The present study is based on the 13th-century treatise of Radwan al-Sa'ati, which describes in details the clock of the Jayrûn Gate. Far from being historical or critical, this technical and analytical study allows, by dissecting the text line by line, to highlight the characteristics of the elements of the clock studied. More precisely, the nature, shape, dimensions, and position of the parts constituting it. Added to that, the precautions to be taken for a good operation of the device. We undertake a thorough study hoping to be able to unveil all the secrets of this clock, uncovered by the orientalist historians who studied this treatise. However, their detailed analyses, each as important as the other, have played a crucial role in paving the path we have followed in this study.
The Editor’s Note: We are pleased to publish this article which is a translation of the French article previously published by the author on this novel water-clock. To ensure that the technical translation does justice to the original manuscript, we invite readers to contact the author for any comments or amendments:
Ever since antiquity, man has sought to assess the passage of time punctuated by years, seasons, days and nights. He measured the time by observing natural phenomena: the succession of day and night, the different positions of the sun in the sky, the phases of the moon, the tides, and others.
To materialize these measurements, man invented instruments, the first being the gnomon, a simple stick planted vertically in the ground.
Over the centuries, the gnomon was perfected to give birth to the sundial. A little later, the clepsydra made its appearance. Indeed, to measure the flow of time, the Greeks of the classical period (5th -4th century BC) used the clepsydra, a simple vase with a hole at its base that let the water escape. The duration of the flow varies, of course, according to the diameter of the orifice and the quantity of liquid to be evacuated. This flow does not suppose any correlation with the solar time sequence and is completely independent of it.
By its simple principle, this instrument is likely to render very varied services.
In the courts, the clepsydra, the haunt of prolix speakers, gives each person his or her speaking time, in proportion to the importance of the debates. In military life, it determines the time of the relief of the sentries. For the researchers, it is used to determine the duration of their observations.
It was during the Alexandrian period (3rd-2nd century B.C.) that the idea of applying the clepsydra principle to the measurement of daytime hours (sundials were unusable in cloudy weather) appeared. Important difficulties arose concerning the regulation of a constant flow of water escaping from the clepsydra orifice. To this end, several processes were adopted, and the clepsydra went beyond its simplicity to give birth to real water clocks measuring, with a fair amount of precision, the day and night hours.
It is this water clock that invaded the East from the 9th to the 14th century and more particularly the Arab-Muslim world.
Among this flurry of clocks, we distinguish the hydraulic clock built in Damascus in Syria, under the Umayyad dynasty, precisely at the Jayrûn gate of the mosque bearing their name.
The clock is described by the traveller Ibn Jubayr1 in his account (Rihlat), as follows:
“To the right of the person coming out of the Jayrûn gate, there is a chamber in the wall of the front gallery in the form of a large circular arch with several bronze windows. These windows, in the number of the hours of the day, are provided with small open shutters arranged with much art. When an hour passes, two yellow copper balls fall into the beaks of two falcons. The two falcons, made of the same metal, stand on a beaker, also made of yellow copper, one under the first door and the other under the last. The beaker has a hole in them to let the balls through, which are piled up in a small box behind the wall; the falcons then stretch their necks towards the beaker and quickly spit out the balls. The mechanism is so curious that it was thought to work by magic.
When the balls fall into the beaker, a sound is heard and the door corresponding to the hour that has just passed, closes immediately to reveal a yellow copper plaque. The mechanism continues with each passing hour until all the doors are closed, the hours pass, and everything returns to its original position.
During the night another mechanism is assigned to the clock. In the archway that curves over the windows, we find twelve openwork discs of yellow copper to which a pane of glass, placed inside the wall of the room, is fitted. All this mechanism is arranged behind the previously mentioned windows. Behind each window, a lamp is suspended where the water rotates for an hour. At the end of each hour, the glass reflects the light of the lamp, and the light rays flood the disc, so a reddish disc is seen. The same mechanism is repeated for the next glass until all the hours of the night are over and all the discs take a reddish color.
The clock is entrusted to a professional overseer who is responsible for ensuring that the clock is working properly; he must reopen the doors and return the counterweights to their original position. It is this clock that people call al-Minjana.”
A more simplified description of this hydraulic clock is written by the traveler Ibn Battuta2
“To the right of the person coming out of Bab Jayrûn, there is a very high room in the form of a big window, containing other small open windows, with the number of hours of the day, having doors, of yellow color on the outside and green on the inside. At each hour, the door turns, and the yellow is replaced by green. It seems that someone, at every hour passed, makes this mechanism work from the inside”.
It was not until the 13th century that Radwan al-Sa’ati became interested in this clock and dedicated a treatise to it, which reveals all the details concerning its construction and operation. This treatise entitled: “‘Ilm al-Sa’at wa al-‘Amal biha” is the subject of the present study.
Radwan ibn Muhammad ibn Ali al-khurasani fakhr al-Din al Sa’ati, was born in Damascus where he spent his entire youth. As a man of science, his father Muhammad was incomparable in his knowledge of astronomy and clock-making. It was he who rebuilt the clock of the great mosque of Damascus during the reign of Nur al -Din al-Zanki who offered him a considerable amount of money. The clock remained under his direction until his death.
His son Fakhr al-Din Radwan was a physicist, poet and physician and a prolific author. He wrote medical treatises in which he noted his remarks on the canon of ibn Sina. But it is his talent as a clockmaker that interests us in this study. The treatise in question was completed in the year 600 H /1203CE.
There are three copies of this treatise
1- Arabic MS 1348, dated 961/1554, Forschungsbibliothek, Gotha
2- MS Istanbul Köprülü.I.949, copied from the original by the Egyptian engineer Abdallah al -Qipjaqi in Cairo 658 H/ 1260 CE.
3- MS 24 Taymûr Sina’a. Institute of Arabic Treatises Cairo Egypt
Other than these treatises, the treaty is mentioned by the historians Sutter3 and Sarton4 as well as by the bibliographer ibn Abi Usaybi’a5
This treatise has attracted the attention of renowned orientalist researchers interested in the history of science in the Arab-Muslim world.
The first study of this treatise was made by Eilhard Wiedman and Fritz Hausser6 in 1915. It is a rather interesting analytical study with well-detailed technical illustrations, revealing some of the secrets of al-Sa’ati’s clock.
In 1981, Donald Hill7 published an article devoted to the Radwan al-Sa’ati clock. In this study, Hill refers to the article by Wiedman and Hausser while borrowing some illustrations from them. Donald Hill carefully details the elements of the clock following Radwan’s descriptions, but unfortunately, he does not highlight the specifics of each element in separate paragraphs, which confuses the non-specialized reader
Despite all the important details included in this study, some technical processes, described by Radwan are omitted. Notably, Hill makes no reference to the الحلزون (al-Halzoun) described in the text. He does not mention the helical circle(الدائرة الحلزونية) in which the pointed head of the nighttime axle rotates, nor does he mention the helical nighttime carrier referred to by Radwan as an important invention of his father.
A bold and rather important contribution is attributed to Fuat Sezgin8, who was the first to build, in the 1980s, a model of this clock and exhibit it in the museum of his institute in Frankfurt.
Abdel Aziz Al-Jaraki9 criticizes this model on the grounds that it does not correspond to the details described by Radwan. For his part, Abdel Aziz analyzes Radwan’s treatise in two articles published on the website “Muslim Heritage» but is content to develop the historical side while simplifying the technical and analytical study.
Muhammad Ahmad Dahman10 publishes a special edition of the treatise without, however, mentioning any technical or analytical comments regarding the construction or operation of the clock.
This treatise entitled “Ilm al-Sa’at wa al-‘amal biha” (the science of clocks and their use) is composed of 198 pages of clear, neat and readable writing, but the order of the ideas leaves something to be desired, and unnecessary repetitions clutter the treatise.
In the introduction to his treatise, Radwan states the following:
“… The clock of the great mosque of Damascus is entrusted, for many years, to several people who take care of its proper functioning. My father is the last clockmaker to take care of it.
This clock has a long water tank with a capacity of twenty jars, the technician had trouble filling it every day. The sun clock was not part of the main clock as well as the semi-circle that hides the illuminated windows one after the other. Only one wheel is connected to the daytime hours. A semi-circle is dedicated to the zodiac sign plates. Only the beaks of the falcons are visible. Other anomalies appear concerning the pulleys, the ropes, the counterweights, and many other elements of the machine
In this state, my father watches, during several years, to its functioning and this until the year 564H/1168CE, when a fire ravages the place of “Labbadine” and the clock is completely destroyed. My father then ensures its reconstruction and takes care of it until his death.
With this reconstruction, my father corrects simple anomalies while adding new features. He replaces the large main tank with a capacity of twenty jars with a much smaller one of two jars. He adds the semicircle that hides successively the illuminated windows. It grants to the diurnal hours two coaxial wheels. He uses a complete circle for the zodiac signs. To the right and left of the machine, he places two falcons that tilt while opening their beaks to spit balls and he adds the sun clock. He associates the helical carrier with the nighttime axle.
After his death, three clockmakers succeeded each other in maintaining it, Ibn al-Nanaccash, his student known as Ibn al-Hajeb and Abi al-Fadl al-Najjar. None of the three ever managed to keep it in good working order, and every damaged part was destroyed without being replaced. Ibn al-Naccach destroyed the semicircle and the sun clock. The sun clock runs parallel to the movement of the sun in the sky. When the sun appears, the sun disk of the clock appears. At noon, this disk is placed in the middle of the zodiac signs…
… Having noticed, that with time, the clock loses most of its functions, I decided, in memory of my father to renovate it. I rebuilt it piece by piece while adding some important new mechanisms. I replaced the Archimedes’ al-Jaz’a pipe with a curved tube, whose role is to regulate the flow of water from al-Jaz’a. I added a tap to the main tank to wash it and rid it of water residues…
… After the revival of the clock, I decided to write this treatise to save all this knowledge for those who want to benefit from it.”
After this long introduction, Radwan divides his treatise into five chapters. The first three chapters are devoted to the description of the different elements of the clock, the fourth details the functioning of these elements and the fifth relates important advice to ensure the proper functioning of the device.
Fig.1.from Radwan’s treatise describes the general aspect of the clock.
On the far right of the figure is the basic mechanism of the clock. This is the main tank (1البنكان ) equipped with its float (2 الطفاف ). The latter is connected to the daylight hours wheel (4), through a chain wound on the groove of another wheel (3) coaxial to the daytime wheel. To this same float, the nighttime wheel (13) is connected through another chain ending with a counterweight (5). This chain is wound on several double groove pulleys (33).
At the bottom left of the figure we find the mechanism of release of the doors indicating respectively the elapsed hours. This mechanism is formed by a metal box, the mouse (7) carrying a rod topped by a crescent- shape index (32). The role of the latter is to indicate the fractions of hours. The mouse (الفأرة) topped by its crescent moves on a rail (23) fixed in front of the doors. Its movement is provided by two chains (29), one is connected to the wheel of daylight hours (4), the other through a pulley, to a counterweight.
In the center, there are twelve doors (9,10,11…) adjusted between the thirteen columns (34). Each door is surmounted by an arch fitted with a shutter. The three right-hand doors can be seen in the figure along with the ropes and the counterweights (24). Above the series of shutters (20,21,22…), we find the canals (12, 19) of the balls. These canals bend on both sides.
Each canal ends in the head of a falcon (6). The latter, mobile around a horizontal axle, bends under the action of the ball, spits it into the appropriate bowl (8) and straightens up under the action of a counterweight (27). The falcons can be seen on both sides of the figure.
Above the shutters, we notice the mechanism of ball release. It is a simple metal rod with a ring at one end and two metal half-discs (31) at the other end, which fit into the two ducts of the main canal and trap two balls (16).
In the center of the illustration and towards the top we see the half-disc of the night hours (30); it is perforated with twelve openings (15) lighting up successively to indicate the flow of the hours. The illumination is ensured by a kerosene lamp (14) hung to the ceiling by mobile chains ensuring its displacement towards the center of the arc of the windows. On the same axle of the half disk is fixed the wheel of the night hours (13). On its groove, the chain (26) winds up to the counterweight (5). To the right are the pulleys (33) necessary to balance the chains. Two pillars (25) are tuned to the clock.
We discuss a technical description of each element of the clock according to the text of al-Sa’ati.
It is a hollow metallic cylinder made of beaten red copper, with a height of 6.5 span (1.43m), a perimeter of four span (88cm) and six joined fingers (12cm), i.e. a perimeter of 1m. Its base is made of a rather thick metal sheet, welded to its crown. This cylinder must be well tinned and mounted on three equidistant feet. A pipe of six joined fingers (12 cm) in diameter and seven span (1.54m) in length is inserted in an opening made in its lateral surface, at six joined fingers (12cm) from its base. For this pipe to be firmly attached to the cylinder, Radwan advises us to weld, on both sides of the surface of the cylinder, a square piece of metal, of suitable thickness and hollowed out in its center of a circle of the same diameter as the pipe, called “al-Lubnah” (fig.2.)
“To make this cylinder, Radwan proposes to use red copper sheets of medium thickness, to roll them in a circle according to the perimeter of the cylinder, to superimpose them and to weld them strongly one to the other to avoid any water leakage. A thick metal sheet with the edge facing upwards forms the base of the cylinder. To consolidate this cylinder, three metal rings are welded around its perimeter.”
This cylinder is equipped with a tap (al-Bathyoun.البثيون) located near the base. This tap is formed by an open pipe inserted in an opening on its lateral surface. A solid metal rod ending in a ball is inserted into this pipe. Through the washers, small cords are attached to the open pipe and its stem to tighten the two parts of the tap to prevent water leakage. A thin rod bent at right angles, welded to the ball serves as an arm to facilitate the opening of the tap. To be raised from the ground of a suitable height, three triangular feet of the same metal, are fixed to its base (fig.2.).
It is a hollow small cylinder-like “al-Kayl” that serves as a capacity measure for dry materials. This cylinder formed of the same metal as the main tank, with height no more than two-thirds of a span (almost 14.7 cm), has a square opening on its side surface facing the open pipe of the main tank (al-Binkan البنكان). This pipe is inserted into this opening while noting that its edges should have a curvature that facilitates the flow of water. A lid with a downward curved edge closes the cylinder tightly. This lid has a large opening in the center to allow water to flow freely. Both sides of the cylinder are tinned with tin (fig.3.).
“Radwan mentions: the bases of the two tanks, main and settling tanks, must be in the same horizontal plane” (fig.9.).
It is a hollow metal cylinder of the same section as the settling tank “al-Kayl” and of double-height. Its bayonet lid is pierced by a circular opening to which is attached, towards the inside, a small hollow metallic cylinder ready to receive the conical index, of the float. On both sides of the center of this lid, two holes are pierced, according to a horizontal diameter, leaving free passage to the silk threads linked to the float. On the surface of this lid and according to a horizontal diameter, a metal blade is fixed with an edge turned downwards.
This blade, which is longer than the diameter, is fitted with hooks at both ends and serves to hold the cover on the cylinder (fig.4.). The latter is mounted on a wooden support in the shape of a truncated cone and, together with its lid, forms the reservoir that regulates the water flow. Its side surface is pierced by a rectangular opening into which a pipe is inserted as the female part (fig.4.). The male part is a cylindrical pipe attached to the center of the al-Jaz’a regulating disk.
To assemble the settling tank “al-Kayl” and the regulating tank “al-Rub'” Radwan proposes the following:
“A canal formed by a pipe of square cross-section pierced, close to its two ends, with two openings, is fixed on the covers of the two tanks. Each opening of this canal must coincide with the corresponding opening of each tank. Thus, the water that rises from the opening of the full settling tank into the canal flows through the opening of al-Rub’ to fill it” (fig.4.).
a – Main tank float
It is formed by a hollow half- sphere weighed by two metal disks of different diameters. The larger one has a slightly smaller diameter than the tank. The small disk concentric to the first one is weighted by lead. This float carries a metal rod topped by a ring followed by a hook and a second ring. The latter receives respectively the chains of the day and night hours (fig.5.).
In the text, Radwan explains how to secure this ballast. He says:
“The lead is melted and spread evenly over the surface of the small disk to have a fairly thick homogeneous layer, and it is left to cool. The mass of the lead to be spread must be five ratl”.
b – Regulating tank float
It has the shape of a very light empty metal pear with a horizontal base, a diameter slightly smaller than that of the tank and a height equal to a third of that of al-Kayl. This pear is surmounted by a cone metal index capable of adhering to the water opening of its tank. Two rings are welded on either side of this index, to which silk threads have been attached (fig.6). The silk wires are attached to the lid, and they emerge from it through the holes associated with them.
a – ( al-jaz’a) “الجزعة”
Definition of Radwan:
الجزعة الجسم المثقوب من جزع او عقيق او ذهب او نحاس ومنها يخرج الماء.
“Al-Jaz’a is the pierced body, of onyx, carnelian, gold or copper, from which water gushed out.”
This pierced body can be a small stone of onyx, carnelian, or gold.
b -The tube of al-Jaz’a “أنبوب الجزعة”
It is a small empty tube with open ends, embedded in the disc of al-Jaz’a, One of its two ends gives access to the curved tube, the other is covered by al-Jaz’a.
c- The disc of al-Jaz’a (صفيحة الجزعة) or regulating disc
It is a metal disc of the same diameter as the hollow part of the zodiac signs’ plate, able to fit into the latter. In the center of this disk is fixed, perpendicular to its plane, a pipe (p) called by Radwan the male part which female part is another pipe (q) inserted in the lateral opening of the regulating tank al-Rub’. The pipe (p), which has a side opening, can rotate freely in the pipe (q) once it is inserted in the latter. The regulating disc that is turned in the zodiac signs’ plate ensures this rotation.
A curved tube (J), having a closed end with a lateral opening, is welded to the opening of the pipe (p) by its open end, so that this tube runs vertically along a spoke at the back of the regulator disc. This last one is pierced by an opening, just in front of one of the curved pipes (J). On the edge, above this opening, an index is fixed. This is a yellow copper pin with a pointed end. Another identical index is fixed to the edge of the disc at a distinct point opposite the first one according to diameter (fig.7.) A very thin empty tube (t), named the whistle الصفارة , is embedded in the hole of the regulating disc, one of its two ends goes through this disc and ends up in the side opening to the curved tube, the other end is covered by al-Jaz’a. This last one will be fixed on the regulating disc while covering the opening of this last one. The rotation of the disc is ensured by a small rod fixed perpendicularly to its plane, near its perimeter.
“the pipe which forms the male part is pierced at its end with a thin rectangular opening. This end is inside the al-Rub’ regulating tank. A thin wooden rod is inserted into this opening to hold the two pipes together when the clock is running.”
d- The curved tube (الأنبوب المعوج)
This is a curved tube (J), equal in length to the radius of the regulating disk or slightly less. It is pierced very close to its closed end with an opening leaving a free passage for water. This tube is inserted into the opening in the tube (p) so that it runs along a radius on the back of the disc while touching it and must be well welded in this position (fig.7.). A very small empty tube (t) is inserted into this pipe and passes through the opening of al-Jaz’a. Then al-Jaz’a is welded onto the regulating disk while hiding the small tube and the opening already made in front of the opening of the curved tube.
e-The plate of the zodiac signs (صينية البروج)
This plate is formed by a metal disc divided into four equal parts by two perpendicular diameters. Each part is divided into three equal sections. The zodiac sign is inscribed in each section, starting with Aries from the trigonometric origin and rotating in the positive direction.
A crown is drawn around the zodiac signs. Six equidistant lines are drawn in this crown above each sign (fig.8). This plate is attached, in its center and perpendicular to its plane, to the pipe (q) forming the female part of the lateral opening of al-Rub’.
Note: Radwan explains the advantages of the full circle of the zodiac signs’ plate over the semicircle as follows:
“We favored the full circle for the zodiac signs over the semicircle proposed by Archimedes. In the case of the full circle the diurnal hours are marked by the position of al-Jaz’ at its initial position and the nocturnal hours by its position at its opposite. Thus, the flow of water is balanced with the elapsed time.
For the semicircle, the situation is not the same, because each sign does not have its opposite. Concerning the full circle: when the sun is on the sign of Gemini, the tube of al-Jaz’a points to the signs belonging to the upper semicircle, in this case the flow of water is low and corresponds to the length of that day. At night the tube of al-Jaz’a will be turned to the opposite sign (the Sagittarius), it will occupy a position in the lower semicircle and the water flow will be stronger. Thus, the length of the day with a low flow corresponds to the short duration of the night with a high flow.”
We can distinguish the main tank al-Binkan,( البنكان) mounted on its three legs. It is equipped with its tap and its float. The latter is suspended by a chain. The settling tank al-kayl (الكيل) is connected to the main tank by a large exhaust pipe. The regulating tank al-Rub’ ( الربع ) equipped with its conical index float, communicates with the settling tank through the rectangular canal that allows the free passage of water. The zodiac signs’ plate with its regulating disc and its bent pipe (fig.9.) is linked to the regulating tank.
“we notice that the bases of the main tank and the settling tank belong to the same horizontal plane.”
The whole mechanism of the monumental clock of Radwan is embedded in a rather rigid rectangular chamber. In the front of this chamber, a rectangular opening is made where twelve equidistant red copper columns are mounted. These columns are flat on the inside and rounded on the outside and are decorated with nails gilded and inlaid with various engravings. They are pierced on both sides, near the top, with small holes capable of receiving the tenons of the shutters. Between two columns stands a door that can be moved around a vertical axle of rotation (fig.10). A clamp maintains this door in its stable balanced position. Two protruding tenons are fixed, on each door, on both sides of its axle of rotation. Two red copper bars are fixed above and below the door. These bars are gilded, and each has twelve holes facing the tenons of the doors. A chamber for the counterweight is built under the doors. The rotation of the doors is ensured by metal chains that link the tenons to the counterweight, and silk cords for the rotation of the shutters.
To recognize the fractions of an hour, Radwan proposes the following method:
“One takes a metal blade of length equal to the distance that covers the twelve doors. It is divided into twelve equal parts, each of which contains twelve equidistant divisions, so that in all there are 144 divisions (12×12). A hole is drilled in each division and sealed with a gold-headed nail. The blade is fixed in front of the doors so that one of its two ends precedes the first door by a distance equal to half its width, the other end is then in the middle of the last door. The displacement of the crescent on this blade allows to measure half of the sixth of the hour,or 5 mn (60 :6=10, 10 :2=5).”
2 – The shutters
In each arch that surmounts the door stands a shutter. The latter is a thin square plate of red copper with a much thicker base in the shape of a trapezium. On either side of this base, a pin is fixed, rotating in the holes of the columns to drive the rotation of the shutter. In the middle of the horizontal side of the shutter, a yellow copper rod ending in a golden crescent is fixed. The rotation of the shutter allows the crescent to occupy two positions: one shows the crescent rod standing vertically while the other shows the crescent rod standing horizontally fig.11.
A long spool (al-barbakh) is associated with each shutter.
According to Radwan:
“This spool is formed by two small disks pierced at their center with an opening that allows an iron rod to pass through, around which they can turn. This rod runs along the length of the horizontal bar above the doors. The two disks are connected by a small rod whose two ends are welded on their surfaces near the perimeter. On this rod, equal in length to the width of a door, is welded a hook to which the silk cords of the shutter are tied.”
“To turn the shutter, it is necessary:
a- Drill a first hole, from front to back, at the bottom of the thin plate of the shutter. A second hole starting at the top of the thick base crosses it and exits through its bottom.
b- Pass a silk cord through the first and second holes and tie it to the hook of the spool corresponding to this shutter. The other end of the rope has a knot to hold it in place.
When the door rotates, its counterweight pulls the chain linked to the spool. The latter turns, attracts the rope of the shutter to tilt it and make its crescent straighten vertically.”
The counterweight of 900 dirhams (2.25 kg) in the form of a metal pear ending on its top by a ring where a rod carrying three small rings is hooked. A silk thread is tied to one of these rings, which is attached to the tenons of the doors. To the second ring, the end of the rope of the ball release mechanism is connected. To the third ring, a string is tied to the shutter spool.
A small sled, called the mouse, according to its shape, carries a small vertical index surmounted by a gilded metal crescent. This sled moves on the horizontal plane formed by the ceiling of the counterweight chambers. Its movement is driven by two chains fixed to two rings attached to its body. One of them is wound on the groove of the big wheel of the daytime hours, the other one is wound on the groove of a pulley and ends with a counterweight.
A small copper rod (al-Bazyoun البزيون) attached to the body of the mouse, (fig.12) can unhook the door spike to allow it a free rotation around its vertical axle. This rotation is driven by ropes that link each door to its counterweight.
The metal body of the falcon is mobile around a horizontal axle. A counterweight ensures that the body is upright when tilted. The beak of the falcon, mobile around a horizontal axle, opens under the action of the weight of the ball and closes as soon as the falcon straightens (fig.13). Below the beak of each falcon, a large metal bowl with a pipe is fixed on well-decorated wooden support, giving access to the ball chamber. In its center, a metallic mirror is fixed. The impact between the mirror and the ball makes a sound that resounds to indicate the passage of an hour.
The balls, numbering twenty-four, are empty metal spheres that, once released, move through canals to reach the head of the falcons. A triggering mechanism is capable of releasing two balls per hour, each of which rolls down its own canal.
The main canal has two well-welded ducts. Each canal is connected to other identical canals in suitable directions, leading to the head of a falcon.
According to Radwan:
“To ensure a good rolling, the diameter of the ducts must be slightly smaller than that of the balls. These ducts are seven in number (fig.1.), five main ones to which we add two others for the falcons.”
The release of the balls, two by two, is ensured by a rod carrying at one of its two ends a ring while the other is provided with two half-discs inserted in the conduits of the principal canal. Each of these half-discs imprisons a ball to release it at the desired moment. A rod identical to the one previously described is associated with each door. It is connected by a chain to its counterweight (fig. 10.)
There are thus twelve equidistant rods in the main canal (fig.1.).
Two coaxial wheels form the wheels of the daylight hours. The larger one is a simple, very light crown made of dry wood, on its throat is wound a chain connected to the mouse. Its perimeter is equal to the distance travelled by the mouse for one day. Its metallic axle made of yellow copper is connected to its crown by wooden rods. On the throat of the small wheel, also made of wood but solid is wound a chain connected to the float of the main tank.
“Radwan specifies that this wheel has a diameter equal to a third of that of the large wheel, so that each turn of the large wheel corresponds to a turn of the small wheel” (fig. 12-15).
The night hours are indicated by the twelve openings made in the wooden facade of the clock. These round openings belong to the crown of a semi-circle. Each opening is closed by an opaque glass disk of the same dimensions. A red openwork copper frame, decorated with gilding, is fitted to these openings in order to fix them in their position during the operation of the clock. These openings are called “al-Jamat” الجامات, a kind of small windows.
A light half-disk of wood, the thickness of the width of the phalanx of the thumb, fixed on the axle of the arch of the windows, covers the latter. Its rotation unveils them one by one. Once all the windows have been revealed, the half-disk forms a complete circle with the semicircle of the window crown.
To this same axle is fixed the wheel of the night hours. It is a heavy wooden disc with a peripheral groove on which a chain is wound. This chain is attached by means of pulleys, on the one hand to the float of the main tank and on the other hand to a counterweight (fig.14).
“Radwan specifies that this wheel must have a thickness three times greater than that of the half-disk.”
The definition is given by Radwan:
“…the solar hours are the most important mechanism of the clock. The solar disk moves parallel to the movement of the sun in the sky, it appears with the appearance of the sun and disappears with its disappearance. At noon, this disk is placed in the middle of the zodiac signs … this disk allows us to verify the proper functioning of the clock so that if the sun rises in the sky the equivalent of two hours, we see a single sign of the twelve zodiac signs. And if the sun is in the center of the sky, the solar disk will be in the middle of the circle of the zodiac signs. Then we see three signs and the time passed is six hours…”.
The solar clock is formed by the following elements:
It is a complete circle, having, on its perimeter, a crown divided into twelve equal parts, each part containing a circle indicating one of the zodiac signs. Each sign is pierced by three equidistant holes so that the sun can occupy, respectively, the three positions for one month. This circle has the same perimeter as the wheel of nocturnal hours, its center is pierced by a circular opening of a diameter slightly larger than that of its axle of rotation (fig.18).
It is a thin metallic disk, slightly smaller in diameter than the small circle of one of the zodiac signs, on which is drawn, in black, the sun and its name. A small thin rectangular plaque, with a square hole at each end, is fixed on the perimeter of the sun disk (fig.18.).
In one of these holes is inserted the axle of rotation, while the other is attached by a nail to one of the holes drilled in the perimeter of the wheel of the zodiac signs. The disc of the sun must extend beyond the perimeter of the wheel.
According to Radwan:
“The disk of the sun is always attached to the circle of the zodiac signs by a nail that fits, on the one hand, in the opening made in the plaque of the sun and, on the other hand, in one of the holes drilled on the zodiac signs’ plate. Every ten days you must move the nail successively in each of the holes previously mentioned. In this way the sun moves 10 degrees from one nail to the other, that is 30 degrees per month. The sun appears at the end of the sign towards the end of the month, in the middle towards the half of the month and at the beginning, the first of the month. (fig.18.)
However, it is important to fix the disc of the sun to its axle of rotation so that when the semicircle hides the windows of the night hours, the sun will be below the east of the horizon.”
It is a thin wooden rectangle decorated with gilding, equal in width to the diameter of the circle of the night hours and of suitable length. It is fixed in front of the wall of the clocks at a distance of four joined fingers so that the side of its width is parallel to the diameter of the circle of the night hours. It is in this space that the solar clock must turn with all its elements.
On both sides of this horizon, two well-decorated columns are fixed, connected by a base, so that the openings of the night hours have the appearance of a mihrab in the center of which appears the solar clock. This horizon hides the lower half of the circle of the zodiac signs. Thus, when the sign of Aries appears, its opposite sign, Libra, will be hidden under the horizon (fig. 16.).
The rotation of the wheels and disks described above is driven by an axle with variable sections. This copper axle has two large square sections of different sizes, each of which is followed by a circular section. To these sections can be adapted wheels one after the other to turn on the same axle. This axle passes through the center of the window arch fixed to the wall.
Radwan specifies how this axle should be perfected:
“… It is an empty metal axle of copper of length 1.5 span. Its round and pointed end turn in al-Lubnah fixed in the center of the semicircle. This axle is filed with a square section identical to that of the complete circle so that the end of the axle is raised to fit and turn in al-Lubnah fixed in the center of the helical circle (الدائرة الحلزونية.( By the same means chisel a circular section, another smaller square and a last circular, which follow the first section… Thus, we end up with the other end of the axle also pointed that turns in al-lubnah of the horizon of the solar hours…”
The assembly of the wheels is done as follows:
A: The pointed end of the axle that turns in al-Lubnah made in the center of the helical circle.
B: The large square section that carries the wheel of the diurnal hours and the semicircle.
C: The large circular section.
D: The small square section which carries the sun disk.
E: The small circular section around which the disc of the zodiac signs rotates.
F: The pointed end of the axle that rotates in al-Lubnah of the solar hour’s horizon.
J: The wheel of the nocturnal hours.
L: The nail that attaches the disc of the sun to that of the zodiac signs.
M: the disc of the sun.
N: The holes in the disc of the zodiac signs.
P: the semicircle that hides the windows.
Q: the zodiac signs disc (the diameter of the central hole must be slightly greater than that of the small circular section of the axle).
Note: In the following sentence: “…the pointed end of the axle rotates in al-Lubnah fixed in the center of the helical circle…”, Radwan does not explain how this circle is helical.
According to Radwan’s description:
“Once all these parts are tuned and the clock is assembled, it is enough to connect the chains to start it. To this end, at dawn the main tank is filled to the required level. The valve of the regulating tank being always closed, one turns the index of the curved tube on the degree which is appropriate for this day. The mouse is placed in its initial position; this position must be at a distance, from the center of the first door, equal to the width of a door.
The al- Rub’ valve is opened by untying the wires that connect it to the rectangular pipe. The water flows and reaches the regulating orifice (al-Jaz’a) through the pipes gathered for this purpose. The flow of water, through the orifice of (al-Jaz’a,) lowers its level in the main tank. The large float follows this level, pulls the chain wound on the groove of the small daylight hours wheel and drives its rotation. The large daylight-saving wheel rotates, causing the mouse to move. The crescent of the mouse runs through the nails of the metal blade indicating the time every five minutes. After one hour, the small rod of the mouse (al-Bazyoun) unhooks the spike of the first door to free it. The latter rotates through 180 degrees and presents its back side. The counterweight descends and triggers two movements: The reel rotates, upsetting the shutter so that its crescent rises vertically, and the half-cylinder rod lifts to release two balls, each of which heads towards the head of a falcon through its own canal. The falcons bend over, spit out the balls and straighten up under the action of the counterweights. When the ball hits the metallic mirror fixed at the bottom of the bowl, a deep sound is hear, to indicate the passage of an hour.
The water flow continues, the mouse follows its path, and the same phenomenon is repeated every hour.
At nightfall each door will have changed face and each shutter will have straightened its index finger vertically and all the balls will have been discharged.
To read the night hours, we disconnect the chain of the mouse connected to the float and we hook it to the chain of the nighttime wheel. As soon as the water flows out of the regulating orifice, the float of the main tank moves, causing the rotation of the nighttime wheel, the sun plate and the half disk. The latter reveals at each hour a window illuminated by the lamp. At the end of the night, all the windows are illuminated
Radwan points out that every five days al-Jaz’a must be moved by one line from the six traced in each sign of the zodiac.
At daybreak, the sun appears occupying its position on one of the zodiac signs.”
Radwan mentions the following:
“Concerning the rotation of the semicircle, the rope of the nighttime wheel should be wound in the following way: A rope is taken from a large, excessively long, well-twisted thread, the tenacity of which is checked by attaching a heavy counterweight to it. The middle of this rope is tied to a nail fixed at the end of a horizontal diameter of the nighttime wheel (fig. 18.). The first half of this rope will be wound on the groove of this wheel by making it pass through the upper half of the perimeter while the other will be wound in the opposite direction. By this winding, when the wheel turns each half of the rope makes the semicircle turn in its own direction.
To be more precise, if one end of the rope is attached to the float and the other to the counterweight, and the clock is running, the semicircle, driven by the rotation of the wheel, will turn from left to right to cover or reveal the windows. However, if the end of the other half is hooked to the float, the semicircle turns from right to left.
By this winding, the semicircle remains in function day and night, during the day the windows will be closed by the semicircle and at night they will be respectively illuminated. If you want to reverse the direction of rotation of the semicircle, you just must fix the nail at the other end of the horizontal diameter of the nighttime wheel.
Moreover, Radwan proposes that the disk of the zodiac signs of the solar clock and the sun plate should be hidden behind the wall of the clock and that a square opening in this wall should reveal the sun in its position on one of the zodiac signs, during the operation of the clock.”
It is in the last chapter of this treatise that Radwan points out these alterations (pages 186 to 197).
Since there are many repetitions, we retain the essential. Moreover, al-Sa’ati begins his chapter with this sentence:
“What is explained in this chapter is already mentioned in the text. I have chosen to repeat it so that we can benefit from it.”
“These alterations affect floats, tanks, water pipes, counterweights, chains, ropes, pulleys, wheels, rails, ball canals, etc.
If a part breaks, it is replaced, but the curved tube must work in the long term, for that it is necessary to take care that the mud does not block it to keep a good water flow. If this happens, it is necessary to disassemble it, wash it and blow strongly inside, even fill the mouth with water and blow into the pipe to get rid of the water residues.
Concerning the chains, you must pay attention to their rigidity, prevent any flexibility and pay attention to the knots that can be formed. Similarly for the ropes, it is necessary to be careful of their extension. To do this, simply draw two parallel lines, one on the rope and the other on the opposite wall. The next day observe their parallelism, if it does not coincide it is necessary to repair this extension. If the ropes harden, they must be softened with soap.
The tanks should always be monitored for water leaks and the welds should be reinforced. The rotation axes of the wheels must be lubricated to ensure a good rotation. When making wheels, choose old wood and let it harden before using it. The wheels can become heavy with the accumulation of dust, they must be brushed.
For the mouse, it is necessary to make sure that its rail is smooth, its trajectory free and its crescent is neither too heavy nor too light.
The lubrication of the door and shutter tenons is very important.
The round shape of the balls must be perfect and their canals shallow to facilitate their rolling.
The semicircle of the night hours must be well fixed to the wall so that it does not fall over during the operation of the clock. The crown of the nighttime wheel must be smooth and narrower than the diameter of the string to ensure proper rotation.
The counterweights must be well adjusted, neither too heavy nor too light.
The winding of the two halves of the rope must be reversed on the groove of the nighttime wheel, otherwise the half disc that reveals the illuminated windows will work in one direction only.”
Radwan also mentions the manufacture of each element while specifying its dimensions, the nature of the material used, the way it is handled. He does not forget to mention the details of the ornaments needed for the clock.
In conclusion, in light of this detailed and precise description, we cannot ignore Radwan’s talent as an expert technician in the field of clock-making mechanics.
Description of the old units of measurement adopted, based on the hand:
1 span varies between 20 and 22cm
Width of a finger varies between 1,85 and 2 cm
A phalanx varies between 2 and 2,22 cm
Other units of measurement
Several values have been adopted for 1 dirham; 1.5g, 2.5g, 3g
In our study we have used the following measurements:
1span = 22cm
Width of a finger = 2cm
1 dirham = 2.5g
1 ratl = 1,5kg (in the Middle Ages, among the Arabs)
N.B.: All the figures taken from the treatise of Radwan al-Sa’ati are annotated.
1- Rihlat Ibn jubayr, Dar Sader edition Beirut-Lebanon 1985 p. 243
2- B.R. Sanguinetti and C.Defremery, “Voyages (Arabic text followed by a French translation)” Paris 1853, vol. I, p.209
3- H.Suter, “Die mathematiker und Astronomer der Araber und ihre Werke”, Abhandl.zur Gesh. Der mathematishen Wissenschaften, 10(1900)
4- George Sarton, “Introduction to the History of Science” (Baltimor,1927-1947), 3vols.
5- Ed. A.Muller Ibn Abi Usaybi’a AU, “Uyun al-anba”, 2 vols (Cairo, 1882)
6- E.Wiedeman, .et F. Hauser, “Üer die Uhren in bereich der Islamischen Kultur” (Hall 1915. 1-272)
7- Donald R. Hill, “Arabic Water -Clock -the University of Aleppo” -Institute for the History of Arabic science Aleppo, Syria 1981.
8- Fuat Sezgin, Eilhard Wiedemann , Gesammelte Schriften, “Zur arabisch-islamischen Wissenschaftsgeschichte Gesammelt und bearb, von Dorothea Girke Frankfurt: Institut für Geschichte der Arabish-Islamischen Wissenchaften”, 1984, 3 volumes.
9- Abdel Aziz al-Jaraki. “From Frankfurt and Cairo to Damascus: Recent Models of the Umayyad Mosque Clock.” https://muslimheritage.com/recent-models-umayyad-clock/ Published on: January 6, 2011
“When Ridhwan al-Sa’ati preceded Big Ben by more than six centuries”
https://muslimheritage.com/ridhwan-al-saati-anteceded-big-ben ) Published on: 6 January 2011
10- Ridwan b. Muhammad al-Sa’ati, “Ilm al-Sa’at wa al-‘Amal biha”, edited by Muhammad Ahmad Dahman. Damascus: Maktab al-dirasat al-islamiyya, 1981″.
In this study, we have limited ourselves to the description of the elements of the clock without going into too much detail about the construction. However, four passages are worth mentioning.
We present here a literal translation of the original text:
On several occasions, Radwan mentions in the text the curved tube الأنبوب المعوج without, however, specifying the shape of the curvature of this tube. The latter is reported in the following pages:
“I invented the curved tube reported inside, so if al-Jaz’a is turned on the zodiac signs mentioned at the upper half of the disc, this tube will be directed vertically upwards, and the flow of water is low. However, if al-Jaz’a is turned to the opposite signs the tube will be directed downwards and the flow of water is stronger…”. (page 16)
“… the large pipe, the female, glued to the plate, the male pipe connected to the disc of al-Jaz’z and the curved tube”. (page 17)
“… the tube of al-jaza’a is the one that is attached to the surface of the disc of al-Jaz’a and coming out of this surface to the curved tube …” (page 19)
“… The curved tube is the tube glued to the back of the disc of al-Jaz’a, one of its two ends is connected to the male pipe from which it receives the water. The other end is connected to the pipe of al-Jaz’a…” (page20).
“…then you make the curved tube. It is a tube of length equal to half the diameter of the disc of al-Jaz’a or a little less. It is slightly larger in diameter than the male pipe. Its end connected to the male pipe has an opening that coincides with the opening of the male pipe so that the water can flow freely. A side opening will be made close to the other end. The curved tube is welded neatly together and lead is added to strengthen it. Then you get the male pipe in a vertical position in the center of the al-Jaz’a disk and the curved tube runs along the radius at the back of the disk whilehaving one end connected to the male pipe to let the water through and the other end releases the water to the al-Jaz’a tube…”. (Pages 55-56)
“My father invented the helical carrier of the daylight hours”. (page 13)
“…then make the carrier of the semicircle of the night and its complete circle from the inside and this in several ways. What my father invented is to make a circle identical to the circle of the night of the same thickness, in the center of which you fix a square, hollow metal piece (al-Lubnah) in which the end of the axle turns. This piece will be pierced with two helical holes in front of two helical holes made in the complete circle. Insert the end of the axle into this piece, then fix the two sets with wooden spiral nails so that they form one piece” (fig-19). (page 108)
“If you want, you can make an alidade of wood similar to the one of the astrolabe, it is rectangular, with square sides, of length equal to the width of the twelve doors. Its ends, like the pegs of the astrolabe alidade, are curved and fixed to the wall of the hours. Fix in the center of this alidade, just in front of the axle, al-Lubnah. Also, in front of this axle, fix a nail in the alidade to which you attach a rope whose other end is fixed to the ceiling (fig-20). This rope is used to raise the alidade to the proper position figure opposite (page 109) or you can make the helix as described above…”
“… Making the (al-Halazoun ) الحلزون : It is a wooden square . It is given a nail that is fixed to the alidade in which is embedded al-Lubnah of the axle. This nail passes from the outside to the inside of the alidade. Four wires are attached to it, which are connected to the four corners of the square of the (al-Halazoun ) الحلزون. To the latter a rope is attached to the ceiling. So, when you pull the four strings and tie them tightly to the nail, the alidade will rise, and the top of the axle will be well embedded in it. The rotation will be well restored without any disturbance…”.
“… the circle of the (al-Halazoun ) الحلزون, it is a circle which replaces the alidade, of the same diameter as the circle of the diurnal hours, bored on each side of a helicoidal hole letting pass a nail also helicoidal in order to fix it to the square of the (al-Halazoun ) الحلزون...”. (Page 36)
The above quotation do not confirm that the hollow of al-Lubnah is milled into spirals or that the (al-Halazoun ) الحلزون , which should be a helicoidal form, is not one. It is simply a circle.
“… the perimeter of the large wheel of the diurnal hours corresponds to the distance covered by the crescent from sunrise to sunset, that of the small wheel is such that if the latter makes a turn, the large wheel also makes a turn…”. (page 24)
“… al-Binkan, its length is six spans plus cinq fingers joined, its length pipe eight fingers joined. The part inserted in the bikam is one and a half fingers, in al-Kayl half a finger, above al-Lunbah two fingers. The remaining free part is 4 fingers. … the perimeter of the Binkam is the length of a string of four spans and five fingers joined. The length of al-Kayl is seven fingers joined, its perimeter is such that a full hand can enter … The large wheel of the daytime hours has a diameter of two spans and seven fingers joined, that of the small wheel is nine fingers joined …” (page 197)
“…the length of the wall of the hours is nine spans and six fingers joined. One removes on each side one span and six fingers joined for the falcons, it remains for the total width of the twelve doors, from the first to the last column, seven spans…”. (page 197)
“…the path of the crescent from the first hour, is the distance traveled from the last column to the origin where al-Bayzoun is at the beginning of the day. This distance is seven spans and seven fingers joined…”. (page 198)
by Dr Mona Sanjakdar Chaarani is a researcher in the history of science and a member of the research group on the scientific Arab heritage affiliated to the CRNS in Lebanon. She was a Professor of physics at the Faculty of pedagogy, the Lebanese University in Beirut, Lebanon.