INTRODUCTION

The last 50 years have seen an explosive increase of small mammal fossils recorded from late Cretaceous and Cenozoic deposits, especially in North America. Two important factors contributed to this explosive increase: development and refinement of screenwashing methods of fossil collection initially advocated by Hibbard (1949), and a virtual army of eager young students to apply those collection methods. We both shared, with C. A. Repenning, numerous experiences and frustrations in the development of these fossil records, and this contribution is a small dividend from among numerous shared experiences.

Vertebrate fossils were initially reported from the Verde Formation (Fm.) of Arizona, USA, by Twenter (1962) and by Twenter and Metzger (1963). Nations et al. (1981) studied the paleoecologic-stratigraphic framework of the Verde Fm., establishing that it represents deposition in a late Cenozoic lake basin. About 1980 NJC initiated serious collecting of small mammal fossils from the Verde Fm., culminating in graduate training and the following publications (Czaplewski 1987a, 1987b, 1987c, and 1990). Czaplewski (1987b) reported six vertebrate localities from the Verde Fm.; these include MNA loc. 181 from the upper part of the Verde Fm which represents the Clarkdale local fauna (l.f.). The Clarkdale l.f. includes the rodents ?Marmota, Spermophilus, Perognathus, Geomys, Peromyscus, Reithrodontomys, Bensonomys, Sigmodon, Repomys, Neotoma (Paraneotoma), and Pliopotamys (Czaplewski 1987c); it is assigned to the middle Blancan interval of the North American Land Mammal Ages (NALMA).

Czaplewski (1987a) described sigmodont rodents from two of the Verde localities curated in the Museum of Northern Arizona (MNA 318 and MNA 319) that include the bulk of the Verde small mammals, which he characterized as the Verde l.f. MNA locs. 318 and 319 occur south of House Mountain, east of Cornville; they are separated geographically by about 1 km, and MNA loc. 318 is about 35 m stratigraphically and about 55 m in elevation above MNA loc. 319. The Verde l.f. includes the rodent genera Perognathus, Prodipodomys, Geomys, Bensonomys, Prosigmodon, Sigmodon, Jacobsomys, Reithrodontomys, Onychomys, Neotoma (Paraneotoma), the arvicoline Ogmodontomys, and another indeterminate arvicoline; it is interpreted as an early Blancan fauna. Assignment to Blancan NALMA is based primarily on presence of the arvicolines in the Verde l.f. Bressler and Butler (1978) produced a magnetic polarity sequence for part of the Verde l.f., allowing the age of the local fauna to be placed in the vicinity of 4.5 to 4.9 Ma. On the basis of the above references, along with current calibration of the Geomagnetic Polarity Time Scale (Gradstein et al. 2004), MNA loc. 318 occurs in the upper part of chron C3n.1r (~4.2 Ma), and MNA loc. 319 occurs in the top of chron C3n.3r (~4.6 Ma), both securely placed in early Blancan (NALMA). Czaplewski (1990) formally defined the Verde l.f. and characterized it with stratigraphic, radiometric, and paleomagnetic data, as well as a comprehensive faunal list.

Collections made by Czaplewski during the 1980s also produced a taxon listed as Copemys sp. (Czaplewski 1990) from three of the Verde sites (MNA loc. 701, 698, and 319). This taxon is described below as Postcopemys repenningi, new genus and new species. Most (22 of the 27 specimens) of the hypodigm for Postcopemys repenningi are from MNA loc. 701, with only three specimens from MNA loc. 698 and two specimens from MNA loc. 319. MNA loc. 701 is located in Buckboard Wash, a dry tributary of the Verde River northeast of Clarkdale; the site was discovered by Twenter (1962), designated his locality no. 6. The fossils come from a brown, highly calcareous mudstone with fragments of the gastropod Lymnaea elodes exposed near the bottom of the wash. MNA site 701 is shown in Figure 1. In addition to the taxon identified as Postcopemys repenningi, this site produced a cleithrum of a catfish (Ictaluridae), an unidentified salamander (rib and tooth-bearing jaw fragment), a partial upper molar of a small bat (?Myotis), several isolated teeth of a rabbit (Hypolagus sp.), an isolated tooth of a geomyoid rodent (cf. Prodipodomys), four isolated teeth of a cotton rat (Sigmodon holocuspis), and molar fragments of a woodrat (Neotoma (Paraneotoma) sp.). The best represented taxon from MNA loc. 701 is Postcopemys repenningi described below.

The paucity of other taxa represented in the Verde l.f. from MNA loc.701, along with the near absence of Postcopemys repenningi from MNA loc. 319 and its absence from MNA loc. 318, suggests that MNA loc. 701 is probably significantly older than MNA locs. 318 and 319 in the Verde Fm. Unfortunately, there is no precise correlation for the stratigraphic level of MNA loc. 701. For convenience (or we should say, ignorance) we continue to recognize MNA loc. 701 as part of the Verde l.f., but realize that if at some later time MNA loc. 701 can be confidently assigned to late Hemphillian LMA it must be removed from the Verde l.f. All of the specimens discussed above are curated in the collections of the Museum of Northern Arizona (MNA), in Flagstaff, Arizona.

The Maxum local fauna from University of California Museum of Paleontology (UCMP) locality V-6889, was collected from the backyard of Dr. B. J. Maxum at 291 Barrett Circle in Danville, California. Dr. Maxum reported the fossil site to paleontologists at the UCMP and the United States Geological Survey (USGS), Menlo Park. The fossils were collected in 1968 when real estate development was active in and around Danville. The fossil assemblage was initially identified by C.A. Repenning (we believe) who noted the presence of an arvicolid rodent [Mimomys (Cosomys) sawrockensis], and abundant high-crowned cricetid rodents described as Repomys maxumi, new species, by S. May (May 1981). The arvicolid rodent, along with the large size of the Repomys, convinced Repenning that the Maxum site fits within the early Blancan NALMA. The site, now covered by grass and construction, is from interbedded olive clay and gritty gray-green clay sediments now assigned to the Tassajera Formation; D. Whistler (a graduate student at the University of California, Berkeley at that time, who helped collect the site in 1968) noted that the sediments dip about 45º to the northeast at the collecting site. The geology of that area is rather complex, with several faults crossing the area. The Lawlor Tuff, dated 4.83 ± 0.02 Ma (Sarna-Wojcicki et al. 2005) is recorded from the Tassajera Fm. but the relationship between the Lawlor Tuff and the Maxum fossil site is indeterminate because exposures in the vicinity of the fossil site are now covered by widespread residential development. Repenning's initial age assignment of the site as early Blancan is unquestioned relative to subsequent knowledge and the additional fossils described below.

The important and undescribed rodents from the Maxum site were "discovered" during a visit to the UCMP by E. Lindsay and M. Woodburne in August 2006, to evaluate fossils from the Warren local fauna in the Mojave Desert. At that time, we discovered that the rodent genus Jacobsomys Czaplewski, described from the Verde fauna, is also recorded from the Maxum and Warren faunas. The Jacobsomys material from the Warren fauna is described by May et al. (2011) in this volume; we recognize it as the same species, Jacobsomys dailyi new species, that we describe from the Maxum fauna below. The Verde and Maxum faunas are considered early Blancan NALMA, whereas the Warren fauna is considered late Hemphillian NALMA. These faunas all appear close to the boundary of the Hemphillian and Blancan NALMAs.

There had been considerable effort by a former graduate student at UC Berkeley (or by C.A. Repenning) to identify the Maxum rodents prior to our "discovery." We tried to locate the person who made these identifications with the help of Dr. P. Holroyd, now in charge of the UCMP mammal collection. None of the obvious leads proved fruitful, so we assume these identifications were made by C.A. Repenning when he was still working at the USGS in Menlo Park during the 1970s and 1980s, and who provided the Repomys fossils for S. May to describe. There are several clues to the paleontologist who initially studied these fossils, including several felt-tip pen marks on the cork associated with each specimen that presumably indicate some morphological attribute of the specimen. Also, long insect pins were used as a base for gluing the specimens. Some specimens were glued parallel to the shaft of the pin, and others were glued to the tip of the pin whose head had been cut off. Virtually all of the pins to which specimens were glued are very long, which makes it difficult to avoid bumping the vial and knocking the glued specimen from the pin when removing the protective vial from the cork into which the pin is embedded. We know that some of these curatorial practices were used by C.A. Repenning, so he is our prime candidate for the researcher who initially studied and identified these fossils. Tooth cusp terminology for specimens described below is from Lindsay (1972).