 2, figure 3). It was recovered from the Chinji Formation, in the general vicinity of the village of Chinji in the Potwar Plateau of present-day Pakistan (the provenance of specimens has been critical to their impact on Sivapithecus taxonomy). Pilgrim said of this specimen, “Though apparently primate, it differs considerably from that of any other genus so far known by the presence of a very distinct cingulum at the postero-internal corner of the tooth” (Pilgrim 1915, p. 48). Interestingly, Pilgrim did not compare this tooth with the canine in the type specimen of Palaeopithecus (now Sivapithecus) sivalensis, GSI D. 1 (Lydekker 1879), which is broken and preserves only the base of the crown. The provenance of the latter is not known with certainty, but it almost certainly comes from the younger Dhok Pathan Formation (J. Barry, personal commun., 1986).
Later, Pilgrim did compare GSI D. 192 to a second, minimally damaged canine of Sivapithecus, GSI D. 196 (Pilgrim 1927, figure 1). Like the canine in the S. sivalensis type specimen, this tooth was also in place in a maxilla so its assignment to Sivapithecus was unambiguous. The new specimen was from Haritalyangar, in present-day India, from sediments that are also younger than those in the Chinji area of the Potwar Plateau. Pilgrim recognized that this newer specimen was very different morphologically from GSI D. 192, and that its greatest resemblance (among both living and fossil primates) was to the partial canine preserved in GSI D. 1. He described GSI D. 196 as having a convex buccal surface, a rounded mesial outline interrupted by a deep sulcus, a concave disto-lingual surface, and being covered with coarse wrinkles. D. 192, by contrast, he described as being only slightly convex buccally and having only a slight sulcus. He therefore assigned D. 192 to a different genus, Dryopithecus, which was then considered to be present in the Siwaliks. Thus, it was also recognized early on that there were two distinct morphologies within the presumed hominoid upper canine sample from the Siwaliks.
For ease of reference in the ensuing discussion, the misidentified non-hominoid canines will be referred to as the “atypical” or “unusual” canines (Table 1), whereas the genuine hominoid canines will be referred to as “typical” hominoid canines, or simply as “Sivapithecus canines,” since several, like the Sivapithecus type specimen, are in maxillae assignable to Sivapithecus.
A second atypical canine, YPM 13809, also from the Chinji area and nearly identical in morphology to GSI D. 192, was described by Lewis (1934, plate 1, figure 3). Lewis also recognized the morphological differences between these and the typical Sivapithecus canines and agreed with Pilgrim that the atypical canines did not belong to Sivapithecus. With regard to assigning them to Dryopithecus, he noted only that the Indian material regarded to be Dryopithecus was quite different from Dryopithecus specimens from Europe.
Two additional atypical canines, K 23/212 (now GSI D. 307) and K 22/448 (now GSI D. 308) were described shortly thereafter by Gregory et al. (1938, plate 2, figures 1, 2, A, B; plate 7, figures A-C). GSI D. 307 is from the Chinji Formation, whereas GSI D. 308 was described as being from older, Kamlial Formation sediments, but was shown in tables 2 and 4 of Gregory et al. as being from “(?) Chinji” (no other hominoid specimens have been reported from the Kamlial Formation). Their discussion of the morphology of these canines is terse and somewhat ambiguous, as they were clearly more focused on size than on morphology.
In the same publication, Gregory et al. (1938) described the first (and still the only) female upper canines of Sivapithecus, from the associated dentition K 29/466 (now GSI D. 299/300) from Haritalyangar (Gregory et al. 1938, plate 1, figure 1; plate 2, figures C-E). Although they acknowledged that the greatest morphological similarity in certain features was between the atypical canines GSI D. 307 and GSI D. 308, they included D. 307 (smaller than D. 308) along with the even smaller GSI D. 299/300 in S. sivalensis, as male and female, respectively. The taxonomic status of the larger GSI D. 308 was left unresolved, but they tentatively assigned it to a different species of Sivapithecus, S. indicus. They were clearly cognizant of the morphological differences among the larger teeth in the Siwalik upper-canine sample described by Pilgrim, but opted to include both morphologies within the genus Sivapithecus. One morphology was described as “massive with posterior shear and deep anterior vertical groove” (represented by GSI D. 196), and the other as “more slender with less accented characters” (represented by GSI D. 307 and D. 308; Gregory et al. 1938, p. 10).
Another apparent atypical hominoid canine, GSI K56/681, was reported by Dutta et al. (1976, figure 5) from Ramnagar in India, from sediments that are roughly contemporaneous with those in the upper part of the Chinji Formation. Dutta et al. actually figured two upper canines, both with broken crowns. GSI K56/681 preserves nearly the entire root with the crown base and appears to have the same morphology as GSI D. 192 and the other atypical canines. The second canine, GSI K56/680 (Dutta et al. 1976, figure 3), is the basal half of a crown that appears to have the deep mesial groove and heavily wrinkled enamel of the typical Sivapithecus canines, such as GSI D. 1 and D. 196. No comments were made concerning the morphology of either canine.
It is important at this point to note the recovery during this same time span of more of the typical Sivapithecus upper canines, identical in morphology to GSI D. 1 and D. 196. One specimen, GSI D. 238, is listed in the compendium of Siwalik primates by Wadia and Aiyengar (1938), but appears not to have been otherwise published. They report it as having been collected in the vicinity of Chinji village and it is catalogued as such at the GSI. Another specimen, GSI 18066, from Haritalyangar was described by Prasad (1964, plate 20, figure 2, but mislabeled as 18065) as having a deep anterior groove, coarse wrinkles and a well preserved lingual cingulum. Lastly, Pilbeam et al. (1977, 1980) described several mostly worn or broken canines (11 total, including two pairs of antimeres in palates), all collected from younger sediments of the Dhok Pathan Formation in the Khaur area of the Potwar Plateau. One of these specimens, however, GSP 11003, is complete and minimally worn and virtually identical to GSI D. 238.
Finally, three more of the atypical canines were described in 1983, two, GSP 17121 and 17122, by Raza et al. (1983, figure 1b) and one, BSPhG 1956 II 38, by Dehm (1983, figure 1H), all from localities in the Chinji Formation. Raza et al. (1983) provided the most complete description yet of the morphological differences between the two types of canines from the Siwaliks, with the atypical canines having less well developed mesial grooves, lower crowns, a sinuous crown-root junction, roots that taper to a triangular apex, and a crown that is more “flexed” with respect to the root compared to the typical Sivapithecus canines. With the greatly expanded sample of typical Sivapithecus canines now at hand, Raza et al. also postulated for the first time that the two morphologies segregated temporally, with the atypical canines being from Chinji sediments and the typical Sivapithecus canines being from younger sediments of the Dhok Pathan Formation, exposed mostly in the area around Khaur and at Haritalyangar. Raza et al. appear not to have been aware of Dutta’s specimen K56/680 (again, with apparently typical Sivapithecus morphology) from sediments at Ramnagar contemporaneous with the Chinji Formation, or chose to ignore it because they could not be certain of its morphology based on the published photograph. However, among the Chinji canines listed by them is GSI D. 238, which is one of the typical Sivapithecus canines that are otherwise associated with younger sediments. They did not discuss possible taxonomic implications of this purported temporal segregation of the two canine morphologies.