6. Sam's Stomach Stones

Stomach stones. The phrase itself sends a knot to my stomach. I can barely imagine carrying a gut full of rocks, and I don't believe I would be much good at selecting and swallowing them.

Several times in my life I have raised chickens for both meat and eggs. These descendants of dinosaurs endlessly search the ground for tidbits of food, tidying up their surroundings. In the process, they deliberately take up sand grains for grit in their gizzard. Chicken food therefore usually includes a component of sand--essential for caged chickens that cannot forage for themselves. Evidently chickens require grit for proper digestion.

I now raise cockatiels (small parrots), an acceptable suburban substitute for chickens. Cockatiels, too, require grit in their diet. Once I forgot to fill their grit bowl, and it lay empty probably a month or longer before I discovered it was bare. I casually filled the bowl, and my four usually sedate cockatiels went into a frenzy before I could remove my hand from the cage. They were desperate for the grit, and I suppose if I had continued to deny them this ingredient in their diet the consequences would have been severe--like a conspiracy to es-cape.

Why do chickens and cockatiels take up grit? To aid digestion. The grit becomes lodged in the gizzard, which is a special chamber at the rear end of the stomach. With its complement of grit, the muscular contractions of the gizzard crush tough seed coatings (in the case of chickens) and perhaps woody or cellulitic stems and branches (in the case of dinosaurs) that have already been softened by chemical digestion in the stomach. Thereafter, the food passes to the intestine, the principal region of absorption and transfer of nutrition.

Seed-eating birds like chickens and cockatiels generally possess the specialized chamber of the stomach called the gizzard, or more technically, the ventriculus. This highly muscular organ is lined with a hardened horny material called koilen. The lining makes the inside of the gizzard somewhat rigid; the heavy muscles surrounding the koilen can produce considerable pressure on the contents in the cavity of the gizzard. This action crushes seeds and plant materials before they are passed to the intestine for further digestion. Gizzards are found only in seed-eating and plant-eating birds; meat-eaters and fruit-eaters do not have (and do not need) gizzards.

Most, or perhaps all, birds with gizzards use grit to facilitate digestion, although the function of the grit has never been fully explained. Apparently, birds with gizzards take up grit when it is available, but they can survive without it. Its real function is elusive. Most biologists assume that grit somehow participates in crushing the contents of the gizzard, a process called trituration by physiologists. Exactly how the trituration of food in the gizzard is improved by grit is an open question, however. Perhaps the grit rotates with muscular contractions of the gizzard, slicing and cutting into the food.

Seed-eating birds have gizzards, and birds are descendants of dinosaurs. Perhaps, then, plant-eating dinosaurs had gizzards. The similarities invite comparison, but differences between birds and sauropods suggest caution. For example, birds are toothless, but sauropods had teeth. Modern birds with gizzards eat seeds, but few (if any) Jurassic sauropods took seeds because nutritious seeds were not available until angiosperms evolved in the Cretaceous Period. Sauropods declined precipitously at the end of the Jurassic and survived only in diminished diversity and number through the Cretaceous. The rise of the flowering plants, the angiosperms, came at the very beginning of the Cretaceous. Thereafter, flowering plants expanded dramatically; they are now, by far, the dominant terrestrial plant.

Some paleontologists have suggested that Cretaceous sauropods, quite unlike many other dinosaur groups, became increasingly restricted to relict habitats left over from the Jurassic. Relicts of the Jurassic would be regions rich in ferns, cycads, and conifers. There are modern analogues: for example, the rich fir and cedar rain forests (with ferns blanketing the forest floor) that thrive along the coasts of Washington and Oregon. Perhaps sauropods would do well there. In any case, we do know that seeds of flowering plants were not available to the sauropods of the Jurassic, when these plant-eating dinosaurs reached their zenith.

There are other important differences between birds and dinosaurs, as well. Birds are small and demand rich foods for their high metabolism. Dinosaurs were large and, because their metabolic demands were lower, probably did not require food as rich as did birds.

A herd of Diplodocus.

Overall, the bird analogy is useful but rough. Are there other organisms alive today that use stones in their digestive tracts? Yes there are: crocodiles, turtles, and some lizards. According to one study, nearly all adult crocodiles of one population had stomach stones. The function of stones in these carnivorous reptiles is debatable, at least for crocodiles. Some authorities suggest that they are used as ballast for buoyancy adjustments like the lead weights that scuba divers use to make them sink. The digestive functions of stomach stones in crocodiles may therefore be entirely incidental. But at least in crocodiles, the stones are truly stones--not just small grit.

Similarly, the function of stomach stones may have been incidental in dinosaurs, too. The hypothesis that stomach stones in sauropod dinosaurs had no function (i.e., were only incidental) is as difficult to test as the obverse, but it cannot be ruled out. Sauropod skeletons sometimes contain stomach stones, but we cannot be certain that sauropods had gizzards. Nevertheless, there is considerable circumstantial evidence (especially from the excavation of Sam) that they had gizzards and used stomach stones, the dinosaur equivalent of grit. Gastroliths  is a better term than stomach stones  because the stones in Sam's digestive tract may not have been confined to the stomach or a chamber of the stomach like a gizzard.

Paleontologist Robert Bakker has argued for gizzards in sauropods and for the use of stomach stones as an aid to digestion. His conclusions, like those of many before him, are based on largely circumstantial (and, in my view, weak) evidence, since no sauropod skeleton before Sam has ever been fully documented as having gastroliths in the visceral cavity.

Sam's gastroliths generally support Bakker's conclusions, but I am still cautious in making physiological extensions from that evidence. Nevertheless, Sam's excavation has furnished some exciting evidence that can be used to bolster the arguments of gizzard advocates. It did confirm that sauropods had (and presumably used) stomach stones, but this does not automatically mean that they had a distinct chamber like a gizzard. Absent soft-tissue evidence we cannot equate the function of dinosaur stomach stones to the presumed function of such stones or grit in birds and crocodiles. Indeed, I question the notion of any direct participation in food grinding by stomach stones in the sauropod dinosaurs, which is the function usually attributed to them. And the idea that they were used for ballast is negated by the overwhelming evidence that sauropods lived on dry land.

Sam had stomach stones--so many that they became a nuisance in the excavation. We have identified and mapped more than 240, ranging in size from about an inch to four inches in diameter, with a median size of about two inches. Most were oblong, and some were spherical. A few were flattened and roughly discoid. Each one was carefully exposed in the sandstone surrounding the skeleton. Each was photographed, its position plotted and mapped, and labeled before removal. Gastrolith documentation was thus meticulous.

Some groups of gastroliths we excavated with the surrounding rock, to keep them in position for later reference. Others were deliberately removed in contact with bones, to preserve their original positions and attitudes. Still others were unknowingly removed along with the bones in the blocks we had established to ensure the integrity of the bones. Some of these, surely, have not yet been "discovered" and won't be until the bones are fully prepared for study with the surrounding sandstone removed. With full preparation of the skeleton, we expect to find more, but the total will probably not be radically different from the 240 we have logged to date.

Location of gastroliths. Simplified quarry map, with ribs shown in original positions. Gastroliths were found mostly in two clusters: one near the front of the pelvis, the other farther forward. In all, more than 240 gastroliths were recovered, but not all are plotted on this map.

Close-up of the anterior region of the skeleton. The semilunar cluster of gastroliths at right center was protected from river currents by a vertebra (not shown on this map so that gastroliths would be evident) that had fallen on top of them. That bone ("Murphy's vertebra") was dorsal vertebra no. 3. Also not shown here are the four isolated and heavily eroded cervical vertebrae that were found (geologically) downstream far to the right of this map.

With the discovery of each stone, the progress of our excavation slowed substantially; without these stones (or with a more casual attitude as to their value) we could have completed the excavation of Sam's skeleton at least a year earlier. But this unexpected bonus in the excavation became a focus of attention when we realized that Sam's gastroliths would be the first to be fully documented for any sauropod dinosaur. And we knew that they could potentially play a role in deciphering the burial history of the carcass, that they could assist interpretation of Sam's anatomy, and that they might suggest something about Sam's feeding habits--and that of sauropod dinosaurs in general.

How can we be sure that these stones are genuine gastroliths? Skeptics, myself included, claim that most stones identified as gastroliths are river rocks that were carried by streams and deposited near the dinosaur skeletons with which they are associated. Therefore, by this line of reasoning, gastroliths may not demonstrably be associated with the dinosaurs. Their presence could be entirely coincidental, perhaps owing to the carcass acting as a barrier to sediment transport or perhaps owing to an eddying effect around the body, which would prompt an abrupt drop in stream velocity, thereby causing stones sliding along the bottom as bed load to drop in the sands surrounding the skeleton. This abiotic explanation for the occurrence of alleged gastroliths may be appropriate in many sites where dinosaur skeletons are common. Thus, a demonstration that Sam's gastroliths are genuine stomach stones, rather than impostors deposited by streams, cannot be marshalled as evidence that all purported gastroliths are real. Instead, each occurrence must be subjected to the same test: can we legitimately rule out deposition by streams or other modes of deposition?

The gastroliths associated with Sam's skeleton surprised us. At first we casually removed them from the excavation and mapped their occurrence. Because no other sauropod excavation in my acquaintance had reported gastroliths in direct association, I was unprepared to accept the first cluster of stones that we found near the pelvis as an indication of more to come. I was naturally skeptical then, but because Sam's skeleton was buried entirely in sand, there seemed to be no other explanation.

Sam's skeleton lay in the middle of a layer of sandstone twenty feet thick, bounded below by shale and capped above by a different, more uniformly cemented and harder sandstone. Except for the gastroliths in the pelvis of Sam's skeleton, no sedimentary materials larger than sand were found in this ancient sand bar. If the stones, generally the size of a plum, were stream-deposited, there should be gravels and cobbles present in the sandstone, too, in gradational layers.

Ordinarily, stream deposits display gradations in grain size, a feature geologists call "graded bedding." It is not unusual to find gravels and cobbles associated with stream sands, but in vertical sequence. Such changes in grain size are almost never abrupt; instead, the sands are overlain by increasingly larger or smaller (clay-size) sedimentary materials that reflect a change in stream velocity and carrying capacity (bed load and suspended load). Where the change is abrupt, without gradation, either no sediments of intermediate sizes were available for transport (a rare and unusual situation), or other modes of deposition must be considered: for example, rafting of cobbles caught in floating blocks of ice or cobbles being carried to their resting place in an animal's carcass. Deposition by stream action and transport to the site in the body of a dinosaur are, however, the only two credible explanations for the stones we excavated. Other explanations, such as ice-rafting, lack evidence and need not be considered as possible.

In our quarry stream deposition seemed a very unlikely explanation for the stones. There were no gravel or cobble layers or "lenses" above or below the skeleton, nor were there any within a radius of a hundred feet from the excavation. The gastroliths in the pelvis lay directly on sand. Under ordinary depositional circumstances, stream-deposited gravels and cobbles dropped from a current would come to rest on other sediments coarser than the sand. The fact that these stones were "matrix-supported" was thus another important piece of evidence against their having been stream-deposited.

All indications thus pointed to the conclusion that the stones had been contained by the carcass, presumably in the digestive tract, that they came to rest on the sand as Sam's carcass decayed, and that the contents of the digestive tract spilled out and settled on the sand that supported the skeleton. My skepticism was dwindling. I was becoming a believer in sauropod gizzards. These were genuine gastroliths, presumably from Sam's gizzard, and we were pleased to discover that we had real documentation, for the first time, that sauropods (or at least Sam) did indeed use stomach stones.

101 gastroliths. These constitute a portion of the anterior set, presumably associated with the crop. Many other gastroliths from this set were collected in situ, with rock and bone, and could not be included in this photograph. The largest gastrolith found in the entire site has been placed in the center of this photograph, partially obscured by another gastrolith. (Scale divisions are 1 centimeter.)

Before the discovery of Sam's gastroliths only five sauropod skeletons were reported with associated gastroliths, and the documentation in each case was meager. Three were in the Morrison Formation in North America and were only casually described. Two were in the Tendaguru beds of Tanzania, and the gastroliths were accurately measured and described. In one of the Tanzanian excavations the gastroliths were found in the neck region of a skeleton of Barosaurus. As there were only five stones, these gastroliths were little more than curiosities, but their position in the neck has an important bearing on arguments presented below that Sam had a crop as well as a gizzard.

After this first discovery of a clump of two dozen stones in the pelvic region of Sam, we did not expect to find more. The pelvis, where this cluster occurred, was exactly where a gizzard would have been in life--if the digestive tract of a sauropod resembled that of a chicken. But a year later Peggy and Wilson Bechtel, the quarry excavation supervisors, began to uncover more gastroliths--some in clusters--and by the dozens. Like the ones found within the pelvis, these stones were all matrix-supported, not associated in any way with gravels or cobbles that would have been deposited by stream action.

With each new gastrolith the map of discoveries became more and more tantalizing: the distribution of gastroliths around the skeleton did not fit the notion that they were confined to a gizzard, and we were finding many more than we expected. By the conclusion of the excavation in 1992, we had mapped more than 240 gastroliths, distributed in two general clusters: one tight group of about two dozen stones in the pelvic region, and a larger, more scattered group farther forward, ranging from roughly the middle of the rib cage toward the base of the neck. What did this mean?

As I mentioned earlier, nowhere in the quarry or in nearby sediments could we find any evidence to support an explanation that these stones were river-deposited--or ice-rafted, for that matter. We therefore held a working hypothesis that these were indeed gastroliths because they couldn't be anything else. They were brought to this site in Sam's body, where they spilled out as the carcass decayed. But this conclusion was supported by several positive lines of evidence, as well.

First, we found sets of gastroliths, including one from the pelvis and several from the more forward position, that came to rest in a line, in contact with each other as though they had been contained by soft-tissues when they came to rest on the sand. Some of the individual stones in these sets were arranged in overlapping fashion, and some were on edge rather than lying flat. These orientations are almost impossible in stream beds, except where a stream bottom is mantled by river stones. Deposition of single lines of stones in isolation seems to be nearly impossible in stream-generated sediments.

This realization that the stones actually reflected the position of stomach anatomy led to another surprise: some of these lines of gastroliths lay well beyond the outline of the carcass as indicated by the skeleton. They seem to indicate the position of entrails distributed by a scavenger that had pulled the carcass apart for easier access. The predator's tugging and pulling at the contents of the stomach brought the gastroliths along too, and they were kept intact in folds of tissue that lay on the bare sand. As the sand later filled in around the soft, probably desiccated tissues, the gastroliths were held in their original positions with respect to each other.

A second compelling piece of evidence supporting the stomach-stone hypothesis was that some gastroliths were in direct contact with bone: ribs and vertebrae. Rather than lying horizontally, some had an amazing orientation. One stone, a flattened disk that I might have chosen as a "skipper" to throw across the surface of a pond, was perfectly on edge, in a vertical position. Others were similarly oriented. Stream action could not have produced such startling positions for stream-deposited stones.

The largest and smallest gastroliths. The largest gastrolith is markedly bigger than all the others. Could it have caused Sam's death by choking? (Scale divisions are 1 centimeter.)

Third, measurements of these stones revealed another contraindication of stream deposits. The statistical distribution of long-axis and intermediate-axis dimensions of a natural population of river stones deposited by a surge in stream velocity would be rather linear: many small ones, fewer intermediate ones and fewer still of the very largest. The sizes of Sam's stones, however, displayed a bell curve. A few were small, the size of a peach pit. More were larger, with the highest frequency being stones roughly the size of a plum. A few were somewhat larger, the size of a small apple, and a very few were larger still--the biggest being the size of a small grapefruit. Only a very unusual stream could have deposited them.

We thus were left with only one reasonable conclusion: these stones were indeed gastroliths, the indestructible dinosaur grit contained in Sam's digestive tract at the time of death. This provides a suitable standard for comparison with purported stomach stones from other sites, even where the gastroliths are not in direct association with a skeleton or their sedimentary context so easily discerned.

Satisfied that the sedimentary context is consistent with the conclusion that the exotic stones associated with Sam's skeleton are genuine gastroliths, we can deduce a great deal of information about their function, their duration in the digestive tract, the anatomy of the digestive tract, Sam's behavior in selecting suitable stones for ingestion, and the history of the carcass between coming to rest on the sand bar and its ultimate burial. (This last topic plays an important role in the chapter on forensics. Here I will concentrate on life-associated issues.)

Gastrolith texture. Several gastroliths here demonstrate the extreme rounding and surface polishing that they undergo in the digestive system. (Scale divisions are 1 centimeter.)

More gastrolith texture. Here one of the gastroliths shows an extreme development of rounding and polish. Its composition is cryptocrystalline quartz.

Sam's gastroliths are all rounded, and some are highly polished. Their surface texture ranges from dull to waxy, in accordance with the degree of polishing. Except for breakage during excavation, none of the gastroliths have sharp edges; even the ones with irregular shapes are highly rounded. Their color ranges from white to black, with a rainbow of intermediate colors including shades of red, yellow, green, and brown. Their mineral composition is less variable: all of the gastroliths in this set originated as igneous or metamorphic rocks. All of the gastroliths, moreover, are varieties of quartz (chert and quartzite)--the hardest and most durable of the minerals commonly occurring at the earth's surface. These facts are all important in deciphering the function of the gastroliths and related aspects of Sam's behavior.

The assumption that Sam acquired these stones from local river beds is implicit here. That is the only reasonable source of exotic cobbles in the Morrison habitat. These cobbles would have eroded out of rock formations from localities more ancient than the Ojito site. Perhaps some were repeatedly subjected to the sedimentary cycle of erosion and deposition, with each rejuvenation of the landscape.

Sam most likely acquired these stones deliberately, while drinking at a stream bank. On the other hand, it is possible that sauropods migrated to source areas for particular rock compositions, an idea that I find highly unlikely but which I cannot dismiss altogether. Several paleontologists have suggested that identification of these source areas, called their "provenance," might indicate the extent of migratory movements among the dinosaurs.

According to some paleontologists, the quality of polish on a gastrolith is a distinguishing feature that separates gastroliths from stream-deposited stones. Therefore all gastroliths should have a high polish, and ones that were in residence for a long time should be waxy. The waxy trait seems to be a consequence of polish, on both the high and low surfaces of the rock. Rocks with such a texture reflect more light than rocks that have polish on the highs only, and they impart a striking quality in reflected light. They feel waxy, too, in contrast to the pitted and slightly roughened texture of rocks that lack this trait.

Not all of Sam's gastroliths, however, display a waxy surface. To the unaided eye, they range from highly polished and waxy to dull and pitted. The waxy ones are easy to distinguish from ordinary river stones, but the dull gastroliths seem not to have lost the surface texture from their original condition when Sam picked them up as river stones in an ancient stream bed.

The origin of this polish, especially the waxy texture, is difficult to explain. If these stones were used for grinding food, as generally claimed, then they should all be pitted and scratched as a consequence of grinding against each other. If the polish came from chemicals in the digestive tract--say, acids in the stomach--they should all have the same degree of polish, or at least all rocks of similar mineral composition should display similar textures.

Tumblers used by rock hounds to polish their stones are not a fair analogue for a dinosaur stomach. Stones polished in a tumbler do not polish each other; instead, the tumbler is furnished with a supply of softer polishing materials (usually culminating with talc), that impart a lustrous surface as the stones roll and tumble. Dinosaurs did not possess such carefully managed polishing materials. I surmise that these stones lay in folds and creases in the digestive tract, particularly in the gizzard (and crop), and that their polish came from the gentle action of muscular contractions.

I cannot explain why some are highly polished and others are dull. There seems to be no correlation with mineral composition, size, degree of rounding, or map position in the quarry. But I can guess, I suppose that the rocks in the specialized chambers of the digestive tract were ingested at different times, as with my cockatiels, which continually replenish their supply of grit. If so, stones in longest residence would be more polished than ones in residence for shorter times. Also, some may have been subject to polishing action more intensely than others just because of position in the digestive tract. One fold of the gizzard may have been more active in its crushing function than another fold, so that some stones in the gizzard were subjected to more of the polishing action than others.

In scientific terms we might express these observations on degree of polish as "a tendency toward high polish and eventually a waxy texture," but one must first exclude the possibility that the tendency is reversed: a waxy texture originally at the time of ingestion; rough and pitted after a long time in residence. Several experiments have documented the high polish of gastroliths; the researchers conclude that the ones with high polish, and especially the ones with waxy texture, are more polished than river stones or beach stones.

Sam's gastroliths are all rounded, a condition more striking and more universal than their polish. None are angular; none have sharp edges. In technical descriptions these stones would be described as well rounded, an extreme condition of rounding that occurs naturally in the rolling motion of rocks and cobbles in stream or shoreline settings. The mechanical action of the rolling and sliding of rocks that are originally angular (say, cubic for discussion here) wears on the corners and edges, gradually changing the shape from cubic to a cube with rounded edges and corners. Eventually the original cube approaches a spherical shape. Because of natural fractures in the original rocks from which they are derived, rocks rounded by the tumbling and sliding in rivers or on beaches tend to be more like squashed spheres than spheres and discoid, like skippers or poker chips.

Of course in nature rocks never become perfectly spherical or even flattened round by this process, but they approach that extreme: this idea is expressed as "sphericity." The sphericity of Sam's gastroliths is uniformly high. Compared with a selection of rounded stream gravels, even those preselected for roundness, Sam's gastroliths are remarkable for their rounding.

Were Sam's gastroliths rounded by rolling and sliding in the digestive tract? Yes, but they were almost surely rounded to begin with, when Sam picked them up, probably from a nearby stream bed. Then too, when feeling the need for more stones, Sam likely exercised some degree of choice in picking them up. In Sam's digestive tract, their sphericity increased as a consequence of renewed mechanical abrasion. They were simultaneously polished and rounded. Now, all these deductions seem reasonable and they fit our expectations. But a pesky problem remains to be addressed: Why are there no angular stones?

If Sam only occasionally and randomly picked up stones, then some should have been broken and angular, reflecting the breakage and angularity of stones in stream beds. In a sample of 240 gastroliths we should expect to find a dozen or more angular stones, with sharp edges not yet rounded by the action of the digestive tract, stones taken up just before death that hadn't been subjected to mechanical breakdown long enough to reduce the sharp corners and edges. None of Sam's gastroliths have sharp, or even dull, edges or corners; all 240 are rounded. I take this collective trait to indicate that all of these gastroliths were in residence in the digestive tract for a long time, perhaps years. If so, by my argument, they stayed in the digestive tract until they were almost totally ground down, say to the size of peach pits, and passed through the digestive system with food.

Even if Sam picked up river stones in quantity, we should see a spectrum of rounding. A set collected eight months before Sam died should have more rounding (and polish) than a set picked up only a month before death; no such pattern of subsets of rounding and polish emerge from our observations. More likely, Sam picked up river stones casually and irregularly.

Considering that gastroliths are poorly documented, they have been used disproportionately in deductive arguments concerning sauropod behavior, feeding habits, and stratigraphy. All of these topics have a bearing on our studies of Sam. For example, was Sam selective? Were some river stones more attractive than others? Certainly a sauropod would select a stone within some size range (too large and it couldn't be swallowed; too small and it would pass through to no effect). Possibly a sauropod would select stones of some minimum level of sphericity. These choices could have been made by sight or by simply spitting out those that felt wrong in the mouth. But some paleontologists have hypothesized that sauropods selected for composition or color too--suggestions that, in my view, move out of the realm of science and into fantasy.

Some popular accounts of sauropods include the suggestion that these dinosaurs undertook long migrations to favorite collecting sites. There is no evidence for this proposal. I see no way to distinguish between river stones picked up in nearby drainages (but derived from sources perhaps hundreds of miles away) and stones collected from afar and carried long distances in the body cavity of a giant dinosaur. The well-intended purpose in these claims is to test the idea that sauropods engaged in long migrations; that may have been true, but gastroliths will not settle the question.

Similarly, some have suggested that sauropods selected only quartz-rich rocks for gastroliths. Again, I disagree. The fact that Sam's gastroliths (the only sauropod gastroliths thoroughly documented) are all quartz does not lead to that conclusion; instead, these may have been the only stones to survive the vigorous action of the digestive tract. Layered rocks would be easily disaggregated because of differential susceptibility of individual layers to pressure and gastric juices; rocks of other composition, such as limestones or shales, would be easily broken down in the acidic environment of the digestive tract. Such stones would not survive for long as gastroliths.

Some paleontologists have relied on moas for making projections of sauropod feeding habits and the functions of gastroliths. Moas, the huge flightless birds of New Zealand that became extinct owing to human occupation of the islands, used gastroliths extensively, and they seem to have been highly selective. These giant herbivores, weighing as much as a half ton, may have selected only white stones; perhaps only white stones were available or were the only available rocks sufficiently durable to withstand the rigors of the digestive tract. Nevertheless, I question the direct application of moa behavior to that of sauropods.

Similarly, my notions on residence time in the alimentary tract (the stones remained until virtually ground down and destroyed) have a bearing on interpretations of purported gastroliths in the Morrison and Cedar Mountain formations that are not positioned directly in association with skeletons. According to geologist Lee Stokes gastroliths occur in great abundance in the Cedar Mountain Formation (lower Cretaceous) of the western United States. This formation overlies the Morrison Formation in much of Utah and Colorado, and the two are difficult to distinguish except by subtle differences. One distinction, according to this argument, is the general lack of gastroliths in the Morrison. I agree, except that I am not convinced that the purported gastroliths are properly identified.

Whatever their origin, the widespread occurrence of these "gastroliths" in the Cedar Mountain Formation is problematic, because sauropod dinosaurs (and all other dinosaurs) are rare in that formation, whereas gastroliths are rare in the Morrison Formation, which contains the world's greatest bounty of sauropods. Paleontologist Robert Bakker has suggested that isolated gastroliths may be the only residual evidence of the positions of skeletons long since dissolved or redeposited, an idea I find attractive but hard to prove. If it can be shown that the conditions for preservation of bone were less favorable in the Cedar Mountain Formation, this idea would become more plausible and deserving of further consideration.

Sam's gastroliths have been the subject of several presentations and papers, but full description of all the gastroliths is probably years away, as the ones still encased along with bone await laboratory preparation and study.

To summarize the facts about Sam's gastroliths: they are all polished, but not all to the same extent, and some are waxy. All are well rounded; none have sharp edges. Some are markedly disk-shaped. All originated from igneous or nonfoliated (unlayered) metamorphic rocks; none are sedimentary or layered; and they range from white to black, including a rainbow of colors. Three-fourths of the gastroliths are between one and three inches at their longest dimension. Long/intermediate/ short axis dimensions of the smallest are .83 inches x .67 inches x .51 inches; and the largest is 3.74 inches x 3.58 inches x 2.87 inches. Large (greater than 3.10 inches) and small (less than 1.20 inches) gastroliths are uncommon.

Not included in the summary above are observations concerning the positions of the gastroliths in the quarry. These facts are just as important as the physical qualities of the objects themselves. Contrary to expectations, Sam's gastroliths were not found in a single pile, as though dumped from a bushel basket and immediately buried. Instead, Sam's gastroliths were spread unevenly over more than 1,600 square feet. We may have missed some in the excavation, especially in the early years before we fully appreciated their implications, but the patterns of their distribution seem clear. In Sam's carcass were two clusters of gastroliths, one large set in the forward region of the digestive tract and a smaller set in the rear region found within the pelvic bones. Only the rear set, consisting of 26 tightly clustered stones, can be attributed to a gizzard, the grinding chamber originating as a specialized rear pocket of the stomach.

The set from the front of the body seems to be centered at the front part of the chest cavity, near the base of the neck. Evidently the neck and head were displaced after the gastroliths spilled out of the body cavity, leaving the gastroliths to mark the temporary position of the front part of the body. Between the front set and the tight cluster in the pelvis was a barren region of at least four feet along the vertebral column where no gastroliths were found. This separation is some evidence of a specialized forward chamber in Sam's digestive system, which I have called the crop in reference to similar anatomy in grain-feeding birds such as chickens and cockatiels.

There are no discernible differences in roundness, polish, size, or composition between the front set (from the crop) and the rear set (from the gizzard). This fact is inconsistent with the usual function of the crop as a storage chamber, where grains and plant matter collect in lumps before passing to the stomach and gizzard for digestion. The inconsistency lies not in the anatomical position, but in the fact that the gastroliths are rounded and polished, and many are waxy. If the crop were simply a storage chamber, then gastroliths should have no function there. The fact that they are highly polished and rounded, like those from the gizzard, indicates that the front gastroliths were equally involved in processing food, a radical suggestion for the function of a crop. One highly unusual bird, the primitive hoatzin of South America, has a crop with a gizzard's function. Although the anatomy of one bird is not definitive evidence that Sam had a crop, it does demonstrate that a crop with a grinding function is possible.

The sauropod digestive tract. This schematic includes the crop (the forward chamber at the base of the neck) and gizzard (between the stomach and small intestine).

Overall, we have sufficient evidence to propose that Sam's digestive system resembled that of grain-feeding birds: esophagus, crop, stomach, gizzard, and intestine. The crop and gizzard both contained gastroliths, where food was probably pulverized in preparation for chemical digestion (gastric secretion in the stomach and absorption in the intestine). Yet, a nagging question remains: how exactly did Sam (and other sauropods) use the gastroliths in digestion?

According to conventional explanations sauropod gastroliths were a substitute for strong teeth. They were used in a grinding function in place of chewing. I disagree, for several reasons.

First, not all sauropod skeletons seem to have had gastroliths, but all sauropods had decidedly weak dentition. I doubt that stones as a substitute for teeth were universally necessary for mastication. In modern animals dentition is suited to diet. Only birds, which lack teeth, have been documented to require grit as an accessory material for grinding food--and only seed-eaters among the birds.

Second, gastroliths must have remained in the lowest creases and folds in the capacious digestive organs, whereas food would have occupied the chambers from bottom to top; only foodstuffs at the bottom of the crop or gizzard would have been subjected to a grinding action of the stones.

Third, the bulk of the gastroliths is surprisingly out of proportion to expectations: the 240 from Sam's carcass scarcely fill a ten-quart bucket. To be important as an accessory grinding device, the bulk of gastroliths should be considerably greater, or the foodstuffs must have been forced through narrow constrictions blocked by gastroliths through which the food had to pass before moving to the next chamber of the digestive tract. For an animal weighing five or six times as much as an elephant and processing many times as much food every day, the gastroliths seem a weak addition to the digestive tract.

Instead, I propose that the gastroliths stirred digestive juices as they rolled and tumbled in the bottoms of their respective chambers, much like a magnetic stirrer in a chemistry laboratory. A small stirring device can create sufficient turbulence to circulate fluids throughout a beaker, preventing unwanted settling and segregation of materials by density. By this argument, gastroliths had a similar role: to keep foodstuffs and the digestive juices mixed, not to grind the food or pulverize it. I subscribe to this notion especially for the crop, which was probably distensible and not highly muscular. Fermentation and putrefaction in the crop might be controlled by adequate mixing of digestive juices, preparing food for more vigorous chemical treatment in the stomach. A grinding function for the gastroliths in the gizzard seems more reasonable, where muscular contractions alone would keep the food moving and prevent stagnation.

In both places, the rounding and polish of gastroliths would arise from their contact with each other, the lining of the digestive tract, and with gastric juices, not with the food materials; thus the grinding (rounding and polishing) of gastroliths is a consequence of their tumbling and rolling with each other, with only an inconsequential grinding of food.

Interpretation of Sam's gastroliths has been a fascinating part of the excavation of Sam. The gastroliths may tell us more about how Sam lived, and whom among the living the great sauropods most resemble from a food processing standpoint. But might the gastroliths also tell us how Sam died?

I off-handedly remarked to a reporter that Sam's largest gastrolith, the size of a grapefruit, was considerably larger than all the others--and maybe it caused Sam's demise by lodging in the esophagus or trachea. This tongue-in-cheek (or, stone-in-mouth) remark received considerable publicity. Did Sam choke on this giant gastrolith? Maybe. The evidence is far from conclusive, but the possibility cannot be ruled out. There is, however, evidence suggesting another cause of death: Sam may have been killed by a predator.