A. VARIOUS DJADOCHTATHERIDS

Taxon: within Djadochtatherioidea, (formerly Djadochtatheria Kielan-Jaworowska & Hurum, 1997) "Kryptobaatar, Djadochtatherium, Catopsbaatar, and Tombaatar form a clade, for which the family Djadochtatheriidae is proposed. Chulsanbaatar is the sister taxon of this clade. Bulganbaatar and Nemegtbaatar are the sister group of all other djadochtatherians.," (Kielan-Jaworowska & Hurum, 1997). Genera: Bulganbaatar, Chulsanbaatar, Nemegtbaatar, other reports Time-Line: Upper Cretaceous: Bulganbaatar, Chulsanbaatar, Nemegtbaatar

Genus: Bulganbaatar Kielan-Jaworowska Z, 1974

Species:	Bulganbaatar nemegtbaataroides Kielan-Jaworowska Z, 1974
Place:	Bayan Zag, Djadokhta Formation
Country:	Mongolia & Kazakhstan?
Age:	Campanian, Upper Cretaceous
Remarks:	Kielan-Jaworowska et al, 2000 have ridden to my assistance with a convenient summary on page 593. My thanks to the supplier. Unlike the members of the more intermite family of Djadochtatheriidae, the wall of the snout for this genus curves in strongly in front of the cheek zygomatic arch), and shows similarity in this respect with Nemegtbaatar. However, it's a smaller critter and its upper teeth aren't as cuspy as those from its friend. Cusp formulae buccal to lingual) Uppers: P4 2:5; M1 5:5:3; M2 1:2:2. Holotype ZPAL MgM-I/25 is employed by the Institute of Paleobiology, Warsaw. A further specimen offered a complete skull, jaws, much of its shoulders and most of the front legs. Additonal notes I’ve seen indications that there might be a second, unnamed species. Sadly, I've forgotten where!
Reference:	Kielan-Jaworowska (1974), Multituberculate succession in the Late Cretaceous of the Gobi Desert (Mongolia). Palaeontologica Polonica, 30, p.23-44.

Link: Acta Palaeontologica Polonica 42(2), 1997 http://www.paleo.pan.pl/acta/acta42-2.htm#AVERIAN Averianov AO (1997), New Late Cretaceous mammals from southern Kazakhstan, APP 42(2), p. 243-256. The abstract.

Genus: Chulsanbaatar Kielan-Jaworowska Z, 1974

Species:	Chulsanbaatar vulgaris Kielan-Jaworowska Z, 1974
Place:	Ukhaa Tolgod, Khulsan, Nemegt, Hermiin Tsav II
Country:	Mongolia
Age:	Campanian (late?), Upper Cretaceous
Remarks:	A multi midget with a 2cm skull. The jaw would fit on your fingertip, as the linked photo shows. Remarkably, it’s been possible to study the ear bones, which shows how well some of the fossils are preserved. Chulsanbaatar is now a resident of modern-day Warsaw, (eg. ZPAL MgM-1/84). There are a fair number of specimens in the collection, which doubtless accounts for the species name. Vulgaris = common. In terms of anatomy, this animal is more or less completely known. Even endocasts of the brain have been preserved. Kielan-Jaworowska et al, 2000... ... contains a summary on page 594. This is the most common djadoch multi, thus the specific name. That has nothing to do with any naughty Vulgarian habits. Chulsan has a skull length ranging from 1.8 to 2.2 centimetres. Its snout is roughly rectangular in outline, and curves inwards before arriving at the cheek area. Cusp formulae (buccal to lingual) Uppers: P4 2:6; M1 4:5:ridge; M2 1:2:2. Lowers: p4 7 ridges; m1 4:3; m2 2:2. Holotype ZPAL MgM-I/139 is a skull being held hostage by the Institute of Paleobiology, Warsaw.
Reference:	Kielan-Jaworowska (1974), Multituberculate succession in the Late Cretaceous of the Gobi Desert (Mongolia). Palaeontologica Polonica, 30, p.23-44.

Links: Hurum JH http://www.2dgf.dk/online/hurum.htm Reconstruction of the petrosal in Late Cretaceous multituberculates (Mammalia). A technical, yet interesting paper. The petrosal is the bony casing of the inner ear. "The long cochlear canal and well developed semicircular canals observed in the sectioned skulls of Nemegtbaatar and Chulsanbaatar, suggests close affinity of multituberculates to other mammals, and the thick ossified pila antotica suggests affinity to monotremes." That’s one way of looking at it. The affinities of multis are very much a continuing field of discussion. A bit more on these ears follows the next couple of links. The University of Oslo, Paleontological Museum http://www.nhm.uio.no/palmus/galleri/montre/english/x562.htm A cast specimen in the Oslo display cabinet. The Prehistoric Data Files http://www.angellis.net/Web/DFG-mam/Chulsanbaatar.htm Time for some more artwork, courtesy of VRW.

A bit on multi ears: A 1998 paper by Hurum JH, (see Bibliography at the end of this page), provides more detail on the earworks of Chulsanbaatarand Nemegtbaatar. The fact that these small features are known so thoroughly, is a good illustration of the remarkable state of preservation of some of these Mongolian fossils. The inner ear of multis is more properly 'mammalian' than is the case for earlier critters such as Morganucodon and Sinoconodon. As for those animals, this useful auditory equipment is housed within one bit of bony casing called the petrosal. Reptiles and non-mammalian therapsids required a number of bones for the packaging. Unlike those couple of aforementioned basal mammals, however, the multi ears are yet more like those of us proper mammals. Each contained three small bones for hearing, (the stapes, incus and malleus). With M. and S. the equivalents of the latter two were externally located on the lower jaw, though they seem to have played important roles in the sense of hearing. A slice of the action Hurum's method involved slicing up sections of two skulls from the collection of the Institute of Paleobiology in Warsaw. (He did ask first.) The Chulsanbaatar skull, all 21mm of it, was cut into a sequence of 885 divisions. The mightier Nemegtbaatar head, (a full 45mm in length), was carved into 1370 sections. These tiny bits were then examined through a microscope, and further information was obtained with reference to other skulls in the collection and a convenient wax model of the brain and inner ear of Nemegtbaatar. The matter-of-fact description of this painstaking process tells of dedication and patience, (p.66-67). (Translated from diplomatic language, that means much of the paper presently evades my understanding.) It might be of interest to know that the cochlear canal of the second mentioned genus has a volume of four cubic millimetres. This was determined by cutting up an extra model, and measuring the volume by means of water displacement. Dissection Hurum then proceeds to dissect the exposed audible hardware of Nemegtbaatar gobiensis, (p.67-74), and Chulsanbaatar vulgaris (p.74-79). Both come complete with photos, diagrams and measurements. He was even able to offer insights on the wiring; the position of veins and nerves. If anyone would appreciate such details, then get hold of the paper. Amongst many more bits and pieces, the mammalian ear contains a cochlear canal. Sound travels along it, so it can be processed. Non-mammals tend to have more of a chamber or cavern instead, which means they have less time to glean information, and a less keen sense of hearing. In multis, (p.83), the slight lateral curvature is similar to what's known from the duckbilled platypus, ( Ornithorhynchus), but this is also a feature shared with other mammals of the Mesozoic.

Genus: Nemegtbaatar Kielan-Jaworowska Z, 1974 'Nemegt hero'

Species:	Nemegtbaatar gobiensis Kielan-Jaworowska Z, 1974
Place:	Khulsan, ?Nemegt, Hermiin Tsav II
Country:	Mongolia
Age:	Campanian, Upper Cretaceous
Remarks:	"Compared to all extant mammals the braincase in Nemegtbaatar and Chulsanbaatar is primitive…" (Hurum, 1998). I cite this quotation for two reasons. Firstly, even I can understand what it means. Secondly, all extant mammals includes the monotremes such as the duck-billed platypus, despite its residual egg-laying habit. Nemegtbaatar was a fairly large member of the clan, with a skull length of up to 4,5cm. At least one specimen is in the Institute of Paleobiology collection of the Polish Academy of Science at Warsaw, (ZPAL MgM-1/76). See the link attached to Chulsanbaatar for insights concerning the inner ear, above. Affinities Kielan-Jaworowska et al, 2000 provides a summary on page 593. This is a middle of the range sized djadoch multi, and its nearest known relative is Bulganbaatar. Both have snouts that curve inwards before the cheek area, and this is in contrast to the straight snouted djadochtatheriids. However, as well as being larger than Bulgan, this genus also has a larger complement of cusps on the upper postcanines. Cusp formulae (buccal to lingual) Uppers: P4 3:6:1; M1 6:7:4; M2 1:3:2. Lowers: p4 7 ridges; m1 5:4; m2 3:2. Holotype The type fossil is ZPAL MgM-I/81, and most of its body may be enjoyed in the collection of the Institute of Paleobiology, Warsaw.
Reference:	Kielan-Jaworowska (1974), Migrations of the Multituberculata and the Late Cretaceous connections between Asia and North America, Annals of the South African Museum, 64, p.231-243.

Links: Acta Palaeontologica Polonica 43(1), 1998 http://www.paleo.pan.pl/acta/acta43-1.htm#Hurum Hurum (1998), The braincase of two Late Cretaceous Asian multituberculates studied by serial sections. APP 43(1), p 21-52. Another complex abstract. T. rex of het einde van de dinosauriërs http://www.cdbeta.uu.nl/model/t_rex.shtml The University of Utrecht invites you to consider the extinction of the dinosaurs, (in Dutch). They suggest N. weighed 500g, though I’ve no idea if that’s accurate. It’s in the Overlevingsmodellen. You can also zie afbeelding, (see a picture). The University of Oslo, Paleontological Museum http://www.nhm.uio.no/palmus/galleri/montre/english/x570.htm A further photo of a cast specimen from Mongolia, via Norway. This could be strong evidence that Dr Kielan-Jaworowska spent a number of years working in Oslo.

Other reports: Xxxxxxxxxxxxxxxxxxxx Xxxxxxxxxxxxxxxxxx

A. Various djadochtatherioideanss B. Sloanbaataridae C. Djadochtatheriidae

B. SLOANBAATARIDAE

Taxon: Sloanbaataridae Kielan-Jaworowska Z, 1974 Reference: Kielan-Jaworowska (1974), Multituberculate succession in the Late Cretaceous of the Gobi Desert (Mongolia). Palaeontologica Polonica, 30, p.23-44. "Kamptobaatar, Sloanbaatar, and Nessovbaatar form a separate clade in the Pee Wee tree," (Kielan-Jaworowska & Hurum, 1997). Genera: Kamptobaatar, Nessovbaatar, Sloanbaatar, other reports Time-Line: Upper Cretaceous: Kamptobaatar, Nessovbaatar, Sloanbaatar

Genus: Kamptobaatar Kielan-Jaworowska Z, 1970 'bent hero' Remarks: 'Bent' here refers to the observable bend of the zygomatic arches in the skull.

Link: DML, Ben Creisler, via Mickey Rowe http://www.cmnh.org/fun/dinosaur-archive/1999Oct/msg00339.html

Species:	Kamptobaatar kuczynskii Kielan-Jaworowska Z, 1970
Place:	Bayan Zag, Djadokhta Formation, ?Ukhaa Tolgod
Country:	Mongolia
Age:	Campanian, Upper Cretaceous
Remarks:	Kielan-Jaworowska et al, 2000 provides some information on pages 594-595. The snout is vaguely rectangular at the front, and there's an inwards curve before reaching the zygomatic arch. Its lower jaw's short with a low coronoid process. Cusp formulae (buccal to lingual) Uppers: P4 3:5-6; M1 5:5:ridge; M2 1:2:3. Lowers: p4 7 serrations; m1 4:3. The second lower molar hasn't been identified. Holotype ZPAL MgM-I/33 is a juvenile skull attending the Institute of Paleobiology, Warsaw.
Reference:	Kielan-Jaworowska (1970), New Upper Cretaceous multituberculate genera from Bayn Dzak, Gobi Desert. Palaeontologia Polonica, 21, p.35-49.

Kamptobaartar kuczynskii by Daniel Bensen This image is the work and property of Mr Daniel Bensen, who kindly allowed its use here. His homepage has plenty more paleo-art to enjoy, and his commentaries are carefully constructed and informative. A visit is recommended. Opus Dinosaur by Daniel Bensen http://maier.gotnet.net/

Genus: Nessovbaatar Kielan-Jaworowska Z & Hurum JH, 1997 'Nessov’s hero' Aka: Nesovbaatar

Species:	Nessovbaatar multicostatus Kielan-Jaworowska Z & Hurum JH, 1997
Aka:	Nesovbaatar multicostatus
Place:	Hermiin Tsav II
Country:	Mongolia
Age:	Campanian, Upper Cretaceous
Remarks:	Kielan-Jaworowska et al, 2000 contains a summary on page 595. For some reason, they consistently use the spelling of Nesovbaatar The dentary is 1.35cm in length and the complete skull would've been longer. Its coronoid process is comparatively small, and this character is similar to Sloanbaatar and Kamptobaatar. The condyle, the bit of bone responsible for forming the lower part of the jaw-skull joint, is set higher than the tops of the molars in this genus, as with a number of other djadochs. Lower teeth The p4 is a larger premolar than other djadoch models. It has nine serrations, eight of which carry weak ridges. The first molar has 4 buccal cusps and 3 lingual ones as with Sloanbaatar, but it differs when it comes to m2; 3:2 rather than 2:2. Other details differ as well. These molars are similar to those of Chulsanbaatar, but the premolar is more like the corresponding tooth of Arginbaatar, a Lower Cretaceous non-djadoch. Among the distinctions from that genus though is the lower number of serrations. Holotype ZPAL MgM-I/103 consists of the lower jaws of a juvenile in the kindergarten of the Institute of Paleobiology, Warsaw. Additional notes The abstract of the reference is linked somewhere towards the top of this directory.
Reference:	Kielan-Jaworowska Z & Hurum JH (1997), Djadochtatheria: a new suborder of multituberculate mammals. Acta Palaeontologica Polonica 42(2), p 201-242.

The following is based upon my reading of Kielan-Jaworowska et al, 2003. Campanian Mammals of the Gobi The Gobi houses an impressive number of localities yielding mammalian fossils from the Campanian stage of the Upper Cretaceous, (ca. 84 - 72 million years ago). There comparative ages have been much discussed, (p.273). The matter is complicated by a complete lack of any radiometric datings or strata containing marine rock, which could provide convenient index taxa known from further afield. As the terrestrial vertebrates only tend to occur in the Gobi, the strongly endemic flavour makes comparisons with elsewhere much less than straightforward. The earliest phase of the Campanian deposition appears to be represented by the Djadokhta Formation. Next in line is the Baruungoyot and youngest is the Nemegt Formation. In each case, there are other formations of similar ages. To make things more entertaining, some locations have been referred to by a variety of names at one time or another, (and indeed simultaneously), often because researchers from different countries and cultures have been active. There's now a concerted effort to stabilise the terminology. A welcome complication has been introduced by the discovery of the exceptionally generous Ukhaa Tolgod beds ('Brown Hills'). Faunal comparisons suggest this dates from between the Djadokhta and Baruungoyot Formations, but more precision isn't yet possible. This locality came into production in 1993, and is numerically the richest Mesozoic mammal mine in the world. In terms of quality it's merely exceptionally excellent. A further interesting area is Bayan Mandahu in Inner Mongolia, (p.274). This has extended the borders of the Gobi Campanian company into China. Research so far suggests it corresponds chronologically with the Djadokhta. Mammal diversity The paper provides a listing of mammals identified so far in the locations, and this acts as a useful overview of similarities and contrasts. The inventory is of course interim, (p.275). More taxa will be described and the living diversity was presumably greater than will ever be known. Twelve multituberculate species represent ten genera. There are also a dozen therians, one of which awaits description; six metatherians (including deltatheriids), five eutherians and one of unclear affinities. Comparisons At least thirteen mammal species occur in the Djadokhta group of locations and nine in the Baruungoyot. Only two are shared, (p.276). This is consistent with differing ages. They also both appear in the Ukhaa Tolgod beds, which have further species in common with both the older and younger localities. Overall, there's more similarity with the Djadokhta and this may be as a result of closer ages. The caveat is that these comparisons only involve mammals. Whether the patterns are reflected by other terrestrial vertebrates hasn't been addressed, (p.277). Ancient environments The geology indicates that rock was deposited under different conditions, which is reasonable enough. The Djadokhta and Baruungoyot Formations are sandstones formed largely by windswept dunes. In contrast, the beds of Ukhaa Tolgod appear to have involved significant flows of debris transported in following heavy rainfall in a generally arid area. Further Mesozoic site summaries can be found at Localities. Meet the mammals of the Campanian Gobi Table 2 (p.277) depicts mammalian distribution for the Upper Cretaceous Gobi Formaitons. The first group of localities listed here are the oldest, and the last are the youngest. All locations are in Mongolia unless otherwise stated. The letters following the species refer to the faunas in which they occur. i. Djadokhta Formation (and equivalents) (11 genera, 14 species) Locations: Bayan Zag (BZ), Törgrög (T), Bayan Mandahu (BM) - PR China. Multis (6 genera, 7 species) - Sloanbaatar mirabilis BZ; Bulganbaatar nemegtbaataroides BZ; Kamptobaatar kuczynskii BZ; Djadochtatherium matthewi BZ, T, BM; Kryptobaatar mandahuensis BM; K. dashzevagi BZ, T; ?Tombaatar sabuli BM. Therian (1 genus, 1 species) - Hyotheridium dobsoni BZ. Eutherians (2 genera, 4 species) - Kennalestes sp. BM; K. gobiensis BZ, T; Zalambdalestes sp. BM; Z. lechei BZ, T. Metatherians (including deltatheroidans 2 genera, 2 species) - Deltatheroides cretacicus BZ; Deltatheridium pretrituberculare BZ, ?BM. ii. Ukhaa Tolgod beds (9 genera, 9 species) Locations: Ukhaa Tolgod. Multis (6 genera, 6 species) - ?Sloanbaatar mirabilis; ?Kamptobaatar kuczynskii; Djadochtatherium matthewi; Kryptobaatar dashzevagi; Chulsanbaatar vulgaris; Tombaatar sabuli. Eutherians (2 genera, 2 species) - Zalambdalestes lechei; Ukhaatherium nessovi. Metatherians (including deltatheroidans 1 genus, 1 species) - Deltatheridium pretrituberculare. iii. Baruungoyot Formation (and equivalents) (9 genera, 9 species) Locations: Khulsan (K), Nemegt (N), Hermiin Tsav I (HT1), Hermiin Tsav II (HT2), Üüden Sair (US). Notes: Baruungoyot aka Barun Goyot; Hermiin Tsav aka Khermeen Tsav. Multis (5 genera, 5 species) - Catopsbaatar catopsaloides K, HT1; Nessovbaatar multicostatus HT2; Nemegtbaatar gobiensis K, ?N, HT2; Chulsanbaatar vulgaris K, N, HT2; Kryptobaatar dashzevetgi HT1 Eutherians (2 genera, 2 species) - Asioryctes nemegetensis K, N, HT2; Barunlestes butleri K, N, HT2. Metatherians (including deltatheroidans 2 genera, 2 species) - Deltatheridium pretrituberculare N, HT2; Asiatherium reshetovi US. iv. Nemegt Formation (and equivalents) (2 genera, 2 species) Locations: Guriliin Tsav (GT), Khaichin Uul-I (KU). Notes: Khaichin Uul-I aka Bügiin Tsav. Multis (1 genus, 1 species) - Buginbaatar transaltaiensis KU. Metatherians (including deltatheroidans 1 genera, 1 species) - 'Guriliin Tsav skull' GT (unpublished, but could be a deltatheroidan or a stagodontid).

Genus: Sloanbaatar Kielan-Jaworowska Z, 1970 Remarks: The genus is named in honour of paleontologist RE Sloan.

Species:	Sloanbaatar mirabillis Kielan-Jaworowska Z, 1970
Place:	Bayan Zag, Djadokhta Formation, ?Ukhaa Tolgod
Country:	Mongolia
Age:	Campanian, Upper Cretaceous
Remarks:	Kielan-Jaworowska et al, 2000 contains a brief introduction on pages 589-590. This genus is represented by a small djadoch skull of 2.5cm in length. The snout's described as narrow and rectangular, with the zygomatic arch of the cheek being wide. The cusp formlae for upper postcanines are as follows (always buccal to lingual): P4 2:5; M1 4:4:ridge* (* as cusp counts are stated for this row, perhaps it's cusped ridge); M2 1:2:3. Lower p4 premolars have eight serrattions: m1 4:3; m2 2:2. Holotype ZPAL MgM-I/20 is a skull, lower jaws and skeleton parts residing at the Institute of paleobiology, Warsaw. It's a rather good specimen but lonely. Certainly in 2000, it was the only one known.
Reference:	Kielan-Jaworowska (1970), New Upper Cretaceous multituberculate genera from Bayn Dzak, Gobi Desert. Palaeontologia Polonica, 21, p.35-49.

Link: Zofia K-J, first recipient of the Walter Granger Memorial Award, 1998 http://users.rcn.com/granger.nh.ultranet/1strecip.html An article on a present day derived eucynodont, which gives the reasons for the award. Best read with sunglasses.

Other reports: Xxxxxxxxxxxxxxxxxxxx Xxxxxxxxxxxxxx

A. Various djadochtatherioideans B. Sloanbaataridae C. Djadochtatheriidae

C. DJADOCHTATHERIIDAE

Taxon: Djadochtatheriidae Kielan-Jaworowska Z & Hurum JH, 1997 Reference: Kielan-Jaworowska Z & Hurum JH (1997), Djadochtatheria: a new suborder of multituberculate mammals. Acta Palaeontologica Polonica 42(2), p 201-242. Multituberculates have traditionally been depicted as herbivores. This is too simplistic, and they’re increasingly referred to as herbivores and omnivores. Certainly in the case of Kryptobaatar, if I’d been a small Mesozoic lizard, I’d have kept a healthy distance between us, just in case. Kielan-Jaworowska et al, 2000 provide a brief outline of the family on page 590. The family snout is unique for known multis: "...in having a subtrapezoidal snout in dorsal view, with a wider anterior margin and lateral margins confluent with the zygomatic arches rather than incurved in front of the arches." That means the sides of the snout run straight back to the cheeks in a diagonal line whereas, for others, there's some inward curvature in front of the cheeks. An additional snoutism also distinguishes family members from other djadochs. It's comparatively longer, in that it contributes at least half of the whole skull length. Genera: Catopsalis (partly = Catopsbaatar & Djadochtatherium), Catopsbaatar, Djadochtatherium (partly = Catopsbaatar), Gobibaatar (= Kryptobaatar), Kryptobaatar, Tombaatar, Tugrirgbaatar (= Kryptobaatar), other reports Time-Line: Upper Cretaceous: Catopsbaatar, Djadochtatherium, Kryptobaatar, Tombaatar

Genus: Catopsbaatar (Kielan-Jaworowska Z, 1974) Kielan-Jaworowska Z, 1994 'Evident hero' Aka: Catopsalis (partly); Djadochtatherium (partly) Remarks: For those of a technical inclination: "One of the most characteristic features of Catopsbaatar (which differentiates it not only from Kryptobaatar but from all the djadochtatherioids in which the zygomatic ridges are known ), is a very deep anterior zygomatic ridge, and a small medial zygomatic ridge, the latter forming about a quarter of a circle and adhering the anterior one from behind," (Kielan-Jaworowska et al, 2002). Name The generic name (Kielan-Jaworowska 1994, p.134) refers to the apparent similarity of this genus to Catopsalis of North America, and that's added to by an oft used Mongolian word for 'hero'.

Species:	Catopsbaatar catopsaloides (Kielan-Jaworowska Z, 1974) Kielan-Jaworowska Z, 1994
Aka:	Catopsalis catopsaloides, Djadochtatherium catopsaloides Kielan-Jaworowska, 1974
Place:	Hermiin Tsav I (aka Khermeen Tsav), Khulsan
Country:	Mongolia
Age:	Campanian, Upper Cretaceous
Remarks:	A short intro with a couple of odd words. The following is based upon my reading of Kielan-Jaworowska et al, 2005. A central concern of that study is a discussion on zygomatic ridges in multituberculates. The cheek bone has ridges which provide information on the arrangement of muscles. Among other things, the details suggest that the masseter superficialis consisted of two parts, and this is a multi autapomorphy dating back at least as far as the Upper Jurassic paulchoffatiids. Anybody concerned with that particular aspect ought to consult the paper directly, which is linked below. Welcome to the family Djadochtatheriidae is a frightening sounding family of mainly Mongolian multis, and contains four genera. They were relatively large representatives at the time, but that only equates to rat-sized. A characteristic presently restricted to these djadochs is an oddly shaped snout (p.487). Generally, mammalian snouts curve inwards in front of the cheeks. With djadochs though, the side of the jaw forms a more or less straight, diagonal line continuing from the zygomatic arch. The authors term the resultant shape subtrapezoid. This makes them wide. Remains, range and some paleo-history Catopsbaatar fossils now reside in Poland, Russia and Mongolia and were mostly collected from Hermiin Tsav in Mongolia. A lower molar (m2) is the only exception. This was found in the fauna from Khulsan. The first specimens were three skulls bagged by the Polish-Mongolian Expedition in 1971, which located two fossil yielding outcrops now named Hermiin Tsav I and II. Two skulls (including the holotype) came from the first and one from the other locality. A Soviet-Mongolian expedition picked up another pair from somewhere in the area in 1975 and a sixth was added in the 1990s. The age seems to correspond with that of the Baruungoyot Formation; either lower or upper Campanian (p.488). The first description appeared in 1974 when the specimen was referred to the related genus of Djadochtatherium. Five years later, a revision shifted the species to the mainly North American taxon of Catopsalis. (Note: page 508 states some Catopsalis material is known from the Gobi Paleocene.) However, that genus was shown to be something of a hotchpotch containing at least five independent lineages. This realisation led to the establishment of Catopsbaatar. That may sound like a bit of a tangle, but paleontologists are generally conservative when it comes to establishing new taxa. For one thing, an over-abundance of names can add confusion to the relevant literature. If fossils can be fitted into already existing diagnoses, then that is what generally happens. A further factor is the possible incompleteness of available specimens. In the majority of cases most the animal is missing. Such realities make some temporary hotchpotches such as the former concept of Catopsalis unavoidable. Family (p.489) Kielan-Jaworowska et al, 2005 provides a revised diagnosis for djadochtatheriids. The shape of the snout differs to all other known mammals, as mentioned above. In contrast to other djadochtatherioideans, the snout is also long, accounting for about half the length of the skull or more. Several details of skull architecture are also mentioned including a prominent nuchal crest at the back. This curves inwards from both sides, and that means the skull is shorter in the middle than at its flanks, when looked at from above. Generic distinctions Despite once having been referred to Djadochtatherium, Catopsbaatar is clearly distinct from all other family members. The eyes are set further back as the snout accounts for around 65% of the length of the skull. It's also the largest known genus in the family, although the difference isn't much in comparison to D.. A further distinction concerns the lower premolar (p4). This tooth is comparatively smaller, subtrapezoidal in shape (rather than a crescent), and lacks the vertical ridges on the sides, which were generally popular with multis. The upper premolars are reduced to an original crew of three instead of four, as is also the case for Tombaatar. It's the P2 that went out of fashion. As with Djadochtatherium, the front most ridge of the zygomatic arch is unusually high but its shape differs in both genera. All the upper incisors of C. are positioned in the premaxilla bone, whereas the alveolus for I3 is partly in the maxilla for Tom. Significant features are also to be found in the construction of the skull. Population census Page 490 provides information on biological ages of the six available skulls. At least two are adult and three juvenile. One of the adults is preserved along with much of the skeleton, which has yet to be described. Approximate skull lengths range from 4.8 to 7cm. Going straight to the head As the authors go into far more detail than I can presently follow, I'm going to content myself with some relatively accessible points. One thing I have done is to quickly and crudely adapt a bird's eye view of a composite skull from page 496. It's not meant to be a good sketch, but it hopefully gives some idea of which bones go where. I haven't attempted paying all too much regard to perspective or precision. Jeer as much as you like at this resultant masterpiece. The skull is wide. In the juvenile holotype, the maximum width (5.6cm) exceeds the length (5.3cm). This isn't so for the two other specimens good enough to provide such measurements; (6.3 long against 5.5 wide and about 5.2 against 4.5). Nevertheless, they're also broad. This is partly a consequence of the expanded zygomatic arch of the cheek. The central part of the long, wide snout is unsurprisingly formed by the nasals. Along the front two-thirds is found a series of small, naturally occurring holes; foramina. This is normal enough as they're the equivalent of holes in the walls of a house for wiring and piping. An unusual aspect is the variation in numbers. Sometimes there's a pair per side, as would be expected for multis. However, one specimen has four foramina on the right bone with only three observable on the left (p.495).This arrangement was very unexpected. The authors describe the various elements of the skull, but I'm going to jump to page 502 for the jaws. Eating utensils There's a lot of information available for the lower jaw, as both complete and fragmentary specimens have been donated by adults and juveniles. To be brief, the mandible is robustly built and comparatively long. Lengths of four were given back on page 490. These range from 3.5 to 4.1cm. Teeth An important function of jaws in many animals is to provide somewhere to keep the teeth. Page 504 contains the complete dental formula for Catopsbaatar: (uppers): two incisors, no canine, three premolars and two molars; (lowers): one, none, two and two respectively. Somewhat perverse as it may sound to some, the two upper incisors are I2 and I3. The ancestors' first incisor was dispensed with during earlier stages of the development of the lineage. The I2 is a strongly built tooth with two roots, and has enamel restricted to a band at the front. Left and right incisors have a slight contact with each other in one specimen, as they gently meet at the midpoint between them. In all cases the I3s are either absent or poorly preserved, and not much more can presently be reliably reported about these. Upper premolars The complete set consists of three; P1, 3 and 4. Number 2 disappeared earlier along the line. However, only juveniles actually had the trio. Positions 1 and 3 were lost as the animals grew up and, in older individuals, all trace of the erstwhile alveoli vanished. Upper molars The M1 is the cuspier of the two teeth, and has a cusp formula of 5-6: 5-6: 4 (p.505). Not terribly surprisingly, the cusps are sharper and less worn in younger animals. They hadn't done as much work. Heavy labour in the mouths of adults turned cusps into concavities. A similar effect occurs on the M2s; cusp formula 2: 2-3: 2-3. Measurements Table 2 (back on page 504) contains a variety of measurements for postcanine teeth and complete tooth rows, as gleaned from seven individuals. Not all dimensions are provided by each animal. The following are all lengths but molar widths are available as well. P4 2.1-4.2mm; M1 6.6-7.5mm; M2 4.2-5mm; upper tooth row ca. 19.3 - ca. 23mm. Lower incisors As is the multi wont, only a single incisor is found on each side. It's a large, diagonally forwards pointing tooth with enamel restricted to the front, (p.506). It continued to grow for a lifetime. In the right specimen of the holotype, there's a large ovoid area of wear on the back. The canines are also typical for multis, as there aren't any. Some more basal representatives did retain such teeth. Lower premolars While there are two per side, (p3 and p4), the first is reduced to a runt. It's pathetic and embarrassedly hides beneath the front of its mightier colleague. The p3 makes do with a single root, which some might say is one more than it deserves. The length isn't worth measuring. On the other hand, p4s range from 3.2 to 3.6mm. This is a more than respectable size although proportionately smaller compared with many multis. When looked at from the side, the top of the crown contains a three-cusped blade for shearing foodstuff prior to milling by the molars. Lower molars The m1 length varies from 5.4 to 5.8mm and the teeth have two rows of democratically distributed cusps, with four in each. These can differ in size relationships even in the same mouth. In one individual the external cusps are larger than the internal ones on the right molars, but the situation is reversed on its left counterpart. The final cusp in a row is the largest, and size decreases along the line towards the front. The m2s are smaller than m1s; 3.2 - 3.5mm. There is generally a pair of rows of two cusps, but a third cusp is present on the external row of one right tooth. The cusps of the internal rows are wider. Lengths of the entire lower tooth batteries vary from 11.3 to 12.7cm (known from four specimens). All animals are individuals The variation in cusp numbers and dimensions isn't unusual for multis, and the specimen skulls differ in size. Juveniles are predictably smaller than adults. Variations between individuals of a more technical nature include the number of foramina (small holes) in the infraorbital and nasal regions of the skull. The latter has been found to differ between the two halves of the same snout. Relatives Lots of characteristics unite Catopsbaatar with other members of the family Djado, (Djados sound much more agreeable than djadoctatheriids), and its closest known relative is Djadochtatherium itself. In both genera, the front of the cheek bone (zygomatic arch) is both large and high, although it is differently shaped (semicircular for C. against near rectangular for D.). This is of course not the only difference (p.507). The most obvious distinction between a Catopsbaatar skull and those of the rest of the family is in the eye of the beheld. The orbit in this instance is relatively small and further back. This comparison holds for all well-known djadochtatherioideans. One consequence is a longer snout; 65% of skull length for Catopsbaatar; 58% for Djadochtatherium; 49-50% Kryptobaatar; 48% Nemegtbaatar; 46% Kamptobaatar; 41% Sloanbaatar and Chulsanbaatar. However, as the snout is also wide, the fact of its length isn't necessarily obvious at a quick glance. (You're welcome to have another look at my resultant masterpiece if you like.) The position of the orbit also has implications for the proportions and meeting places of neighbouring skull bones. As I've just learnt from attempting to tidy our cellar, moving one piece of furniture can require the shifting of many. Dental count divergence and p4s Djadochtatherium (and most djadochtatherioideans) had a relatively traditional number of teeth; per side (uppers): 2-0-4-2; (lowers): 2-0-2-2 (p.508). The known exceptions are Catopsbaatar and Tombaatar, both of which have (as a maximum) three upper premolars rather than four. P2 had been dispensed with. As stated, mature C. individuals also gave up P1 and P3. It's not known whether Tom did likewise, as the sole specimen is uninformative on the matter. In all djadochtatherioideans for which it's known (with one exception), the main lower premolar (p4) has a crescent shape when seen from the side and a number of ridges. This also applies for the relatively small tooth in Djadochtatherium. Catopsbaatar is the exception, as the outline is more trapezoidal and the tooth comparatively even smaller (for a multi). There are three cusps along the top but no ridges on the sides. Losing premolars The loss of premolars in line with the advancement of age is a characteristic in common with taeniolabadids, to which the genus was previously referred. Taeniolabidids, however, took things further, as they ended up with just one premolar per side both up and down. There's a chance that Gobi Paleocene genera (assigned to the taenios) might perhaps have connections with the djados, which are known from but a snapshot of their history. However, such an assumption would be shallowly rooted in crumbly soil. Those animals (Catopsalis and Prionessus) aren't presently known from skull material. (You can't make comparisons with what isn't available.) The p4 of Prionessus is relatively small and (as far as can be told) has more in common with taenio teeth. Upper molars though have more similarities with djadochtatherioideans. Holotype ZPAL MgM-I/78 is a near complete skull with both lower jaws in the collection of the Institute of Paleobiology, Warsaw, (p.490). It requires foster parents as this individual is a juvenile. Size and cusp formulae (buccal - lingual) According to Kielan-Jaworowska et al, 2000 (p.592), this was a large djadoch multi with a skull length of six cemtimetres. The same source provides cusp formulae for the postcanines. Uppers: P4 5:1; M1 5:5:4; M2 2:3:3. Lowers: m1 4:4; m2 2:2. They also congratulate the animal on its exceptionally long snout.
References:	Kielan-Jaworowska (1974), Multituberculate succession in the Late Cretaceous of the Gobi Desert (Mongolia). in Results of the Polish-Mongolian Palaeont. Expeditions - Part V. Palaeontologica Polonica. (30), p.23-43.
	Kielan-Jaworowska & Sloan (1979), Catopsalis (Multituberculata) from Asia and North America and the problem of taeniolabidid dispersal in the Late Cretaceous. Acta Palaeontologica Polonica 24, p.187-197.
	Kielan-Jaworowska (1994), A new generic name for the multituberculate mammal "Djadochtatherium" catopsaloides. Acta Palaeontologica Polonica 39, p.134-136.

Links: Acta Palaeontologica Polonica, 50(3), p.487-512 http://app.pan.pl/acta50/app50-487.pdf Skull structure in Catopsbaatar and the zygomatic ridges in multituberculate mammals, Kielan-Jaworowska, Hurum & Lopatin, 2005. The full paper is freely available on-line. The University of Oslo, Paleontological Museum http://www.nhm.uio.no/palmus/galleri/montre/english/x564.htm There are some good photos around of these creatures. This is a cast of Cat. cat. from the Gobi. Institute of Paleobiology, Polish Academy of Science, Warsaw http://www.paleo.pan.pl/collect.htm#Mon-mammalia Included in this collection are specimens of Bulganbaatar nemegtbaataroides, Catopsbaatar catopsaloides, Chulsanbaatar vulgaris, Kamptobaatar kuczynskii, Kryptobaatar dashzevegi, Nemegtbaatar gobiensis, and Sloanbaatar mirabillis. APP 1994, 39(1), p.134-136 http://www.app.pan.pl/archive/published/app39/app39-134.pdf The 1994 nomenclatural note is presently freely accessible on-line. I've added little from it to the above entry, seeing as that involves more recent research. An exception is the meaning of the generic name. However, as the nomenclatural note contains a proposed diagnosis, it's more than a matter of a straightforward name change. A difficulty with the brief discussion is, at the time, this critter was still interpreted as being within the family of Taeniolabididae. As that's incorrect, some of the points mentioned have lost relevance.

Genus: Djadochtatherium Simpson GG, 1925 'Djadokhta beast' Remarks: A replacement puzzle The following is based upon my reading of Rougier et al, 1997 (p.12). In 1926, Gregory and Simpson referred a second specimen to their genus of Djadochtatherium. Doubts were subsequently expressed by Kielan-Jaworowska (1970) and Rougier and colleagues in this study. Further preparation revealed new information. At this juncture, I admit to finding a sentence puzzling. "The maxilla in the referred specimen is representative of an adult specimen, because P1, the last replaced premolar in multituberculates (GWR personal obs.), is freshly erupted while P2 already shows evidence of extensive wear." This runs counter to what I've read previously, some of which postdates this paper. According to Hahn & Hahn, 2000, the replacement of upper premolars in Jurassic multis they studied began with the P1 (p.101). Szalay, 1965 describes a specimen of Upper Cretaceous Cimolodon maxilla and, in this case as well, the P1 appears to have been replaced before P2 (p.6). The P2 was still in the process of erupting behind its deciduous predecessor. Both these cases contradict Rougier's personal observation.) Based on the conclusion that this referred specimen is adult, the authors found it too small for Djadochtatherium; specifically the maxilla and premolars. They point to a number of details of skull construction, and find more similarity with Nemegtbaatar. As (adult or not) it's too large for that animal, then it may represent a different taxon. However, the preservation isn't good enough for the skull to be used as a type fossil.

Reassigned species: D. catopsaloides Kielan-Jaworowska, 1974 see Catopsbaatar catopsaloides

Links: The Prehistoric Data Files http://www.angellis.net/Web/DFG-mam/Djadochtatherium.htm A sketch by VRW. The Godwana Studios, Tasmania http://www.gondwanastudios.com/stars.htm A pictorial study of the skull. DML, Ben Creisler http://www.cmnh.org/fun/dinosaur-archive/1999Mar/msg00514.html Japanese finds in Mongolia. Amongst the material was apparently an almost complete djado skull. Nevertheless, according to Kielan-Jaworowska et al (2002), this genus is "very poorly known", (p.559).

Species:	Djadochtatherium matthewi Simpson GG, 1925
Aka:	Catopsalis matthewi (Simpson GG, 1925)
Place:	Bayan Zag, Tögrög, Ukhaa Tolgod & Bayan Mandahu
Country:	Mongolia & China
Age:	Campanian, Upper Cretaceous
Remarks:	The following is largely based on my reading of Simpson, 1925. Simpson states that describing the holotype of Djadochtatherium was a "great privilege" (p.1), and the specific name honours one of the people to whom he felt "deeply indebted"; Dr Matthew. The second recipient of this gratitude was Henry Osborn. Both had allowed him to describe "this important form". Take into consideration the poor condition of the critter's teeth left plenty of room for uncertainty, and a cynic might conclude that this junior member of staff was blatantly toadying to his superiors in the hierarchy of the American Museum of Natural History. Both the honoured were above him in the pecking order. That assumption may be attractive to some, but there was a solid reason for Simpson to feel this opportunity was a very rare privilege. A head from the pack Simpson was in the early stages of his distinguished career and, with this study, he became the first person ever to get a crack at the skull of a Mesozoic mammal. Despite about a century of effort, none had previously been found. He didn't quite know that at the time, for he thought this was the second. The privilege was rarer than realised. Richard Owen, in the very late stages of his career, described the partial skull of Tritylodon in 1884, and referred it to Mammalia. The available remains were certainly more like a mammal than anything else for comparison, so it was hardly an unreasonable conclusion. The prevailing view in 1925 saw titrys as multituberculates although, if I may allowed myself a quick leap to page nine, this appears to be implicitly doubted. "Djadochtatherium is plainly very distantly related to this group [the Tritylodon group] if at all." Settling whether Tritylodon was a mammal (let alone a multi) required better specimens for clarification. More informative trity skulls from China were subsequently found, and then interrogated by the pioneering CC Young. In the early 1940s they eventually confessed to being non-mammalian cynodonts. They hadn't been deliberately trying to mislead people, and it's not their fault they happen to be extremely mammal-like in many regards. However, that was in the future. In the parlance of 1925, Multituberculata was understood to include two fundamental divisions; Tritylodontoidea and Plagiaulacoidea. Additionally, there were the haramiyids. It's now eighty years later and their affinities are still debatable. Discovery The original Djado specimen was found hiding in Mongolia by AF Johnson, who was part of the AMNH Third Asiatic Expedition. This series of excavations in the 1920s was an ambitious undertaking, (insanely ambitious comes to mind, as normal people wouldn't have taken part in the war-torn circumstances), and richly rewarded. Most attention focused on the spectacularly good, sexy dinosaurs. Even if we fall into line with then current practice, and recognise Djadochtatherium as being the second known skull of a Mesozoic mammal, it still managed a couple of notable firsts. Tritylodon wasn't accompanied by its lower jaw, and no Meso mammal material had previously been reported from the whole of Asia. The work clearly qualified as a "great privilege" to a serious student of ancient mammals. The preparation of the fossil was accomplished by Albert Thomson. The soft bones and unhelpful sandstone combined to make this task very difficult. Getting to meet Djado Djadochtatherium was a modestly sized multi of around rat-dimensions. The front of its skull was well preserved but a bit crushed, and the tip of the snout was missing (p.2). The sutres (joins) between the various bones were generally clearly visible. Skull (An amateurish sketch of a skull based on Catopsbaatar may help some people with their orientation, although the artist clearly has little talent for drawing: A not exactly convincing skull of a djadochtatheriid.) Much of the wide snout us formed by the unusually large nasals, the rear ends of which form part of the orbit rims. Some of the frontal bones were presumably also present, although distortion of that area is relatively bad (p.3), and leaves some room for doubt. The parietals, which form the roof of the braincase, weren't available. However, they couldn't have been in contact with the maxillae of the upper jaws. The maxillae and the premaxillae were long, with the latter providing more than the usual share of the front of a mammalian mouth. The former were doing sterling work in the cheek area by contributing about half of the impressive zygomatic arch, which commences from a level by the third premolar. No sign of bones termed the lacrymal or jugal could be found, so they were either small or entirely absent. Each of the nasals has a hole towards the front, and these are called foramina. They're the equivalent of holes in the wall of your house for wires and pipes. Simpson wasn't completely sure these had been present during life, but the corresponding positions suggested they probably were (p.4). Subsequent djadochtatheriodean skulls provided plenty of confirmation. They are indeed foramina. Dentary The lower jaw is close to being standard multi issue, if such a phrase makes sense. Multis were evidently well served by the design flair of Jurassic evolution, as they didn't come up with radical innovations. It's cited as being not much different from the earlier Ctenacodon, and even more similar to the later Ptilodus. Both of those genera are known from North America. Teeth Simpson gives an estimate of dental formula but, as time had been very unkind to the teeth, it required some educated guesswork. Better material has since produced a count for djadochtatheriodeans. Kielan-Jaworowska and Hurum 2001 reports: (uppers): 2 incisors, 0 canine, 3 - 4 premolars and 2 molars; (lowers): 1, 0, 2 and 2 respectively. Simpson found only one lower premolar, and perhaps that's all there now is. Nevertheless, these animals persisted in growing a thoroughly pathetic p3 which, in some cases, is small enough to be hidden beneath the front of p4. He thought more than three was correct for the upper premolars and offered four as the probably number. There aren't any traces of the final one, but the gap in the row would've been odd if it were a genuine diastema, and the lower p4 needed a colleague to be there in order to do any work (p.7). Holotype The type fossil, AMNH 20440, is the front of a skull with dentary. It lives in New York. The specific name presumably honours WD Matthew, although this isn't actually stated. Size and cusp formula According to Kielan-Jaworowska et al, 2000 (p.591-592), this is a middling-sized djadoch multi with a skull length of about four centimetres. The only postcanine for which a cusp formula is available is the first lower molar: 4:3 (buccal: lingual). Djadochtatherium is a larger animal than Kryptobaatar and Catopsbaatar, but the snout shares the family characteristics. As with Krypto, the lower premolar is arcuate and there are four uppers. Catops and Tombaatar had only three of those teeth per side. The lower incisor is robustly built.
Reference:	Simpson (1925), A Mesozoic mammal skull from Mongolia. American Museum Novitates 201, p.1-11.

Links: Mesozoic Moments by Kelly Taylor http://www.nhm.ac.uk/hosted_sites/paleonet/vop/mmoments/tday.html Tyrant’s Day: a nice image with a couple of shadowy Catopsalis critters in the foreground. I’m cheating. C. is a valid genus of North American multituberculates. American Museum of Natural History Archive http://digitallibrary.amnh.org/dspace/bitstream/2246/3185/1/N0201.pdf Simpson 1925 is presently freely accessible in pdf format.

Genus: Kryptobaatar Kielan-Jaworowska Z, 1970 'hidden hero' Aka: Gobiaatar, Gobibaatar ('Gobi hero') Kielan-Jaworowska Z, 1970; Tugrigbaatar Kielan-Jaworowska Z & Dashzeveg D, 1978 Remarks: "Gobibaatar has page priority but was selected as the junior synonym by Kielan-Jaworowska in 1980," (McKenna & Bell, 1997). Kielan-Jaworowska et al, 2000 (p.590) report: In terms of size, this is the baby of the family with skull lengths ranging from 2.5 to 3.2cm. It's most similar to Djadochtatherium. The lower p4 premolar is arcuate. Seen from the side, the crown's arch-shaped. Proportionately, this tooth is longer for Krypto and has up to eight serrations. A count of four upper premolars is more than known from Catopsbaatar and Tombaatar and, in contrast to the former, the lingual ridge of the M1 upper molar runs for less than half of the crown's length. The M1 cusp formula is put at 4-5:4:3-5 (buccal to lingual). The snout is somewhat short for the family, but still longer than known from other djadochs (meaning the superfamily), and the lower incisor is also a thinner tooth.

Reassigned species: K. saichanensis (Kielan-Jaworowska Z & Dashzeveg D, 1978) see K. dashzivegi

Link: Ciencia Hoy 6(32) 1996, Los mamiferos mesozoicos, Rougier GW http://www.cienciahoy.org/hoy32/mamiferos_01.htm An article, in Spanish, on Mesozoic mammals. It includes what appears to be a photo of a near complete krypto skeleton.

Species:	Kryptobaatar dashzevegi Kielan-Jaworowska Z, 1970
Aka:	K. saichanensis (Kielan-Jaworowska Z & Dashzeveg D, 1978), Tugrigbaatar saichanensis Kielan-Jaworowska Z & Dashzeveg D, 1978
Place:	Bayan Zag, Ukhaa Tolgod, Tögrög Shiree and Hermiin Tsav I
Country:	Mongolia
Age:	Campanian, Upper Cretaceous
Remarks:	The following has been culled from my reading of Wible and Rougier, 2000 (see Bibliography). These authors state (p.9): "Although K. dashzevegi is known from more skulls than any other Mesozoic mammal, its entire cranial anatomy has yet to be documented in detail, and therefore we present a bone-by-bone description of the skull exclusive of the ear ossicles and reconstruct the course of the major cranial nerves and vessels." This is a promise they fulfilled in ferocious detail. Unfortunately, much of the resultant analysis of the skull is above my head. This particular species was first described on the basis of a partial skull, (minus braincase), and lower jaws from Bayn Dzak, (aka Flaming Cliffs). The original interpretation placed the genus within Eucosmodontidae: Taeniolabidoidea; which was also the case for most multis from the Djadokhta and Barun Goyot Formations of Mongolia. The present view of the affinities dates from 1997. Body Further fossils of this species include a partial skeleton with nearly complete pelvis and back legs. This provided the first record of epipubic ('marsupial') bones from a multituberculate. Rather than being simply typically marsupial, these frontal pointing attachments to the pelvis are typically mammalian, (excepting for placentals, though some basal eutherians retained them). Whilst Mrs kangaroo and many of her sisters may use such pegs in part to support their pouches, the primary function probably had more to do with anchoring leg muscles. A further feature is the presence of a spur of bone from the back of the foot, which was first interpreted as a possible fragment of jaw. This is actually a displaced tarsal spur; an anatomical element found among the existing duck-billed platypus and the like, (monotremes). It's also known from the zhangheotheriid Zhangheotherium and the gobiconodontid Gobiconodon. In monotremes, this spur is generally the privileged possession of only adult males, and this was very possibly also the case for these earlier mammals. The nine specimens at the core of this study were catalogued in the collection of the Institute of Geology, Ulaan Baatar. Along with about 400 other multi specimens, these were harvested by the Joint Mongolian-American Museum Expeditions, which began work in the early 1990s. Most are the right sort of size for this species, but only a handful have as yet been prepared enough to be diagnosable. PSS-MAE 101 looks particularly fine, and is the indirect model for my atrocious attempt at a sketch below. It consists of an exquisitely preserved skull and the front half of an articulated skeleton, which is a great rarity. Many multis Drs Wible and Rougier came to the conclusion that K. saichanensis, (based on a single specimen from a locality presently known as Tögrög Shiree, though also called Tugrugeen Shireh), is actually a largish member of this species. They see it as a junior synonym, and this is advice I've followed. Kielan-Jaworowska & Lancaster, 2004 concur, (p.179). An incomplete and poorly preserved skull has also been recovered from a location known as Hermiin Tsav I, (Kielan-Jaworowska et al 2003), which is also in Mongolia. This site is probably somewhat younger than the other localities, though likely still lower Campanian. K-J et al 2003 (p.275) report that two distinctive types of skull structure are known for this species; wide snouted and narrow snouted. This may be due to the state of preservation, but the age of the individuals could also be of significance. As adults had wider skulls, their snout could well have been relatively narrower. Brain casts (endocasts) of Kryptobaatar The following is based upon my reading of Kielan-Jaworowska & Lancaster, 2004. The star of this section is GI PST 8-2, a resident of The Geological Institute of Ulanbaatar. This individual is known from a nearly complete skull, dentaries and fragments of the skeleton. It's the holotype of Tugrigbaatar saichanensis. However, closer comparisons have now resulted in a consensus that it fits within the variation range of K. dashzevegi, (p.179). After the original description of this specimen, Kielan-Jaworowska was able to remove parts of the skull so as to reveal more of the brain endocast. This had been formed by very coarse sand, but some details of structure are clearly preserved. The length of the skull was probably about 3.5cm, with the somewhat deformed endocast being half that length. Most characteristic are features known as the olfactory bulbs, which are associated with the sense of smell. They're relatively long (6mm) and narrow. The cerebral hemispheres (more informally known as the right and left lobes) are slightly less than 9mm in length. Other specimens were examined and showed possible individual variation in the size of the olfactory bulbs, as has also been observed in Chulsanbaatar vulgaris and Nemegtbaatar gobiensis. In both those cases, those bulbs are relatively shorter; accounting for about 25% of the endocast length, (p.181). However, they're broader. The shape in K. seems most similar to the North American, Ptilodus montanus. Measure for Measure Page 182 contains a table with various measurements and estimates. These suggest the specimen was part of an individual with a body length of about 13.5cm and a weight of something like 65 grammes. However, various other methods of estimating mass have been explored over the years; predictions derived from upper molars, lengths and diameters of limb bones and so on. The answers received are also dependent upon which animals are selected as a basis for comparison. Judged in terms of upper molars, (perhaps close to Tooth Row Length which has been used in rodents), a weight estimate of 37g was calculated, (p.183). Results for the various available limb bones, (the comparison was again with a basket of rodents, with Chinchilla featuring prominently for some reason), range from about 33 - 268g, with an average of around 85. I think I'll settle for three standard mice, (blind or otherwise). That's around 75g plus or minus something or other. This enabled the authors to attempt an estimate of the animal's EQ (Encephalization Quotient), which is related to (but isn't the same as)the proportion of the bodymass accounted for by the volume of the brain. It's not just size that matters [To be useful, Encephalization Quotients have to take into account factors such as body size and build. For example, the brains of small mammals tend to be proportionately larger than in larger critters, seeing as a brain has to have a certain mass in order to function. If EQ only concerned a crude division, then small birds appear to be the most intelligent animals on Earth, and I'd compare poorly with a tree shrew. I'd be about the equal of a mouse. Various other factors conspire to stop EQ being the same as intelligence, which is dependent upon more than simply volume.] Brain boxes For Kryptobaatar, the authors calculated an estimated EQ of between 0.71 - 0.73, (p.185). This is about the same as was previously reached for the eutherian Zalambdalestes. Results obtained in other studies for multituberculates have included Ptilodus, (0.62 -a maximum figure according to Kemp 2005, p.160), and Chulsanbaatar, (about 0.55). A way of life In their conclusions, the authors gave some consideration as to possible lifestyle. The aforementioned olfactory bulbs suggest a strongly developed sense of smell, and the large ear region implies sharp hearing. "Finally very large paraflocculi suggest well-developed sensorimotor adaptations, in particular coordinated movements." These features support a nocturnal interpretation. Holotypes and size Kielan-Jaworowska et al, 2000 (p.590) assures me: The type fossil's known to its friends as ZPAL MgM-I/21, and lives at the Institute of Paleobiology, Warsaw. The skull length reaches 3.2cm. The type of K. saichanensis (now considered a synonym), is PSS 8-2 PST. It's a damaged skull with lower jaws and partial skeleton from Tögrög. Perhaps due to injuries, it opted to remain in Mongolia as an employee of the Institute of Geology, Ulaan Baatar. The 'distinctions' (p.591) included the first upper molar being a bit longer for this proposed species, while the second lower has one more buccal cusp (4:2). Additionally, the premaxilla and maxilla met up in front of the first premolar, rather than level with it. Cusp formulae (buccal to lingual) Uppers: P4 2:4; M1 4-5:4:3-5; M2 1:2:3. Lowers: p4 with 7 serrations; m1 4:3; m2 3:2.
References:	Kielan-Jaworowska (1970), New Upper Cretaceous multituberculate genera from Bayn Dzak, Gobi Desert. In: Kielan-Jaworowska (ed.), Results of the Polish- Mongolian Palaeontological Expeditions, pt. II. Palaeontologica Polonica 21, p.35-49.
	Kielan-Jaworowska & Dashzeveg (1978), New Late Cretaceous mammal locality in Mongolia and a description of a new multitubercualte. Acta Palaeontologica Polonica 23, p.115-130.

Links: DMG Projects, University of Texas http://digimorph.org/specimens/Kryptobaatar_dashzevegi/ A report, photo and animation from the X-ray CT Facility. Kielan-Jaworowska & Lancaster, 2004, Acta Palaeontolgica Polonica http://app.pan.pl/acta49/app49-177.pdf The full paper's freely accessible on-line and the citation's in the Bibliography of this directory. Martin Jehle, Multituberculates: Heyday of the longest lived mammalian order http://www.paleocene-mammals.de/multis.htm As regards these sharp lower incisors, (which are characteristic for multis), "the recent rat-kangaroos, which share this feature, include not only herbivores but also omnivores that feed on plants, insects or even carrion." This article is a fine introduction to the order.

Kryptobaatar dashzevegi PSS-MAE 101 This critter had a skull length of between 2,5 to about 3,5cm.

Species:	Kryptobaatar gobiensis
Aka:	Gobibaatar parvus Kielan-Jaworowska Z, 1970
Place:
Country:
Age:	Upper Cretaceous
Remarks:	This is wild speculation, but I suspect this is from the Gobi, Mongolia. "Kielan-Jaworowska and Hurum (1997) recognized two species of Kryptobaatar, K. dashzevegi and K. saichanensis, with the most substantive difference being the cusp formula (in the labial and lingual rows) of the second lower molar; 3:2 in K. dashzevegi versus 4:2 in K. saichanensis", (Wible & Rougier 2000, p.11). As well as having their doubts upon the correctness of this, (these authors found three cusps for the latter species), this quote also implies that K. gobiensis is probably not considered valid. Logic appears to suggest this material must have been referred to Kryptobaatar dashzevegi.
Reference:	Kielan-Jaworowska (1970), New Upper Cretaceous multituberculate genera from Bayn Dzak, Gobi Desert. In: Kielan-Jaworowska (ed.), Results of the Polish- Mongolian Palaeontological Expeditions, pt. II. Palaeontologica Polonica 21, p.35-49.

Species:	Kryptobaatar mandahuensis Smith T, Guo D-Y & Sun Y, 2001
Place:	Bayan Mandahu red beds, Inner Mongolia
Country:
Age:	Campanian, Upper Cretaceous
Remarks:	Based on several well preserved skulls, at least one of which exhibits a partly exposed endocast of the brain. This is IMM 96BM-II/3, a resident of the Inner Mongolian Museum, Hohhot, (Kielan-Jaworowska & Lancaster 2004, p.181). The location is about the same age as the Djadokhta Formation of Mongolia.
Reference:	Smith et al (2001), A new species of Kryptobaatar (Multituberculata): the first Late Cretaceous mammal from Inner Mongolia (P. R. China). Bulletin de l’Inst. royal des Sci. nat. de Belgique, Sci. de la Terre Supplement 71, p.29-50.

Genus: Tombaatar Rougier GW, Novacek M & Dashzeveg D, 1997 'big hero' Remarks: My orignal opening line for this entry was: 'Tom was a relatively large multi.' Over four years went by until I learned that's the thinking behind the generic name. I didn't know 'tom' is Mongolian for 'big'.

Species:	Tombaatar sabuli Rougier GW, Novacek M & Dashzeveg D, 1997
Place:	Ukhaa Tolgod
Country:	Mongolia
Age:	Campanian, Upper Cretaceous
Remarks:	The following is based upon my reading of Rougier et al, 1997. The Nemegt Basin can be found in southern Mongolia, and several sites have been praised for their excellent Cretaceous fossils. Various paleontologists have probably thanked their lucky stars for the place, and could hardly have hoped for more. But more turned up anyway in the shape of a further locality called Ukhaa Tolgod; 'brown hills'. The main treasure chest there is compacted into about 4 square kilometres of a landscape with low hills and wide slopes; an Upper Cretaceous cemetery boasting 20 sub-localities. "These sites together account for an unprecedented wealth of high-quality specimens representing a diversity of Mesozoic vertebrate groups..." (p.2-3). This opulent praise is in acknowledgement of (as in 1997): 150 dinosaur skeletons, 240 mammal skulls, 300 lizards, nests, eggs, embryonic and juvenile dinos, and several specimens of a weird large bird called Mononykus. Further fossiling has since increased the count significantly. This is a desert landscape today, and that's much how things stood about 80 million years ago too. The excellent preservation of fossils shows these dead animals can't have been lying around exposed for long, as scavenging and weathering would've caused much destruction. The burials must've been quickly conducted and signatures in the rock show sand was the main undertaker. Tripartheid Fossils are most numerous in the lower and middle parts of the section, which are formed of red sandstones. The upper area contains that as well, but river action contributed mudstones and conglomerates. Large dinos are scarce in this attic, which is in contrast to multituberculate mammals and lizards. Big Tom In English, Tom is a shortened name somehow evocative of a big man. Even Tom Thumb was a giant in terms of spirit. A Tom would surely tower over a Thomas. In Mongolian it literally means 'big', as Tombaatar was a large multi for the time of Earth. Imagine a big rat and that'd probably be not quite enough. The single specimen so far is a partial skull, so there's some scope for uncertainty about body size. Nevertheless, it's at the upper end of the local scale for mammals, and the authors place it in the same league as Catopsbaatar and Djadochtatherium (p.1). A length of around 35cm is probably near the mark. A more intimate peep in Tom The preserved length of the skull is four centimetres (p.4), so the complete head must've been longer. Differences from Djadochtatherium and Catopsbaatar include the construction and position of the alveolus for the incisor, I3. This is formed by both the premaxilla and maxilla. The premaxilla has a process at the rear; the skull's higher, the orbit is deeper; and a process at the anterior of the frontal bone is shorter and more slender. As for C., but in contrast to D., the P2 premolar had been dispensed with. P1 and P3 are large (as for D. and not C).. The first molar (M1) also contrasts with Catopsalis. A lobe at the rear on the lingual side is shorter for Tom and houses only two cusps. C. allowed the lobe to extend further forwards along two-thirds of the crown, and it supported four cusps. A short selection from the skull Preserved are the upper jaws and front part of the braincase (p.5). There is damage and bits are missing (especially from the right), but even such faults can be useful. For example, the absence of some of the bone from the nasal cavity conveniently allowed a look at the inner workings; the maxillary sinus and infraorbital canal. The premaxilla has a tall facial process (p.7), and this contacts only with the nasal and maxilla bones. No septomaxilla is involved. As usual for multis, the maxilla is impressively developed, forming much of the jaw and zygomatic arch (cheek bone). It predictably houses the postcanine dentition (although a canine wasn't actually present) and, to the front, even lends a hand with part of a hole for the root of an incisor (p.8). Eager to be of service, this bone also found time to provide the rear of the nasal cavity, part of the cranial cavity and a floor for the choanal passage; a feature I'm present unfamiliar with. Another of its duties concerned construction work on the bony palate ceiling for the inner sanctum of the mouth. A bit of the back of the maxilla is broken off, but comparisons with relatives suggest not much is absent. The upper snout is the responsibility of the aptly named nasal bone and, as usual for multis, this element is very large (p.9). It tapers from rear to front, and connects with the premaxilla, maxilla, frontal and probably the lacrimal. There's a small rectangular fragment of a bone, and that's its likely identity. The situation is obscured by damage. Dentition 'Big hero' suffered both tooth damage and loss, but the uppers from both sides combine to contribute information. The formula per side is (p.10): (uppers): 2 incisors, 0 canine, 3 premolars and 2 molars. That's one less premolar than some of its relatives. Table 1 (p.23) contains a range of measurements, and the following are lengths from the left jaw: P1 1.45mm; P3 1.92; P4 3.73; M1 5.46; M2 3.7. Incisors There are two incisors, I2 and I3. The ancestral I1 was dispensed with earlier along the line. This tally is shared with cimolodontids (as used in this study). Both the left and right I2s had been snapped off, but it was clearly a narrow tooth. As the root curves back to meet the maxilla, the crown would've been strong. What remains of the left incisor has thin enamel limited to the front. I3 is a backwards curving cylinder positioned between the premaxilla and maxilla. There's no trace of enamel remaining. This may be an effect of weathering, as this was completely exposed to the elements. Premolars This trio consists of P1, P3 and P4. A short diastema separates the first pair, and there's no indication a tooth ever developed there. All premolars are double-rooted (p.11). For P1, the first root slants to the front and is a bit compressed; typical for this position in multis. Its rear partner is wider. The crown has a triangle of cusps with two set lingually and one labially. Enamel damage obscures the labial side. P3 leans a bit backwards. Four cusps are on its low crown, with two on each side. The lingual cusps are the taller. Both roots of the P4 are compressed. It's a narrow tooth with a line of five cusps running from the rear labial area of the crown to the lingual side of the front. The centrally located third is the strongest. Two further cusps form a short row labial of the front three. These are small. More may have been positioned on the lingual surface at the back, but that area of the tooth is heavily worn. Molars There are two per side and M1 is about 50% longer than its colleague. These teeth on the left are best preserved. All cusps are broad and near hexagonal in shape. Multis varied with styles of cusp, and this may have reflected differing tastes to some degree. For example, some favoured more crescent-shaped cusps. As is typical for upper molars, the cusps are arranged in three longitudinal rows, with the formula in this case (labial to lingual) being 4-5-2. The labial cusps increase in size from the front to the third, while the final one breaks the trend. A similar progression occurs in the middle row going back to the fourth, and this wins the prize for the largest cusp on the crown. The lingual row is restricted to a lobe on the rear half of the tooth; a bit of a kink supporting just two cusps. The shorter M2 also has three rows, but the labial one is more like a ridge without distinct cusps (p.12). The middle row comprises two particularly large cusps, while a trio of relatively tall but smaller ones are found on the lingual margin of the crown. Holotype The type fossil, PSS-MAE 122A, is the front of a skull in the collection of the Mongolian Academy of Sciences, Ulan Baatar. A couple of probable colleagues are among masses of mammal skulls awaiting preparation. The specific name is Latin for 'sand'. There was a lot in the area at the time, and that's also the case today. The specimen was arrested during fieldwork in 1993. Size and cusp formula buccal - lingual) Kielan-Jaworowska et al 2000 provides a convenient summary on page 592. The skull length of this giant ammounts to seven centimetres, and further specimens may await formal identification. The cusp formula for the M1 is presently unique within the superfamily: 4:5:2. An alveolus for the third upper incisor is in a rather archaic position between the premaxilla and the maxilla. Extant mammals wouldn't be seen dead with such an arrangement. We keep all our incisors housed in the premaxilla. Additional notes Some similarly aged skulls from Bayan Mandahu, China, may belong to this genus, (Kielan-Jaworowska et al 2003, p.276).
Reference:	Rougier, Novacek & Dashzeveg (1997), A new multituberculate from the late Cretaceous locality Ukhaa Tolgod, Mongolia: Considerations on multituberculate interrelationships. American Museum Novitates 3191, p.1-26.

Links: DMG Projects, University of Texas http://digimorph.org/specimens/Tombaatar_sabuli/ Another gem from the Lone Star State lab. The AMNH, New York, Dispatches 16.7.1998, Amy Davidson http://www.amnh.org/exhibitions/expeditions/gobi/dis716.html Excavations in Mongolia. This includes a photo of a well preserved tomba skull. If anyone’s got one they don’t want, please feel free to send it in. Ukhaa Tolgod has now yielded more mammal skulls (over 800) and skeletons than any other Mesozoic site in the world. American Museum of Natural History Archives http://digitallibrary.amnh.org/dspace/bitstream/2246/3591/1/N3191.pdf Rougier et al, 1989 is presently freely accessible in pdf format.

Other reports: Xxxxxxxxxxxxxxxxx Xxxxxxxxxxxxxxx

Mammalian Phylogeny, The University of Louisville http://www.louisville.edu/medschool/anatomy/Research/NSF/MammalPhylogeny.htm This is a bonus link for any specialists or masochists. Collaborative Research: A Morphological Database for the Higher-Level Relationships of Fossil and Recent Mammals. Here you can check up on the orientation of narial apertures, should you feel inclined.

Help: Should anybody have any further information, I'd be pleased to hear of it. Regarding references and Bibliography: I haven't and can't verify all the references, so beware. Traditional papers used in constructing this page are in the bibliography. If you feel these are too few, then send some more. With thanks to all the featured sources. back to top Trevor Dykes, December 2001. Last update: 27.12.2009. Ktdykes@arcor.de

With further thanks due to: The Prehistoric Data Files http://www.angellis.net/Web/PDfiles/marsups.pdf A handy metaindex, now in pdf format. Martin Jehle, Paleocene mammals of the world, Class Mammalia http://www.paleocene-mammals.de/pal1.htm A great help through the mazes of multidom, with much more besides. The Paleocene is the age which followed the Mesozoic, between about 65 and 54 million years ago. BIOSIS, The Index to Organism Names http://www.biosis.org.uk/triton/indexfm.htm The Society of Vertebrate Paleontology BFV Online, (John Damuth) http://www.bfvol.org/ Mongolian multis seems to be a bit of a blind spot, as far as original citations are concerned. Nevertheless, later references for a number of these genera are listed.

Bibliography: Averianov AO & Archibald JD (2003), Mammals from the Upper Cretaceous Aitym Formation, Kyzylkum Desert, Uzbekistan. Cretaceous Research 00 (2003), p.1-21. Hahn G & Hahn R (2000), Multituberculates from the Guimarota mine, p.97-107 in Martin T & Krebs B (eds), Guimarota - A Jurassic Ecosystem, Verlag Dr Friedrich Pfeil, München. Hurum JH (1998), The inner ear of two late Cretaceous Multituberculate mammals, and its implications for multituberculate hearing. Journal of Mammalian Evolution, 5 (1), p.65-93. Kemp TS (2005), The Origin and Evolution of Mammals, Oxford University Press, pp.331. Kielan-Jaworowska Z (1994), Nomenclatural note, A new generic name for the multituberculate mammal 'Djadochtaterium' catopsaloides, Acta Palaeontologica Polonica, 39(1), p.134-136. Kielan-Jaworowska Z & Hurum JH (2001), Phylogeny and systematics of multituberculate mammals, Palaeontology, Vol 44 (3), p.389-429. Kielan-Jaworowska Z, Hurum JH, & Badamgarav D (2003), An extended range of the multituberculate Kryptobaatar and distribution of mammals in the Upper Cretaceous of the Gobi Desert. Acta Palaeontologica Polonica 48(2), p.273-278. Kielan-Jaworowska Z, Hurum JH, Currie PJ, & Barsbold R (2002), New data on anatomy of the Late Cretaceous multituberculate mammal Catopsbaatar. Acta Palaeontologica Polonica 47(3), p.557-560. Kielan-Jaworowska Z, Hurum JH & Lopatin AV (2005), Skull structure in Catopsbaatar and the zygomatic ridges in multituberculate mammals, Acta Palaeontologica Polonica, 50(3), p.487-512. Kielan-Jaworowska Z & Lancaster TE (2004), A new reconstruction of multituberculate endocranial casts and encephalization quotient of Kryptobaatar, Acta Palaeontologica Polonica 49(2), p.177-188. Kielan-Jaworowska Z, Novacek MJ, Trofimov, BA & Dashzeveg D (2000), Mammals from the Meozoic of Mongolia, p.573-626 in Benton MJ, Shishkin MA, Unwin AM & Kurochkin EN (Eds.), The age of dinosaurs in Russian and Mongolia, Cambridge University Press. McKenna MC & Bell SK, (1997), Classification of Mammals Above the Species Level. Columbia University Press. Rougier GW, Novacek MJ & Dashzeveg D (1997), A new multituberculate from the Late Cretaceous locality Ukhaa Tolgod, Mongolia. Considerations on multituberculate relationships, American Museum Novitates, 3191, p.1-26. Simspon GG, (1925), A Mesozoic mammal skull from Mongolia, American Museum Novitates, 201, p.1-11. Szalay FS (1965), First evidence of tooth replacement in the Subclass Allotheria (Mammalia), American Museum Novitates, 2226, p.1-112. Weil A (1999), Multituberculate phylogeny and mammalian biogeography in the Late Cretaceous and earliest Paleocene Western Interior of North America, Ph.D. Dissertation, University of California, Berkeley, p.1-243. Wible JR & Rougier GR (2000), Cranal anatomy of Kryptobaatar dashzevegi Mammalia, Multituberculata), and its bearing on the evolution of mammalian characters. Bulletin of the American Museum of Natural History, 247, p.1-124.