On average, extant primates have relatively larger brains than those of other mammals of similar body mass (Martin, 1990). Factors proposed to explain selection for relatively large brains in primates include aspects of diet (e.g., Clutton-Brock and Harvey, 1980; Aiello, 1997; Isler and van Schaik, 2009; DeCasien et al., 2017), arboreality (Falk, 2007), modifications to the visual system (e.g., Barton, 2004; Kirk, 2006), and more complex social interactions (e.g., Byrne and Whiten, 1988; Barton and Dunbar, 1997; Dunbar, 1998; Dunbar and Shultz, 2007). Each of these might have played a role in selection for relatively large brains among extant primates at various points in the evolution of this trait, and direct evidence for the timing of brain size increases relative to these factors depends on what can be inferred from the fossil record.

A number of studies focused on brain size evolution have incorporated data from Eocene fossil euprimates (crown primates, i.e., omomyoids, and adapoids; e.g., Radinsky, 1967, 1970, 1977; Jerison, 1973, 1979, 2012; Gingerich and Martin, 1981; Gurche, 1982; Martin, 1990; Kirk et al., 2014; Ramdarshan and Orliac, 2016; Harrington et al., 2016, 2020). Historically, these data have been derived from natural and latex endocasts, with a growing contribution from virtual endocasts derived from x-ray computed tomography (CT) scan data in more recent years. Reconstructions of six nearly complete virtual endocasts of middle Eocene notharctine adapoids from North America (Smilodectes gracilis and Notharctus tenebrosus) demonstrate that their neuroanatomy is similar to that of the late Eocene European adapine Adapis parisiensis (Gingerich and Martin, 1981) in certain inferred plesiomorphic features, such as small frontal lobes and cerebra that do not overlap onto the cerebella (Harrington et al., 2016). Among Omomyoidea, partial natural endocasts for Necrolemur antiquus (Hürzeler, 1948) and Tetonius homunculus (Radinsky, 1967) have been described. Virtual endocasts for the late Eocene putative omomyoid Rooneyia viejaensis (Kirk et al., 2014), and the microchoerid omomyoids Microchoerus erinaceus (Ramdarshan and Orliac, 2016) and Ne. antiquus (Harrington et al., 2020), are also available. Of these, the endocast of Rooneyia most closely resembles that of extant primates, overlapping in relative size with that of living strepsirrhines (Kirk et al., 2014). Although these adapoids and omomyoids do retain some apparently plesiomorphic traits, they also share a number of more derived euprimate features, potentially inherited from their last common ancestor. These features include relatively large brain size in at least some cases (e.g., Kirk et al., 2014), demarcation of the temporal lobe by a clearly defined sylvian sulcus (but see Gazin, 1965; Harrington et al., 2016), lack of exposure of the midbrain, a posterior location (toward the hypophyseal fossa) of the optic chiasma (Ramdarshan and Orliac, 2016; but see discussion below), and olfactory bulbs that account for less than 3.5% of the total volume of the brain (Kirk et al., 2014; Ramdarshan and Orliac, 2016; Harrington et al., 2016, 2020).

To map out the pattern and timing of when these features emerged with respect to the evolution of crown primates, it is necessary to look at fossil members of Euarchonta (i.e., the group that includes Primates, Scandentia, and Dermoptera) that retain informative plesiomorphic traits. Plesiadapiforms are a para- or polyphyletic group of euarchontan mammals consisting of 11 diverse families that existed in the fossil record of North America, Asia, and Europe during the Paleocene and Eocene epochs, ∼65–37 million years ago (Russell, 1964; Beard and Wang, 1995; Silcox and Gunnell, 2008; Silcox et al., 2017a). The Microsyopidae is the second longest-lived family of plesiadapiforms, persisting for over 20 million years (Silcox et al., 2017a). Classification of Microsyopidae has been controversial, with some suggesting a relationship to early primates (Szalay, 1969a; Bown and Gingerich, 1973; Bown and Rose, 1976; Gunnell, 1989; Bloch et al., 2007; Silcox and Gunnell, 2008; Silcox, 2008; Silcox et al., 2010a,b; Chester et al., 2015, 2017), and others inferring a closer relationship to dermopterans (Szalay and Drawhorn, 1980; Szalay et al., 1987; Ni et al., 2013, 2016). Here we follow Bloch et al. (2007), Silcox (2008), Silcox et al. (2010b), and Chester et al. (2015, 2017) in considering microsyopids to be stem primates (Fig. 1). Regardless, as there are no Paleogene cranial fossils of unambiguous treeshrews or colugos, plesiadapiforms are the best-known primitive euarchontans (possibly along with apatemyids; von Koenigswald et al., 2009; Silcox et al., 2011) from the fossil record and can be used to help establish a plesiomorphic baseline for studying primate brain form and size evolution.

Microsyopidae includes three subfamilies: the generally larger body-sized Microsyopinae, and the smaller Uintasoricinae and Navajoviinae (Gunnell, 1989). Previous endocast research for this family was limited by availability of known fossils, with endocasts reported only for the microsyopines. In 1969, Szalay described a partial latex endocast for Megadelphus lundeliusi (AMNH 55284), and produced a composite reconstruction of the brain using details of the size and morphology of the olfactory bulbs derived from a specimen of Microsyops annectens (AMNH 12595; Szalay, 1969a). These species are of notably different inferred body mass, and are now recognized as different genera (Gunnell, 1989), suggesting that combining data from them is problematic. Szalay (1969a) also provided no quantitative data from his endocranial reconstruction, and unfortunately the latex endocast that he produced has since degraded. Silcox et al. (2010a) reconstructed a virtual endocast of M. annectens from the most complete and best preserved plesiadapiform cranium known (UW 12362; Silcox et al., 2020). This provided the first quantitative endocranial data for a microsyopid, coupled with observations of a natural endocast of M. annectens (UW 14559), and a partial natural endocast for an earlier occurring species of microsyopid (cf. Microsyops elegans; UM 99843). Endocasts derived from these specimens showed some similarities to those of other plesiadapiforms (Plesiadapis cookei, Ignacius graybullianus, Plesiadapis tricuspidens; Gingerich and Gunnell, 2005; Silcox et al., 2009a; Orliac et al., 2014) in having relatively larger olfactory bulbs than those of extant euprimates, and relative brain size at the low end or below the range for fossil adapoids and omomyoids from the Eocene. The relatively rostrocaudally short cerebrum (i.e., as documented by some degree of midbrain exposure) and poorly defined temporal lobes observed in these plesiadapiform species were interpreted as evidence for less developed visual processing areas relative to those of euprimates (Silcox et al., 2009a, 2010a; Orliac et al., 2014).

Overall brain size can influence aspects of cerebral scaling, and endocranial data from the smaller microsyopids may thus provide new insights into the evolution of the brain in early primates, which are also inferred to have been of small body mass (see discussion in Silcox and López-Torres, 2017). Here we report an endocast derived from the first partial cranium of a uintasoricine, Niptomomys cf. Niptomomys doreenae (USNM 530198), from the late Paleocene Clarkforkian North American Land Mammal Age (NALMA) of Wyoming. Despite fragmentation and slight distortion of the specimen, the resulting endocranial model (Fig. 2) allows for description of size, shape, and anatomical details, supplying novel insight into the brains of small-bodied, primitive euarchontans from the late Paleocene.

The present study aims to consider the taxonomic affinity of USNM 530198, describe its endocranial anatomy, and consider the meaning of the resulting data in the broader context of early primate brain evolution. In particular, functional regions that can be identified on the endocast, and that have been identified as being of interest in early primate brain evolution (e.g., olfactory bulbs, petrosal lobules, cerebellum, neocortex), are compared to a range of comparative taxa and (when possible) quantified to allow for consideration of the effect of function on size and shape.