The Egyptian Drill: A Unique Dual-Mode Device

37

ER

TUOAYDFK

GHJLZICVN

Ethnoarchaeology, Vol. 4, No. 1 (Spring 2012), pp. 37–52.Copyright © 2012 Left Coast Press, Inc. All rights reserved.

The Egyptian DrillA Unique Dual-Mode Device

Stephen C. Saraydar, Department of Anthropology, State Univer-sity of New York College, Oswego, NY 13126 ([email protected])

Abstract An Egyptian drilling tool used primarily for the production of alabaster, calcite, and limestone vases and first known from the early Third Dynasty has been the subject of speculation and debate among Egyptolo-gists and students of ancient technology for more than one hundred years. Researchers once thought that this tool, which was used with both stone and tubular metal drill bits, represented the earliest crank-driven device, however, that idea has been abandoned in favor of other possibilities. An intriguing argument centering on the ability of the drill to translate linear input into rapid, unidirectional rotary output was advanced by Hartenberg and Schmidt (1969) but never tested with a replica made from authentic ma-terials. Recent experiments conducted by Stocks (2003) using replicas made from materials available in ancient Egypt, led to a rejection of that claim and demonstrated slow drilling and boring of stone by means of alternating twists of the tool in clockwise and counterclockwise directions. The pictorial evidence left by the ancient Egyptians and the persuasiveness of the analysis by Hartenberg and Schmidt of the drill’s geometry and operational dynamics pointed to the need for further investigation. Accordingly, I performed an experiment in which a wood and flint replica of the tool was used success-fully to drill blocks of alabaster and limestone. I conclude that the evidence partially supports Hartenberg’s and Schmidt’s theory that the drill repre-sents a unique technological achievement by virtue of its dual-mode capabil-ity and ability to produce crank motion more than two millennia before the earliest true crank tool.

38 Stephen C. Saraydar

An intriguing and controversial Egyptian drilling/boring device dat-ing in its earliest form to the Third Dynasty (2686–2613 BC), and known only from tomb paintings, hieroglyphs, and surviving stone weights and drill bits (see Breasted 1919: 571–572; Childe 1954: 193; Hester and Heizer 1981:28, 30, and plates XXII and XXIII; Stocks 2003:139), has provided a particularly instructive example of how we may move from early conjectures to more carefully reasoned hypotheses, and from these to experiments with replicas of varying fidelity to ancient designs (see Figures 1–4). While there has been general agreement that this tool was used primarily for hollowing out alabaster, calcite, and limestone vessels, a variety of conflicting claims have been made concerning its mode of operation. Until recently it had generally been assumed that the drill accom-plished its work by being rotated by one hand while the other held its shaft in a vertical position. A significant variation on this theme has been successfully demonstrated by Stocks (1993, 2003), in which bidirectional rotation (a partial turn clockwise followed by a partial turn in the counterclockwise direction) was used in the replication of Egyptian stone vases. But is there another possibility?

Figure 1. Depiction of standing operation of drill from tomb at Saqqara, Late Old Kingdom (adapted from Childe 1954:193). Tomb was not identified.

The Egyptian Drill 39

Figure 2. Depiction of seated operation of drills from a tomb relief (from Breasted 1919: 572). No provenience provided but most likely Saqqara, Late Old Kingdom.

Figure 3. Hieroglyphic representations of drill (adapted from Hartenberg & Schmidt 1969: 157).

(a) (b)Figure 4. (a) Typical crescent-shaped flint bit (from Childe 1954:193). Reprinted with the permis-sion of Oxford University Press; (b) Worn limestone figure-of-eight “vase borer” from Hierakonpo-lis. (adapted from Quibell and Green 1902, plate LXII, image 6). Length approximately 10 cm.

40 Stephen C. Saraydar

A novel claim made by Hartenberg and Schmidt (1969) that this drill was ca-pable of high speed, unidirectional rotation produced by a simple linear (back and forth) motion of the operator’s arms has been largely ignored and expressly rejected by Stocks. The goal of my experiment was to determine if a replica of the device using a crescent-shaped flint bit could be made to drill stone effectively in the manner described by Hartenberg and Schmidt. If so, its ability to transform linear motion into rotary motion would make it unique for its era.

A Brief History of Interpretations and ExperimentsIt seemed evident to Egyptologists such as Ludwig Borchardt (1897: 107) in the late nineteenth century that that the device was a crank-driven drill that required two-handed operation:

Die Arbeit mit diesem Bohrer muss übrigens recht schwer gewesen sein, da wir auf den Darstellungen sehen, dass die Kurbel meist mit beiden Händen gepackt werden musste. Translation: The work with this drill, incidentally, must have been fairly difficult, because we can see from the illustrations that the crank had to be gripped by both hands.

George Reisner (1908: 134), writing in the early twentieth century, shared Borchardt’s opinion, referring to the Egyptian drill as a “crank and shaft borer” and later describing it as a “boring stone fixed in a forked shaft weighted at the top and turned by a crank . . . .” (1931: 179). James Breasted’s (1919: 571–572) in-terpretation of this early Egyptian implement was in agreement with Borchardt’s and Reisner’s, and he went so far as to attribute the invention of the crank to the ancient Egyptians. 1

In the view of historian of technology Lynn White, Jr. (1968:116), “Next to the wheel, the crank is probably the most important single element in machine design”, and a far more complex one in terms of its kinematics than most people today would imagine (see also White 1962: 103–104, 115; 1968: 117–119). In its purest form, the crank may be defined as a device in which continuous rotary motion is produced via a handle that turns a shaft held firmly in position (typi-cally vertically or horizontally) by one or more bearings. The shaft will normally have an arm attached to it at a right angle, with the handle attached to the arm, also at a right angle. For example, a mechanical pencil sharpener and a windlass for obtaining water from a well both have a handle attached to an arm that is

The Egyptian Drill 41

continuously rotated through 360 degrees to turn the shaft that either sharpens the pencil or winds/unwinds a rope attached to a bucket to raise or lower it.

In the late 1960s the earlier assertions regarding the tool’s method of opera-tion as a crank came to the attention of mechanical engineer Richard Hartenberg, who found them unconvincing. On the basis of its configuration as depicted in tomb paintings, he concluded that the tool could not have been crank-driven:

If such a one-piece drill is assumed to have been cranked to turn about its vertical axis, an upper steady bearing at C [Figure 5] would have been necessary to support the vertical part; but illustrations . . . show the one hand busy with the weight. The picture has an uncomfortable look because of the high position of the weights, balanced on a point at OA and supported from the off-axis handle at A. The drill is precariously balanced and, unless restrained by a hand on the weight, would tend to rotate about the axis OAA. A cranking motion imposed on a top-heavy system whose axis of rotation is supported at only one point is unthinkable. (Hartenberg and Schmidt 1969: 159)

Figure 5. Drill rotational axis. (adapted from Hartenberg & Schmidt 1969: 156).

42 Stephen C. Saraydar

To test what he had inferred from the pictorial evidence, Hartenberg enlisted the assistance of instrument maker John Schmidt, Jr. to create a replica of sorts, one that did not use authentic materials but did mirror the overall geometry and proportions of the device as depicted in Figure 2. This replica was constructed from electrical conduit (bent to the appropriate shape), a 5/8-inch twist drill bit, and two bricks that were attached to the shaft with a brazing rod and a wooden clamp. The result, in brief, was that efforts to crank the device failed. However, as attempts were made to maintain a vertical orientation while struggling with the “wobbling course of the bricks” (Hartenberg and Schmidt 1969: 161), they discovered that the drill would rotate rapidly when both hands were used to push and pull the device in a straight line. This linear, back and forth movement by the operator was easy to sustain and resulted in a block of wood being drilled rapidly with minimum effort. As to the rotation of the drill:

. . . [it] rotated about the axis OAA between bit and hand because of the drop and rise of the eccentric center of gravity . . . after the center of gravity reached its lowest position, it was raised by pulling the handle straight back at the proper instant as the weight swung around. The rotational inertia of the weight helped swing the center of gravity beyond its lowest and highest positions. The hands exerted no cranking action, only a push-pull. (1969: 161)

These results led Hartenberg and Schmidt (1968:164–165) to conclude that “the ancient device is still with us, at least in principle” in the form of a carpenter’s brace (Figure 6). A carpenter’s brace (bit-brace), with its four right angle bends, can be cranked using two hands (one on the top to hold the device vertical and the other to rotate the shaft). However, it is equally useful for producing much more rapid rotation, simply by moving the handle at its upper end back and forth in a one-handed, push-pull operation (the handle provides an upper bear-ing for the shaft and does not rotate with it). When used in the latter mode, the operator’s hand clearly does not execute a cranking motion. It is worth not-ing that prior to Hartenberg’s and Schmidt’s experiment, Needham had offered the suggestion that the Egyptian device may have been a “drill of the primitive brace-and-bit type” (1965:114). Hester and Heizer (1981: 39) have documented the use of simple braces fitted with variously shaped and sized metal bits in small workshops in contemporary Egypt to hollow out the bodies of replicas of an-cient alabaster vessels. This process uses the brace in crank-mode rather than the push-pull mode that intrigued Hartenberg and Schmidt.

The Egyptian Drill 43

Understanding of the drill has been advanced considerably in recent years thanks to the exhaustive experimental studies of ancient Egyptian stonework-ing technology conducted by Denys Stocks (1993, 2001, 2003). Stocks used wooden replicas of the drill variously equipped with tubular copper borers, flint and chert crescent drill bits, and ground-stone borers in the form of an elongated disc with notches at the midpoint to accommodate a forked drive shaft (hence the term “figure-of-eight” borer. See Figure 4b). With these, he has succeeded in producing stone vessels that mirror ancient examples in all key attributes, includ-ing characteristic striations on the interior surfaces resulting from rotation of the borer. He has labeled his technique of boring the “twist/reverse twist” method (a turn of the device 90 degrees clockwise followed by

a 90 degree turn counterclockwise). Of particular relevance here is that in his ex-periments Stocks revisited the work of Hartenberg and Schmidt. He comments:

They concluded that the tool rotated in one direction, and that the handle was used as some form of crank.2 Their tests were carried out on a bent tube weighted with two house bricks. However, the present tests, on tools reconstructed from materials in use by ancient craftworkers, demonstrate that not only does a continuous rotary action cause the drill to wobble alarmingly, but is difficult for a human to perform and, indeed, to control. The stone weights fly outwards and increase the wobbling action. Such use of the tool must cause serious damage to the vessel, not to mention the extreme tapering of the cores and the hole when in use with a tubular drill. This is at variance with the archaeological evidence for parallel-sided cores and holes in ancient vessels. This proposed use of the tool must be firmly rejected. (Stocks 2003: 148)

Making and Testing a ReplicaSeveral years after Hartenberg and Schmidt published their article, I had the op-portunity to test a near duplicate of their drill. I was greatly impressed with how

Figure 6. Carpenter’s brace (bit-brace).

44 Stephen C. Saraydar

efficiently linear (back and forth) motion was translated into rapid rotation and speedy drilling.3 While a stubborn adherence to a favored hypothesis may or may not be a useful trait for those engaged in scientific pursuits (Mitroff 1981), my hands-on experience with the replica of Hartenberg’s and Schmidt’s drill had made a sufficiently strong impression on me, as had their analysis of the device’s rotational dynamics, that I was unable to accept Stock’s total dismissal of their results without a test of my own. In addition, I felt that the differences in the pic-torial evidence left by the ancient Egyptians were significant and pointed to two possible modes of operation. The seated craftsmen in Figure 2 could be operat-ing the device in the two-handed manner described and successfully tested by Stocks. In contrast, the operator in Figure 1 has one hand on the handle and the other appears to be giving the weight a push to initiate rotation. Further, if the central body of the drill is accurately depicted, the central section of the shaft in Figure 1 is too large in diameter to be grasped in the manner shown in Figure 2, where there is ample room below the two small weights near the top of the drill for the craftsman on the left to hold the much thinner central shaft in a vertical position. I must make it clear that I do not dispute Stock’s findings concerning the manufacture of vessels by slow drilling and boring. The goal of my experi-ment was simply to determine whether or not a replica made of wood and stone rather than metal and bricks would operate effectively as a drill in the manner described by Hartenberg and Schmidt.

To make my drill, I fitted a crescent-shaped flint bit (measuring 48 mm at its widest point) modeled on the example in Figure 4a to a slot cut in one end of a 27 cm. long, 20 mm diameter section of a small ash tree stem (Figure 7). I then used a drawknife to make flat surfaces near the other end of this piece and the lower portion of a 108.5 cm long, 25 mm diameter section of another ash tree stem. The latter piece was forked at one end and was chosen to serve as the upper shaft and handle of the drill. The two pieces were lashed together with cord to form a simple scarf joint. The total length of the completed drill (measured from the bit to the end of the handle) was 96.5 cm. I loosely suspended two smoothly rounded stones obtained from the shore of Lake Ontario in cloth bags from a stub near the top of the upper body of the drill; the weights were not restrained in any other way (Figure 8). I obtained a block of alabaster (the fine-grained, massive form of gypsum) from Colorado which, while not identical to Egyptian alabaster, provided an appropriate material to be drilled.

The Egyptian Drill 45

Figure 7. Lower drill shaft fitted with flint bit.

Figure 8. Replica drill fitted with two 1-kg weights and crescent flint bit, drilling in alabaster.

46 Stephen C. Saraydar

With the drill in place on the alabaster, a small rotation of the handle was sufficient to start the operation; a push on one of the weights worked equally well. It was no surprise that as the drill began to spin the resulting centrif-ugal force caused the weights to fly outward from their resting po-sitions (Figure 9). However, con-trary to Stocks’ experience with a replica quite similar to mine, I did not find this to be a problem. The drill was easy to control and keep spinning, inducing little fatigue so that I was able to maintain consis-tent arm movement; the rotation-al speed that felt most comfort-able was approximately 80 rpm, although slower rotational speeds were easily managed and could be quite useful for more delicate operations.

Contrary to the results reported by Hartenberg and Schmidt, whose replica featured brick weights clamped in place on the drill’s shaft, a purely push-pull arm action was not sufficient to sustain rotation. I quickly found that I had to supplement the push-pull by small movements of my arm from side to side to compensate for the freely swinging weights. This resulted in my hand tracing out a 360 degree circle as it also moved forward and back. These results indicate that the drill comes considerably closer to being a crank tool than Hartenberg and Schmidt believed. Another way in which my experience differed from that of Hartenberg and Schmidt is that I operated the drill with only one hand; I found using two hands awkward and ultimately unworkable. I found it helpful to wear a leather glove for protection against abrasion from the rotating handle. Alter-nately, a sheath of some kind placed loosely around the drill’s handle would serve the same purpose.4

Figure 9. Drill in motion, showing weights flying outward.

The Egyptian Drill 47

The block of alabaster was neatly drilled. The best results came from starting the operation by rotating the drill slowly from the end of its handle to establish a circular depression on the block’s surface. Throughout the drilling I kept the stone wet and regularly poured water on it to wash away the thick paste that developed (ancient examples of similar flint bits caked with gypsum were found in Old Kingdom workshops in the Fayum by Caton-Thompson and Gardner 1934:105). Drilling did not become more difficult once the bit was entirely below the surface of the stone.

Fitted with weights totaling 1.4 kg, a depth of 24 mm was reached in 30 minutes (including time spent rewetting the alabaster); the volume of the stone removed was 32 cm3. A second trial was conducted with heavier weights, total-ing 2.0 kg; 30 minutes of drilling resulted in a depth of 30 mm; the volume of stone removed was 39 cm3 (Figure 10). This trial concluded after an additional 30 minutes of drilling, which increased the depth of the bore to 55 mm and its vol-ume to 85 cm3 (Figure 11). Clearly, heavier weights (perhaps necessitating a more robust drill bit) would further increase the rate at which stone was removed. In

Figure 10. Drilled alabaster block. Depth of bore on right 24 mm, depth of bore on left 30 mm.

48 Stephen C. Saraydar

both cases the portion of the bore through which the bit had fully passed was essentially cylindrical (sides parallel).

In the first trial, the diameter of the bore at the surface of the stone varied between 49 and 50 mm; in the second, the range was slightly greater, ranging between 50 and 53 mm. By limiting the stroke, that is the length of back-and-forth movement of the hand working the drill, a more nearly circular bore will be produced, with the added benefit of being gentler on the stone being drilled (a long stroke, and hence a larger rise and fall of the weights, is not required to make the device work). Shifting the operator’s position relative to the stone be-ing drilled by 90 degrees part way through the operation might also tend to limit such variations (not tested). In contrast, the replica created by Hartenberg and Schmidt relies on a longer stroke and is more violent in its action, with poten-tially damaging consequences when drilling a thin-walled vessel.

Furthermore, at the end of the drilling, the flint bit showed no significant signs of wear or damage. Given alabaster’s hardness (2 on the Mohs scale), I would not expect it to cause rapid wear on a flint bit (hardness of 7). It should

Figure 11. Alabaster drilled to a depth of 55 mm, showing striations on interior surface.

The Egyptian Drill 49

be noted that Hester and Heizer (1981: 33) cite the absence of “arduous wear” on the crescentic flint bits they examined to argue against their use in vase bor-ing. However, they acknowledge the strong evidence to the contrary reported by Caton-Thompson and Gardner noted above and suggest that further use-wear studies are warranted. More recently, Stocks (2003: 139) has observed that the depth, width, and direction of striations in a broken gypsum vessel in the Robert H. Lowie Museum in Berkeley, California correspond closely to those that would be produced by flint and chert crescents.

Lastly, I had not originally intended to do anything more than determine whether or not the drill would operate in the manner detailed above. However, in light of Stocks’ report that “tests on calcite [hardness of 3] showed that it is too hard effectively to be bored with flint or chert crescents” (2003:139), I decided to experiment a bit further to see if this applied only to drilling using the twist/reverse twist method. In using that technique, Stocks found that “when flint and chert crescents are forced against hard stone vessel walls [of calcite or hard limestone], the scraping action breaks their edges” (2003:140). I was unable to obtain a large enough piece of calcite but did acquire a suitable block of limestone, whose hard-ness I determined to be approximately 4, as it was not scratched by calcite but was scratched by lazurite (hardness of 5 to 5.5). I was able to drill the limestone in the same way as I drilled the alabaster, although progress was, not surprisingly, slower. After 30 minutes of work, the depth of the hole was 20 mm. I did notice that a few very small flakes (approximately a millimeter or two in diameter) were detached from the flint bit, but the edge remained in good condition. Finally, a brief test of the twist/reverse twist method confirmed that it would quickly ruin the edge of the bit, as noted by Stocks, even when working with stone as soft as alabaster.

ConclusionsI have demonstrated that my flint and wood replica of the Egyptian drill is capa-ble of rapid, unidirectional rotation and is effective in drilling at least two types of stone, limestone and alabaster, comparable to those known to have been worked in ancient times. When fitted with a crescent-shaped bit, the drill can be oper-ated with one hand and with minimal effort. With a change to a figure-of-eight bit, this device is also well suited to slow-speed boring by the twist/reverse twist method demonstrated by Stocks.

Less clear is how the drill should be classified as to the type of machine it

50 Stephen C. Saraydar

represents when used in the unidirectional mode. It does not appear to have been designed as a crank. If it had been, it would earn the distinction of being the earliest crank-driven device by over 2,000 years. However, the scene depicted in Figure 1 does not suggest the use of a true crank, as the drill’s weights are placed in a location that would make it extremely difficult at best, and probably impos-sible, to use one hand to rotate the handle continuously through 360 degrees while using the other to maintain the tool in an vertical position, even had the shaft been thinner. Likewise, the workers in Figure 2 are not grasping the upper part of their drills in a way that would make continuous rotary motion possible. The drill is not equivalent to a primitive carpenter’s brace either, despite the simi-larities noted by Hartenberg and Schmidt. The apparent close correspondence to that device can be attributed largely to their highly compromised replica (in particular, the brick weights that were clamped to the drill’s shaft). The forces created by stone weights suspended from a single point dictate a more complex motion of the operator’s arm to make the drill revolve in a controlled manner. If not quite a crank or a carpenter’s brace, then what?

White (1962: 107) makes the important distinction between crank and crank motion. He asks, “If not the crank itself, can we at least find crank motion in Antiquity, apart from China?” In the case of the Egyptian drill, the answer is “yes”. To make it both spin rapidly and drill a circular hole, the operator is forced to move the device in a way that incorporates some of the motions of a crank (360 degree circular movement of the hand) and a bit-brace (push-pull move-ment of the arm). However, there is no evidence to suggest that the drill’s osten-sible potential to lead to the development of true crank-driven devices was ever realized in ancient Egypt. White (1962:110) comments, “The mechanical crank is extraordinary not only for its late invention, or arrival from China, but also for the almost unbelievable delay, once it was known [referring to Europe], in its assimilation to technological thinking.” It is easy for us today to see an obvious step from the Egyptian drill to a true crank tool simply because we are already familiar with crank-driven devices.

Regardless of any debates we might have on how this piece of technology should be classified, the essential simplicity and elegance of its design is appar-ent, as is its effectiveness in producing rapid rotary motion when so desired by its operator. Replicative experiments have shown that the Egyptians should be credited with creating an ingenious and effective device whose operational me-chanics appear to be unlike those of any other implement in antiquity.

The Egyptian Drill 51

Acknowledgments

I thank master flintknapper Ken Wallace for making the flint bit used in this experiment. I also thank Robert Ascher for his very helpful suggestions on improving an early version of this manuscript, John Lalande II for his translation of Borchardt, and Mary Rolland for the drawings. I thank Kathryn Bard for her comments on the provenience and dating of the tomb reliefs in Figures 1 and 2. I am also grateful for the valuable comments and sug-gestions made by Kathryn Arthur and the two anonymous reviewers of my manuscript.

Notes

1. The earliest appearance of a true crank is generally believed to have occurred in Han Dynasty China (206 B.C. to A.D. 220). For further discussion of the crank and its later development see White 1962: 103–116, Lucas 2005:5 note 9, and Ritti, Grewe, and Kessener 2007.

2. I am confident that Hartenberg and Schmidt would deny having reached a conclu-sion that the handle was used as a “form of crank”, as this idea is precisely what they objected to in the first place. Their analogy with the carpenter’s brace went only as far as the technique of producing rapid rotation by rocking the device back and forth; it did not extend to the brace’s unique features that also enable it to be cranked.

3. Hartenberg’s and Schmidt’s experiment has also provided a useful case study for purposes of instruction in the design of replicative experiments, particularly on the effects of compromises in authenticity of materials on the confidence that may be placed in the data they produce and the subsequent limitations on the conclusions that can be drawn from them (Saraydar 2008: 26–28; 37–41).

4. Another possibility is that a hollow horn or wooden tube of some sort fitted over the curved upper part of the drill could serve as a handle that would allow the drill to rotate without rubbing against the hand. This idea is a modification of a suggestion made by V. Gordon Childe who, in puzzling over the drill’s construction, suggested that “the curved piece at the top of the spindle is very likely not part of the spindle at all, but a handle of hollow horn in which the spindle turns freely” (1954:192). Hartenberg and Schmidt (1969: 160) dismissed this two-piece construction, noting that it would not make sense to have a curved handle if the goal was to rotate the spindle in a vertical position. I would add that with Childe’s design the free hand would have to push the weights to turn the spindle, which would be awkward and, if the weights were suspended (as appears to be the case) rather than lashed tightly to the spindle, all but impossible.

References Cited

Borchardt, Ludwig. 1897. Beiträge zu GRIFFITH’ Benihasan III. Zeitschrift für ägyp-tische Sprache und Altertumskunde XXXV. Leipzig: J. C. Hinrichs.

52 Stephen C. Saraydar

Breasted, James H. 1919. The origins of civilization. Scientific Monthly IX:561–577.Caton-Thompson, Gertrude and Elinor W. Gardner. 1934. The desert Fayum. London:

The Royal Anthropological Institute of Great Britain and Ireland.Childe, V. Gordon. 1954. Rotary Motion. In A history of technology, volume 1, eds.

Charles Singer, Eric J. Holmyard, and Alfred R. Hall, 187– 215. London: Oxford University Press.

Hartenberg, Richard S. and John Schmidt, Jr. 1969. The Egyptian drill and the origin of the crank. Technology and Culture 10:155–165.

Hester, Thomas R. and Robert F. Heizer. 1981. Making stone vases: Ethnoarchaeological studies at an alabaster workshop in Upper Egypt. Monographic Publications of the Near East, Occasional Papers 1(2). Malibu, CA: Undena Publications.

Lucas, Adam. R. 2005. Industrial milling in the ancient and Medieval worlds. Technology and Culture 46:1–30.

Mitroff, Ian I. 1981. Scientists and confirmation bias. In On scientific thinking, eds. Ryan D. Tweney, Michael E. Dougherty, and Clifford R. Mynatt, 170–175. New York: Columbia University Press.

Needham, Joseph. 1965. Science and civilization in China, volume IV. Cambridge: Cam-bridge University Press.

Quibell, James E. and Frederick W. Green. 1902. Hierakonpolis, part 2. Egyptian Re-search Account, fifth memoir. London: Bernard Quaritch.

Reisner, George A. 1908. The early dynastic cemeteries of Naga-ed-Dêr. Leipzig: J. C. Hinrichs.

____. 1931. Mycerinus. The temples of the Third Pyramid at Giza. Cambridge, MA: Harvard University Press.

Ritti, Tullia, Klaus Grewe, and Paul Kessener. 2007. A relief of a water-powered stone saw mill on a sarcophagus at Hierapolis and its implications. Journal of Roman Archaeology 20:139–163.

Saraydar, Stephen C. 2008. Replicating the past: The art and science of the archaeological experiment. Long Grove, IL: Waveland Press.

Stocks, Denys A. 1993. Making stone vessels in ancient Mesopotamia and Egypt. Antiq-uity 67:596–603.

____. 2001. Testing ancient Egyptian granite-working methods in Aswan, Upper Egypt. Antiquity 75:89–94.

____. 2003. Experiments in Egyptian archaeology: Stoneworking technology in ancient Egypt. Oxford: Routledge.

White Jr., Lynn. 1962. Medieval technology and social change. Oxford: Oxford University Press.

____. 1968. Dynamo and virgin reconsidered. Cambridge, MA: The MIT Press.