Primary Faiyum of Egypt location for 1907 American fossil expedition (V.Morgan after T.Bown/M.Krause).

The American Museum's Faiyum of Egypt expedition party reached their fossil base camp, situated on the north rim of the Faiyum Depression above Birket Qarun and along the base of the Gebel el Qatrani escarpment, on February 5, 1907. They began excavating in quarry pits left from the earlier British expeditions which collected fossil vertebrates such as the Ancodon, Arsinöitherium, Crocodilus, Geniohyus, Megalohyrax, Möeritherium, Paleomastodon and Sagatherium. Author and illustrator* John R. Lavas describes these and other ancient mammals of the Faiyum in the essay that follows:

What occurred was this. Lord John had shot an ajouti -- which is a small, piglike animal -- and, half of it having been given to the Indians, we were cooking the other half upon our fire. There is a chill in the air after dark, and we had all drawn close to the blaze. The night was moonless, but there were some stars, and one could see for a little distance across the plain. Well, suddenly out of the darkness, out of the night, there swooped something with a swish-like an aeroplane. The whole group of us were covered for an instant by a canopy of leathery wings, and I had a momentary vision of a long, snake-like neck, a fierce, red, greedy eye, and a great snapping beak, filled, to my amazement, with little, gleaming teeth. The next instant it was gone -- and so was our dinner. A huge black shadow, twenty feet across, skimmed up into the air; for an instant the monster wings blotted out the stars, and then it vanished over the brow of the cliffs above us.

This tantalising sighting of a pterodactyl by a group of British explorers led by the indefatigable "Professor Challenger" is recounted in such evocative style by Sir Arthur Conan Doyle, in his classic 1912 novel The Lost World that many readers at the time unquestioningly accepted it as an actual expedition narrative. The first edition appropriately contained photographs which purported to show the newly-discovered land. Before being published in novel form, The Lost World was serialised in the Strand Magazine and eager London readers queued three blocks deep to buy the latest report from the pen of Doyle's alias Ed Malone, the Daily Gazette reporter who had accompanied the exploring party. One American university later even organized an expedition to locate the surrealist world, an isolated plateau in northern Brazil putatively inhabited by Jurassic dinosaurs and a variety of prehistoric mammals that had made their way up to the summit over subsequent geological periods.

Inspiration for The Lost World stemmed from the spectacular series of plateaus or "tepuis" of the Gran Sabana that separates the Orinoco and Amazon river systems. Tepuis are rugged, uplifted blocks of land girdled with perpendicular cliffs set in rolling landscapes of grassland (even today, few tepuis have been properly explored but, as some vindication of Doyle's original concept, those that have been are sometimes found to harbour endemic species unique to each plateau). Maps of 1912 showed many such tracts of little-explored regions awaiting discovery, and with a pre-World War I public still enjoying the rather romantic age of empires and expeditions to exotic regions, they (including not just a few academics) had yet to be convinced that prehistoric animals, dinosaurs or otherwise, did not still linger in some forgotten, far-flung corner of the globe.

Doyle's fictional "Challenger" expedition was supposed to have taken place in around 1907 (although clues in the text also suggest 1908, using a 100-year calendar--J.R.L.11Feb.'00), but in that very year on the other side of the Atlantic in North Africa, an entirely authentic "Lost World" was being actively explored by members of an American Museum of Natural History from New York. Many of the extinct denizens of this true "Lost World" of the Old World were so unusual they may well have flowed from the deft pen of Conan Doyle himself. In 1907, the study of extinct forms of vertebrate animal life (that is, the science of vertebrate palaeontology), had already developed to a reasonably advanced level both in the United States as well as a number of European countries. But though American palaeontologists had undertaken quite extensive explorations of both Mesozoic and Cenozoic strata within the United States, they had gained no practical experience at sites abroad.

The variety of vertebrate life evident in the Faiyum could naturally not be sustained in the environmental conditions that now characterise northern Africa. Geological evidence from the area has been the subject of debate in the past, but recent research has thrown further light on the topic. An assemblage of fossil small bird life described recently (Rasmussen, Olson & Simons, 1987) indicates these fossils were trapped in deposits formed in slow water current conditions with their associated freshwater vegetation. The avian fauna is very similar to that found in tropical swamp and river areas of Central Africa. In late Eocene and early Oligocene times, it is suggested that the Faiyum was a subtropical to tropical lowland coastal plain with seasonal rainfall, and contained one or more slow-flowing, swampy rivers which were heavily overgrown with papyrus, reeds, and floating plants like Salvinia and water lilies (Bown & Kraus, 1988; Rasmussen, Olson & Simons). Terrestrial vegetation included liana vines, tall trees, and probably mangroves (Wing, 1982). The only part of Africa that today most closely approximates these conditions is a section found in Uganda, northwest of Lake Victoria. Previously, some authors argued that the Faiyum was always treeless and arid even in ancient times. But those who have recently described the geology of the Faiyum and its assemblage of fossil primates agree with a tropical, heavily-forested concept of the ancient Faiyum because of the type of environments now occupied by similar species (Bown, et al., 1982).

The most unusual members of the ancient Faiyum include several of the larger mammals, many of which were very archaic, and probably endemic to that region. Some of the better known -- Arsinöitherium zitteli, Möeritherium lyonsi, Palaeomastodon, Phiomia, and the basilosaurs -- are worthy of brief discussion at this point. The giant subungulate Arsinöitherium zitteli from the late Eocene and early Oligocene (40-32 million years ago, or mya) was a 3.4 meter long creature that stood taller than the largest rhinoceros of modern times. Its large skull featured a huge pair of bony horns growing upon the snout side-by-side rather than in an anteroposterior line, as occurs in rhinoceroses. Although outwardly rhinoceros-like in appearance, Arsinöitherium was not at all related to that lineage. Instead, it is classified as one of the largest of the few known representatives of the order Embrithopoda -- an evolutionary dead-end with no known ancestors or descendants. The exact placement of this order is not yet known, but it is generally positioned closest to the proboscideans and their relatives, the tiny rock hyraxes.

The Arsinöitherium probably inhabited the Faiyum's marshy areas and the surrounding low jungle and thick forest vegetation. It was heavily-built and slow-moving. Its dentition consisted of 44 high-crowned teeth that formed a continuous series from molars to incisors, these being considered primitive in construction though suitable for coping with a diet of the low foliage that grew around the margins of swamps and forests. Other remains of this unusual animal have been found recently in Romania and Turkey, so it was not limited to the Faiyum in distribution. The generic name comes from the Greek-born, Egyptian princess Arsinöe (wife of the pharaoh Ptolemy I, Soter, of the late 4th and early 3rd century, b.c.), whose name also predominated a number of religious centres within the Faiyum oasis.

Another large Faiyum mammal was Möeritherium lyonsi Andrews, also known previously as the "Dawn Elephant" because it was once believed to have been the direct ancestor to later proboscideans. Now it is instead considered to be from an aberrant line that was none-the-less very close to the true elephant lineage (which still remains unknown). Several closely-related species have been found from localities north of the ancient Lake Moeris from which the animal derives its generic name. None has been found elsewhere, so the möeritheres may have been endemic to the Faiyum. The Möeritherium was a heavily-built, stout, piglike animal that stood about 1 meter high at the shoulder. Short, thick legs terminated in broad feet with flat hooves. There was no perceptible trunk, but there was a highly mobile upper lip similar to that of a tapir, and a set of simple, cusped teeth. Both sets of second incisors in both jaws were enlarged into short, boar-like tusks. Möeritherium possibly frequented thickets at the edges of the Faiyum's lakes and marshes, living not unlike a hippopotamus or tapir during the late Eocene (45-36 mya).

The early Tertiary Faiyum sediments have also produced important proboscideans more advanced than the Möeritherium -- the mastodonts. The genera Palaeomastodon and Phiomia are from the late Eocene and early Oligocene age and are closely related to each other, their descendants possibly including modern elephants. Both were much larger animals than Möeritherium. Phiomia stood up to 2.4 meters high at the shoulder, and had long legs and an elephant-like skeleton. The nasal bones of Phiomia were retracted up on a bulbous skull, suggesting the existence of a trunk. There were two downward projecting tusks from the upper jaw, as well as two horizontal tusks protruding from the lower jaw. The upper jaw tusks may have been derived from the much shorter tusks of Möeritherium. Faiyum proboscideans are particularly valuable due to the general paucity of fossil mastodonts in later strata, until the record picks up again elsewhere in lower Miocene strata.

The most important marine mammals from the Faiyum are undoubtedly the basilosaurs, previously known as the zeuglodont whales. Over 240 skeletons of the whale-sized Basilosaurus are known, and until recently, they have always been restored showing whale-like flippers. In 1990, however, a Basilosaurus skeleton with external hindlimbs and individual pelvic limb, foot, and toe bones was described for the first time (Gingerich, Smith & Simons, 1990). Modern whales were probably descended from land-dwelling hoofed-ungulates called mesonychids (Van Valen, 1966), but until these new Faiyum finds, there was nothing to link these animals with modern cetaceans (whales lack external hindlimbs though they retain vestigial ones as embryos). The tiny legs of the basilosaurs may have been derived from mesonychids, and may have been used as copulatory guides or to aid movement through shallow water.

In 1994, two other ancestral whales where found in Pakistan which provide evidence on the position of the basilosaurs within the cetacean lineage. Ambulocetus natans was older than Basilosaurus, and was a small seal-like animal with a typical basilosaur-like skull (Thewissen, Hussein & Arif, 1994). It had well-formed fore- and hindlimbs and it may have been able to move on land as well as swim (which is assumed to be the main means of locomotion). Progression through water was probably by a combination of paddling the hindlimbs and undulating the spine up and down. The Archaeocetes (the group which includes the basilosaurs) now thus appears to fall in line somewhere between Ambulocetus and the modern toothed (Odontoceti) and baleen (Mysticeti) whales.

The Eocene whale Rodhocetus was also found in Pakistan later in 1994 (Gingerich, Raza & Arif, 1994). It shows adaptations more suited to swimming than those of Ambulocetus, and thus appears to lie between Ambulocetus and Basilosaurus, because it retains ancestral features of parts of the vertebral column and pelvic girdle from the former, but shows adaptations for swimming more akin to the latter. These three whale fossil types have supplied much-needed information on the chronology of cetacean evolution, and on how and when their terrestrial ancestors returned to the oceans.

Installment No. 3 will continue Granger's narration of the 1907 American expedition's hunt for fossils in the Faiyum. The work party is now sited at their base camp near two quarries (A & B) previously worked by the British. The area is at the foot of the Gebel el Qatrani escarpment about 10 kilometers west northwest of the Qasr el Sagha Temple ruins. Birket Qarun is about 15 kilometers to the south. Quarries A & B are located on either side of a wadi later known as Wadi Markgraf, so named after Faiyum fossil collector Richard Markgraf. Wadi Markgraf is a north-south draw that connects to a larger wadi, later known as Wadi Granger, which courses into the bottom of the Faiyum Depression. (See map below.)

*John's major artistic influence has been Zdenek Burian (1905-1981), the excellent Czech artist whose impressive work is familiar, but who remains little-known to the public. For further information, see http://www.bpib.com/illustrat/burian.htm.