Jumat, 01 Februari 2013
in Arabic-Islamic Civilization
DONALD R. HILL
and
AHMAD Y. al-HASSAN
Water-raising machines
The earliest machine used by man for irrigation and water supply is the shaduf.It is illustrated as early as 2500 B.C.E in Akkadian reliefs and about 2000 B.C.E. in Egypt. It has remained in use until the present day and its application is world-wide, so that it is one of the most successful machines ever invented. Its success is probably due to its simplicity, since it can easily be constructed by the village carpenter using local materials. For fairly low lifts it delivers substantial quantities of water. It consists of a long wooden pole suspended at a fulcrum to a wooden beam supported by columns of wood, stone or brick. At the end of the short arm of the lever is a counterweight made of stone or, in alluvial areas where stone is not available, of clay. The bucket is suspended to the other end by a rope (see Figure 6). The operator lowers the bucket into the water and allows it to fill. It is then raised by the action or the counterweight and its contents are discharged into an irrigation ditch or a head tank.
Fig. 6
The scoop drum or tympanum was probably invented in Egypt in the first half of the third century B.C.E. Two large timber discs were fixed to a wooden axle which had iron pegs protruding from its ends. The pegs were housed in iron bearings supported on two columns. The space between the discs was divided into eight segments by wooden boards. The perimeter was closed by wooden boards, there being a slot in each segment to receive the water. Circular holes were cut around the axle in one face of the drum, one hole to each segment. The whole machine was coated with tar (see Figure 7). As the drum was rotated by a tread-wheel, the water was scooped from the source, entered the compartments when they were at the bottom of their travel and was discharged from them when they approached the top. The water ran into a channel and then into a head tank. The scoop drum is rarely mentioned by Muslim writers in connection with irrigation, and its main use seems to have been in de-watering mines. It is ideally suited for this purpose since it can be operated in a fairly restricted space. It was necessary to use a series of drums: the first raised the water into a tank on a platform, a second wheel raised it from this tank to a second tank and so on, until the water was discharged into a drain at the head of the mine.
Fig 7
The screw or water-snail was probably invented by Archimedes (c. 287-212 B.C.E.) when he was living in Egypt and it is therefore appropriate that the machine is often called the ‘Archimedean screw’. A wooden blade is fitted spirally to a long cylindrical wooden rotor. A wooden case is made to fit around the blade, constructed like a barrel, the planks painted with pitch and bound with iron hoops. The rotor is supplied with iron spigots which rotate in iron journals. The screw is set at an angle with one of its ends in the water and as it is rotated the water flows along the helix and discharges from the other end. The smaller the angle to the horizontal, the greater will be the rate of discharge. We do not know precisely how the machine was turned in earlier times – it may have been by a tread-wheel, the power transmitted through a pair of gears. Nowadays it is usually operated by a crank, but the crank is not known to have been in use before the sixth/twelfth century. The screw was in common use throughout the Muslim world until quite recently, but now seems to be becoming rarer (see Figure 8).
Fig 8
The word saqiya is used here to denote the chain-of-pots driven through a pair of gear-wheels by one or two animals harnessed to a draw-bar and walking around a circular track. This very important machine was invented in Egypt, probably about 200 B.C.E., but did not come into widespread use until the fourth or fifth century C.E., with the introduction of the pawl mechanism and earthenware pots. Although it is fairly easy to explain the operation of the machine, it should be emphasized that its construction is quite complex, consisting as it does of over 200 separate components. Only the basic constructional details will be given here. The draw-bar to which the animal is harnessed passes through a hole in an upright shaft to which the horizontal gear -wheel is fixed by spokes. The shaft rotates in a thrust bearing at ground level and another bearing above the gear -wheel located in a cross-beam which is supported on plinths. The gear-wheel is a lantern-pinion, i.e. two large wooden discs held apart by equally spaced pegs. The vertical gear- wheel carries the chain-of-pots and is often called the potgarland wheel. It is supported centrally over the well or other source of water on a wooden axle. On one side of it are the pegs that enter the spaces between the pegs of the lantern-pinion and these pegs pass through to the other side of the wheel, where they carry the chain-of-pots. This consists of two continuous loops of rope between which the earthenware pots are attached – sometimes chains and metal containers are used (Figure 9). In order to prevent the wheel from going into reverse, the machine is provided with a pawl mechanism, which acts on the cogs of the potgarland wheel. This mechanism is essential, because the draught animal is subjected to a constant pull both when moving and when standing still. The pawl is activated in two cases – when the animal is to be unharnessed and in the event of the harness or traces breaking. Without the pawl the machine would turn backwards at great speed and, after one revolution, the drawbar would hit the animal on the head. At the same time, many of the pins of the lantern - pinion would break and the pots smash.
Fig. 9 – A saqiya at Ma’arrat al-Nu’man near Aleppo, Syria
As the animal walks in a circular path, the lantern-pinion is turned and this rotates the potgarland wheel. The pots dip into the water in continuous succession and discharge at the top of the wheel into a channel connected to a head tank. Although the main function of the saqiya is for irrigation it can also be important for water supply when, for example, buildings are some distance above the source of water. The longer the chain-of-pots, i.e. the lift, the lower the rate of discharge will be. For domestic water supply this may not be a crucial factor, but in fact one of the problems of water-raising engineering is that of raising large quantities of water through a small lift. The problem can be solved by using a spiral scoop-wheel (see Figure 10), which raises water to ground level with a high degree of efficiency. This machine is very popular in Egypt nowadays, and engineers at a research station near Cairo have been trying to improve the shape of the scoop in order to achieve maximum output. Although it appears very modern in design, this is not the case, since a miniature from Baghdad dated to the sixth/twelfth century shows a spiral scoop-wheel driven by two oxen. The transmission of power is the same as that employed with the standard saqiya.
Fig. 10
The saqiya was widely used in the Muslim world from the earliest days onwards. It was introduced to the Iberian peninsula by the Muslims, where it was massively exploited. Not only was it diffused into many parts of Europe but it was also taken to the New World by Christian Spanish engineers. It has advantages over the diesel-driven pump: it can be constructed and maintained by local craftsmen and does not require the importation of fuel. The long history of the saqiya is by no means ended, and there are welcome indications that its advantages will ensure its survival for the foreseeable future.
The noria is also a very significant machine in the history of engineering. It consists of a large wheel made of timber and provided with paddles. The rim of the wheel, inside the paddles, is divided into compartments or, in another type of noria, earthenware pots similar to those of the saqiya, are lashed to the rim. The wheel is mounted on an axle over a running stream, so positioned that the paddles and the compartments, at the lowest part of their travel, are immersed in the water. The force of the current acting on the paddles causes the wheel to rotate, the compartments fill with water and discharge their contents when they reach the top of the wheel. The water usually collects in a head tank and is then conducted through a feeder channel to the irrigation system or to an urban water supply (see Figure 11). Being driven by water, the noria is self-acting and requires the presence of neither man nor animal for its operation.
Fig. 11 – Noria at Hama, Syria
The earliest description we have of the noria occurs in the writings of Vitruvius in the first century B.C.E., in words that imply that it had already been in use for some time. It was invented around 200 B.C.E. in the Near East. Its use was widespread in the Muslim world, wherever conditions were appropriate; there are attestations for its use in Iraq, Iran, Mesopotamia, Spain and elsewhere. The most famous norias are those at Hama, on the River Orontes in Syria. These are an impressive sight, the largest being over 20 metres in diameter; they discharge into the end of an aqueduct that carries the water to the town and the fields. These machines are known to have been in operation since the third/ninth century, but there were probably norias at this site much earlier than this. The large-scale use of norias was introduced to Spain by Syrian engineers. An installation similar to that at Hama was in operation at Toledo in the sixth/twelfth century and the machine was heavily exploited all over Muslim Spain. It was diffused to other parts of Europe and to East Asia, and like thesaqiya has shown remarkable powers of survival into modern times.
Five water-raising machines are described in al-Jazari’s great book on machines, composed in Diyar Bakr in 602/1206. One of these is a water-driven saqiya, a type of machine that is known to have been in everyday use in medieval Islam. Three of the others are modifications to the shaduf, obviously intended to raise the output of the traditional machine. These are important for the ideas they embody, ideas which are of importance in the development of mechanical engineering. In one of them, for example, the concept of minimising intermittent working is implied and another incorporates a crank, the first known example of a crank used as an integral part of a machine. The fifth machine is the most significant. This is a water-driven twin-cylinder pump (Figure 12). A paddle wheel is mounted on a horizontal axle over a running stream, with a gear-wheel mounted on the other end of the axle. This meshes with a horizontal gear-wheel installed in a large triangular wooden box which is mounted over a pond supplied from the stream. On the face of the second gear-wheel, near the outside, is a peg which enters a slot-rod pivoted at one corner of the box. The connecting rods were fixed to the sides of the slot-rod by staple-and-ring fittings. On the end of each connecting rod was a piston, consisting of two copper discs with a space of about 6 cm between them, the space filled by coiling hempen cord until the gap was filled. The cylinders, made of copper, were provided with suction and delivery pipes, all provided with non-return clack-valves. The two delivery pipes were joined together above the machine to form a single delivery pipe, which discharged the water at a height of about 14 m above the level of the stream. The action was as follows: when the paddle wheel turned it caused the gear-wheel on its axle to turn, and this rotated the gear-wheel in the box and the peg made the slot-rod oscillate from side to side. When one piston was on its delivery stroke the other was on its suction stroke. The important features embodied in this pump are the double-acting principle, the conversion of rotary into reciprocating motion, and the use of true suction pipes. The hand-driven pumps of classical and Hellenistic times had vertical cylinders which stood directly in the water which entered them through plate-valves in the bottoms of the cylinders on the suction strokes. The pumps could not, therefore, be positioned above the water level. A quarter-scale working model of this pump was made for the 1976 World of Islam Festival in the Science Museum, London. The construction is the same as that of the machine described by al-Jazari, except that the drive was electric. The pump works perfectly, with smooth transmission and the discharge of a steady stream of water from the delivery pipe.
Fig 12 -Al-Jazari’s two cylinder suction pump
Fig. 13 – Taqi al-Din’s six cylinder pump
Power from water and wind
There are three basic types of water-wheel, all of which had been in use for centuries before the advent of Islam and the question of their origin, and diffusion, which is still unresolved and controversial, need not concern us here. The first – the undershot wheel – is a paddle wheel mounted on a horizontal axle over a running stream (Figure 14a). Its power derives almost entirely from the velocity of the water, and it is therefore affected by seasonal changes in the rate of flow of the stream over which it is erected. Furthermore, the water level may fall, leaving the paddles partly or totally out of the water. The efficiency of the undershot wheel is not high perhaps as low as 22 per cent – because so much of the energy is dissipated by turbulence and drag. The fact that it retained its popularity over many centuries is due to the simplicity of its construction and to special measures that can be taken to increase its performance. These will be discussed later.
The overshot wheel is also vertical on a horizontal axle. Its rim is divided into bucket-like compartments into which the water discharges from above, usually from an artificial channel or ‘leat’ (see Figure 14b). Its efficiency can be as high as 66 per cent, provided all the water from the leat falls into the buckets and there is no spillage.
Fig 14 a and b
When used for corn milling, both types of vertical wheels require a pair of gears to transmit the power to the millstones. A vertical toothed wheel is mounted on the end of the water-wheel’s axle inside the mill house. This engages a lantern-pinion, whose vertical axle goes up through the floor to the milling room; it passes through the lower, fixed millstone and is fixed to the upper, rotating stone. The corn is fed into the concavity of the upper stone from a hopper (see Figure 15).
Fig. 15
The third type of wheel is horizontal, and can be subdivided into two main types. In the first of these a wheel with curved or scooped vanes is mounted at the bottom of the vertical shaft and water from an orifice fitted to the bottom of a water tower is directed on to the vanes, the flow being therefore tangential and radial (Figure 16). The second type is a vaned wheel, also fixed to the lower end of a vertical axle, and installed inside a cylinder into which the water cascades from above, turning the wheel mainly by axial flow. Axial-flow wheels can also be driven by vertical jets of water directed on to them from below. The first sub-type was known in Europe and western Asia by the sixth century C.E. at the latest. The second appears in an Arabic treatise of the third/ninth century, but is not known to have been used in Europe before the tenth/sixteenth century.
Fig. 16
The Muslim geographers and travellers leave us in no doubt as to the importance of corn-milling in the Muslim world. This importance is reflected not only in the widespread occurrence of mills, from the Iberian peninsula to Iran, but also in the very positive attitude of the writers to the potential of streams for conversion to power. The Tigris at its source, says al-Muqaddasi, would turn only one mill, and al-Istakhri looking at a fast-flowing stream in the Iranian province of Kerman, estimates that it would turn twenty mills. It is as if these travellers were rating streams at so much ‘mill-power’. This concern is understandable when we remember that the great cities of Islam such as Baghdad, Fustat and Cordoba depended upon the produce of a thriving agriculture to feed their large populations and to provide the finished materials for a thriving commerce. We therefore find that all large urban communities were provided with flour by factory milling installations, either close to the city or accessible to it by good communications. To take but one example: in the fourth/tenth century Upper Mesopotamia was the granary for Baghdad and the corn it produced was ground in large ship-mills moored on the Tigris and the Euphrates. Each mill had two pairs of stones and could produce 10 tonnes of flour in 24 hours. Nothing approaching this scale of corn-milling was known in contemporary Europe.
Water power was also used in Islam for other industrial purposes.. It seems that the use of water power in industrial applications was already established at an early date. Jabir Ibn Hayyan (b.721- d.815 C.E.) in Kitab al-Sab’in (Book of the Seventy) describes in treatise XVIII a spherical vessel for melting metals. This sphere was erected on a river and it was rotated by a water wheel. The sphere rotated continuously while it was heated by fire from underneath it. Jabir describes a similar continuous rotating motion on a river in treatise XXXIV of the same book.
In the year 134/751, after the battle of Atlakh or Talas, Chinese prisoners of war introduced the industry of paper-making in the city of Samarkand. The paper was made from linen, flax or hemp rags. Soon afterwards paper mills on the pattern of those in Samarkand were erected in Baghdad, the Yemen, Egypt, Syria, Iran, North Africa and Spain. Without doubt, the raw materials in these mills were prepared by pounding them with water-powered trip-hammers. Writing about the year 435/1044, al-Biruni tells us that gold ores were pulverized by this method “as is the case in Samarkand with the pounding of flax for paper”. Water power was also used in the Muslim world for fulling cloth, sawing timber and processing sugarcane. (It is interesting to note that the transfer of technology between China and Islam was a two-way traffic; according to Marco Polo, the Muslims taught the Chinese how to refine sugar.) It has not yet been established to what extent industrial milling in Europe was influenced by Muslim practices. A likely area of transfer is the Iberian Peninsula, where the Christians took over, in working order, many Muslim installations, including the paper mills at Jativa.
Windmills were probably known in Seistan (the western part of modern Afghanistan) before the advent of Islam. According to al-Mas’udi, a Persian claimed to the Caliph ‘Umar I that he was able to build a windmill. ‘Umar made him substantiate his claim by building one. Mills in Seistan are mentioned by the Arab geographers of the fourth/tenth century, but the first full description occurs in a book written about the year 669/1271. These were not the European type of windmill with a horizontal axle and a pair of gears. The mills were supported on substructures built for the purpose, or on the towers of castles or on the tops of hills. They consisted of an upper chamber in which the millstones were housed and a lower one for the rotor. The axle was vertical and it carried twelve or six arms covered with sails. The walls of the lower chamber were pierced with funnel-shaped ducts, with the narrower end towards the interior in order to increase the speed of the wind when it flowed on to the sails (Fig. 17). This type of windmill spread throughout Islam, and to China and India. In medieval Egypt it was used in the sugar-cane industry, but its main application was to corn-milling.
Fig. 17 – Wind mill in Seistan as described by al-Dimashqi
FINE TECHNOLOGY
The expression ‘fine technology’, applied to earlier times, embraces a whole range of devices and machines, with a multiplicity of purposes: water clocks, fountains, toys and automata and astronomical instruments. Some were designed to tell the time, as aids to scientific investigation and some to delight and amuse. What they have in common is the considerable degree of engineering skill required for their manufacture, and the use of delicate mechanisms and sensitive control systems. Many of the ideas employed in the construction of ingenious devices were useful in the later development of mechanical technology.
When we look into the origins of fine technology, we inevitably find our attention directed to the Hellenistic world and, in particular, Alexandria. Thus we find that the first complex water clock and the first musical automaton are both attributed by Vitruvius to Ctesibius, an Egyptian engineer who worked in Alexandria about 250 B.C.E, and the first major treatise on ingenious devices was composed by Philon of Byzantium a contemporary of Ctesibius. Philon’s work was continued and extended by Heron of Alexandria, who flourished in the middle of the first century C.E. The origins of the astrolabe can be firmly placed in the school of Alexandria. It was almost certainly known to Ptolemy and was described by Theon of Alexandria (c. 350 C.E.), whose writings are preserved in the treatise of Severus Sebokht, who composed his book in Kinnesrin in the north of Syria before 660 C.E. i.e. a few years after the Arab rule..
The tradition of fine technology continued uninterrupted and was further developed under Islam. Monumental water clocks in Syria continued to be installed in public places. The Abbasid Caliphs were interested in clocks and ingenious devices. The story of the clock that was presented by Harun al-Rashid (786-809), to Charlemagne in 807 C.E. is well known. It is reported also by Ibn Abi Usaybi’a that Al-Mutawakkil (d. 861) was so obsessed with ingenuous devices, literally Alat mutaharrika (moving machines), that he favoured the Bani Musa who took advantage of the Caliph’s obsession and persecuted their opponents. The Banu Musa wrote their book on al-Hiyal during this period, and we infer from this story that they had actually constructed and operated their devices to please and satisfy the Caliph. In Kitab al-hayawan, al-Jahiz(776-867 C.E.) when discussing the measurement of time, says: ” Our kings and scientists use the astrolabe by day and the binkamat (clocks) by night” It is reported also by historians that Nasir al Dawla of Diyar Bakr (d. 1061 C.E.) had constructed a public binkm (clock) for the city of Mayyafariqin in the year 1012 C.E. When al-Jazari wrote his book 200 hundred years later in Diyar Bakr, he was describing a tradition that was firmly established in that area. Other clocks in public places were installed in some other cities of the Islamic east.
The technology of clock-making was transferred to Muslim Spain and to Al-Maghrib. About the year 1050 C.E. al-Zarqali constructed a large water clock on the banks of the Tagus at Toledo in Spain. The clock was still in operation when the Christians occupied Toledo in 1085 C.E. A manuscript describing Andalusian monumental clocks was written in the eleventh century by Ibn Khalaf al-Muradi. Water clocks were constructed for public places in al-Maghrib. The remains of a public water clock in Fas can still be seen.
In Damascus, Muúammad al-Khurasani al-Sa’ati (the clockmaker) built a monumental clock around 1160. The clock was described by several travellers, It was still in operation when Ibn Batuta visited Damascus in the 14th century. Ridwan al-Sa’ati, the son of the clock’s maker, re-built the clock and gave a detailed description of its construction in 1203. Al-Jazari’s book was written in Amid, Diyar Bakr, in 1206. This book together with Kitab al-hiyal of Bani Musa, are the most important treatises on fine technology that came down to us. It is reported also that the astronomer ‘Ali al-Qushji (d. 1474) wrote a tadhkira or treatise on spiritual machines. The last important writer on the same subject was Taqi al-Din ibn Ma’ruf al-Dimashqi who wrote a book on water clocks in 1552 and another on mechanical clocks in 1556.
In addition to water clocks and ingenious devices, the making of astrolabes and of geared astronomical mechanisms in Islam continued throughout the centuries as will be shown later. Our knowledge of Islamic fine technology will continue to improve with the publication of more research results.
It is not easy in the space available to demonstrate the achievements of Muslim engineers in the field of fine technology. This can probably best be done by considering a few of the more important Islamic works, placing emphasis upon their innovative features. The Banu Musa were three brothers – Muhammad, Ahmad and al-Hasan – who were members of the courtly circles of the Abbasid Caliph al-Ma’mun (198/813-218/833) and his successors. This was the period that witnessed the first flowering of Arabic science, both in the translation of Greek and Syriac works into Arabic and in original scientific and technological works, and much of this activity was carried out under the direction and patronage of the Banu Musa. They were also scientists and engineers in their own right and wrote about twenty treatises, only two of which are known to have survived. One of these, The Book of Ingenious Devices (Kitab al-hiyal), written in Baghdad about 235/850, is our present concern. It contains descriptions of a hundred devices, most of which are trick vessels, together with self-feeding and self-trimming lamps, a gas mask for use in polluted wells and a mechanical grab. The trick vessels exhibit a bewildering variety of effects, for example:
Model 26. A jar with an outlet pipe: during inpouring of a liquid the pourer can, according to choice, allow the liquid to discharge, or prevent its discharging.
Model 43. A jar with a tap, into which three different liquids can be poured without mixing. When the tap is opened the liquids discharge in the sequence in which they were poured in.
Model 77. A basin beside a closed reservoir. When moderate quantities of water are taken from the basin, like quantities run into it from a pipe at the bottom of the reservoir; if, however, a large quantity is taken no replenishment occurs.
These effects, and many others, were produced by the ingenious combination of a number of hydraulic and mechanical components, two of which are shown in Figure 18. Figure 18(a) shows a double concentric siphon: pipe bd passes through plate f which divides the upper chamber from the lower, the joint being airtight. Pipe a-cc is placed over end b of pipe bd and held to it by soldering pieces of copper wire between the two pipes. End a of this pipe is closed. Another wide piece of pipe e-gg, its end e closed, is fixed over end d of pipe bd. The effect of introducing this device into a flow system is to create an airlock once the flow of liquid is interrupted, so that the flow cannot be restarted except under certain conditions. Onlookers could therefore be startled by unexpected events. The double concentric siphon does not occur in any of the Greek writings nor, as far as we know, in any Arabic work except the Banu Musas book. The fluid mechanics of its operation are surprisingly complex.
Fig. 18 a and b
The second mechanism is shown in Figure 18 (b). Seat b of a conical valve is soldered to the end of pipe a, and plug c of the valve is soldered to the end of a vertical rod, the other end of which is soldered to the top of float f. Above the float and integral with it is the small tank d which has a small hole e in the bottom of one of its sides. The float rests on the surface of the water in the slightly larger tank g. It works like this: water is poured in at a and runs through the valve into tank d; the weight of the liquid in tank d stops the valve-rod from rising and so the valve remains open. When pouring is stopped tank d empties into tank g through hole e, float f rises and the valve closes. No further inpouring can then take place. Conical valves do not appear in the works of Philon and Heron. They were made by casting the plug and seat together in a single mould, the material nearly always being bronze. Plug and seat were then ground together with emery powder to a watertight fit.
A distinguishing feature of the Banu Musa’s work is therefore their confident use of conical valves as integral parts of flow systems. More generally, they show an astonishing empirical mastery of the use of small variations in hydrostatic and aerostatic pressures to produce a variety of effects. Although their work was well-known in the Muslim world for centuries after their time, none of their successors ever attempted to emulate them. They had, in fact, taken the subject to its limits with the materials and techniques then available, and nothing similar was done until the introduction of pneumatic instrumentation in modern times.
A most important treatise was written in Muslim Spain in the fifth/eleventh century by al-Muradi. Unfortunately, the only known manuscript of the work is so badly defaced that it is impossible to deduce from it precisely how any of the machines was constructed. Most of the devices were water clocks, but the first five were large automata machines that incorporated several significant features. Each of them, for example, was driven by a full-size water wheel, a method that was employed in China at the same period to drive a very large monumental water clock. The automata were of the types that were common in water clocks, for example a set of doors in a row that open at successive intervals to reveal jackwork figures. The text mentions both segmental and epicyclic gears. (In segmental gears one of a pair of meshing gear-wheels has teeth on only part of its perimeter; the mechanism permits intermittent transmission of power). Although the illustrations are in other respects quite incomprehensible, they clearly show gear-trains incorporating both these types of gearing. This is extremely important: we have met simple gears in mills and water-raising machines, but this is the first known case of complex gears used to transmit high torque. It is also the earliest record we have of segmental and epicyclic gears. Sophisticated gears for transmitting high torque, first appeared in Europe in the astronomical clock completed by Giovanni de’ Dondi about 1365.C.E.
Al-Jazari completed his magnificent book on machines in Diyar Bakr in the year 602/1206. This is the most remarkable engineering document to have come down to us, from any cultural area, until the Renaissance. In one respect it is unique: it was written at the wish of al-Jazari’s master, the Artuqid Sultan Nasir al-Din Mahmud ibn Muhammad, so that a record of the fragile devices could be available for succeeding generations of craftsmen, long after the devices themselves had perished. Each of the fifty chapters – text and illustrations – was therefore composed with detailed instructions for manufacture, so that the machines could be reconstructed by later craftsmen; Al-Jazari was successful in this aim because several of his devices including a monumental water clock, have been constructed by modern craftsmen working from al-Jazari’s instructions. The works of other writers, while they often describe the operation of the machines quite adequately, give only the sketchiest details of construction. Al-Jazari’s willingness and ability to communicate the knowledge gained by training, experience and informed experiment therefore endow his work with immense value.
AI-Jazari was an engineer who took pride in continuing a long tradition of mechanical technology, and his work may in many respects be regarded as the epitome and the summit of the Islamic achievement in this field. With one or two notable exceptions, such as the complex gear-trains of al-Muradi, it is safe to assume that he dealt with most of the machines that had been known to his predecessors, while introducing innovations and improvements of his own. Indeed, he often acknowledges the work of earlier engineers, such as Archimedes or the Banu Musa in connection with a particular technique or type of machine, describes the earlier construction accurately and then tells us how he improved upon it. For instance, a certain type of flow regulator was used in water clocks by both Hellenistic and Muslim engineers. Al-Jazari found by experiment that it was inaccurate, and describes how he made an accurate instrument by carefully calibrating a small orifice to produce correct rates of flow under various heads of water.
One example will have to suffice to give some idea of al-Jazari’s methods and the type of devices he constructed. This is the water machinery and some of the associated mechanisms from his third and fourth water clocks. These were driven by the submersible bowl or tahrajar a device that was normally used for timing the period of allocation of irrigation water to farmers. These two clocks are the only examples we have of the adaptation of the tahrajar for timekeeping, and it seems likely that this system was al-Jazari’s own invention. Figure 19 shows the basic principles; it would be impossible to describe either of the clocks fully, since they had a number of automata together with the mechanisms for activating them, some of them very ingenious.
Fig. 19
The bowl A with a calibrated orifice in its underside rests on the surface of the water in tank N, to which it is connected by the three flat pin-jointed links H. A rod is soldered across a diameter of the bowl, with hole K in its centre. At the top of the clock, supported on four columns, is the ‘castle’, a square brass box with a detachable dome. Inside the castle is a ball-release mechanism (not shown) from which a channel leads to the head F of a bird. The tail of the serpent, in effect a pulley, rotates on an axle that rests in bearings in transoms fixed between each pair of columns. The open mouth of the serpent is just below the head of the bird. A light chain D runs from the underside of the bowl to a staple in the tail of the serpent. A wire H is tied to hole K and to the ball-release. At the beginning of a time period – an hour or half an hour – the empty bowl is on the surface of the water. It sinks slowly until at the end of the period it suddenly submerges. Wire H operates the ball-release, and a ball runs into the bird’s mouth and out of its hinged beak into the mouth of the serpent. The serpent’s head sinks, and chain D lifts the bowl, which tilts due to the combined action of the chain and links B and discharges all its contents. The ball drops from the serpent’s mouth on to a cymbal, and the serpent’s head rises to its previous position. The empty bowl is again horizontal on the surface of the water, and the cycle re-starts. This is therefore a closed-loop system, since the clock will continue working as long as there are balls in the magazine. The concept of continuous operation occurs elsewhere in al-Jazari’s work; in his first clock, for example, the head of water over the orifice is kept constant by a hydraulic feed-back control system.
A number of ideas and techniques appear for the first time in al-Jazari’s work. These include the double-acting pump with suction pipes and the use of a crank in a machine (both already mentioned); accurate calibration of orifices; lamination of timber to minimize warping; static balancing of wheels; use of paper models to establish a design; and casting of metals in closed mould-boxes with green sand. There is also an indication that he knew of a method for controlling the speed of rotation of a wheel by an escapement of some kind. This is very significant when we consider a clock described in a Spanish work compiled in 1277 C.E. in which all the chapters are translations or paraphrases of earlier Arabic works. The clock consisted of a large drum made of walnut or jujube wood tightly assembled and sealed with wax or resin. The interior of the drum was divided into twelve compartments, with small holes between the compartments through which mercury flowed. Enough mercury was enclosed to fill just half the compartments. The drum was mounted on the same axle as a large wheel powered by a weight-drive wound around the wheel. Also on the axle was a pinion with six teeth that meshed with thirty-six oaken teeth on the rim of an astrolabic dial. The mercury drum and the pinion made a complete revolution in 4 hours and the astrolabic dial made a complete revolution in 24 hours. Clocks incorporating this principle are known to work satisfactorily, since many of them were made in Europe in the seventeenth and eighteenth centuries. This type of timepiece, however, with its effective mercury escapement, had been known in Islam since the fifth/eleventh century, at least 200 years before the first appearance of weight-driven clocks in the West.
The astrolabe was the astronomical instrument par excellence of the Middle Ages; from its Hellenistic origins it was brought to perfection by Muslim scientists and craftsmen. Essentially it consists of a circular plate inside a raised annulus that is divided into degrees, the annulus being soldered to the rim of a base-plate. The first plate is engraved with the lines of azimuth and altitude for the latitude of the observer. Turning on a central pin above the plate is the rete or spider, made of open metalwork. This is essentially a star map, the principal fixed stars appearing as holes or gemstones; there is also a circle for the sun’s ecliptic. Rotating above the rete was an alidade. Both plate and rete were marked out by stereographic projection. The instrument was usually made of brass (see Figure 5). A number of astronomical problems, which otherwise have to be solved by tedious computation, can be solved very quickly by using the astrolabe. It has been established that the first European treatises on the astrolabe were of Arabic inspiration and were written in Latin at the beginning of the fifth/eleventh century in the abbey of Ripoll in Catalonia. From this centre the knowledge of the instrument was diffused to the rest of Europe.
Other computing instruments were devised in the Muslim world in the later Middle Ages, perhaps the most important of these being equatoria, which were invented in Muslim Spain early in the fifth/eleventh century. The objective of the equatorium was the determination of the longitude of any one of the planets at a given time. This was done by constructing to scale and by mechanical and graphical means the Ptolemaic configuration for that particular planet at the given instant. As with the astrolabe, knowledge of equatoria was diffused into Europe from the Muslim world.
An important aspect of Islamic fine technology is the tradition of geared astronomical instruments which were described in Arabic literature. The most notable example is the astronomical geared mechanism that was described by al-Biruni (Fig 20) and called by him huqq al-Qamar (Box of the Moon). From al Biruni’s text we understand that these mechanisms were known in Islamic astronomy. A surviving example is the geared calendar made by Muhammad b. Abi Bakr al-Ibari al-Isfahani dated 1221/2 C.E. which is found on the back of an astrolabe. This instrument is in the collection of the Museum of the History of Science at Oxford. The tradition of geared astronomical instruments continued until later centuries and we find a description of a geared mechanism in Taqi al-Din’s Al-Turuq al-Saniyya (1552 C.E.) under the same name of al-Biruni’s instrument: Huqq al-Qamar. Taqi al-Din says “as to the method of use, there are special treatises for this purpose and we need not elaborate on it here”
Fig. 20 – Al-Biruni’s huqq al-qamar
Derek J. de Solla Price when describing the Antikythera mechanism (90 C.E.) remarked that " It seems likely that the Antikythera tradition was part of a corpus of knowledge that has since been lost but was known to the Arabs. It was developed and transmitted by them to medieval Europe, where it became the foundation for the whole range of subsequent invention in the field of clockwork”
The astronomical gearing mechanisms just mentioned were manually driven. But as mentioned above, al-Muradi’s machine, (Model 5), shows a system of gears for transmitting torque that is much more complex than any other power-driven gears known to have existed so early.
Many of the ideas that were to be embodied in the mechanical clock had been introduced centuries before its invention: complex gear trains, segmental gears in al‑Muradi and al‑Jazari, epicycle gears in al‑Muradi, celestial and biological simulations in the automata‑machines and water clocks of Hellenistic and Islamic engineers; weight‑drives in Islamic mercury clocks and. pumps, escapements in mercury docks, and other methods of controlling the speeds of water wheels. The heavy floats in water clocks may also be regarded as weights, with the constant‑head system as the escapement.
We know that the Christians in Spain learned about Muslim water clocks, not only by the translation of Arabic works into Spanish but also by the inspection of actual clocks in Toledo. This knowledge was transferred to Europe and there was a substantial advance in the fifth/eleventh century in the techniques of hydraulic time-keeping. In a treatise written by Robertus Anglicus in 1271, it is mentioned that the clockmakers – i.e. the makers of water clocks – were trying to solve the problem of the mechanical escapement and had almost reached their objective. The first effective escapement appeared a few years later. This evidence, circumstantial though it is, points strongly to an Islamic influence upon the invention of the mechanical clock.
In the middle of the sixteenth century, mechanical clocks from Germany and other European countries reached the Ottoman Empire. Taqi al-Din wrote in 1556 C.E. his book Al-Kawakib al-Durriyya fi al-binkamat al-Dawriyya (The Brightest Stars for the Construction of Mechanical Clocks) in which he described a clock incorporating some of his inventions. He refers to the pocket watch (that uses a spring)from which we infer that they were quite common.
