Figure 1. Cananor Fort from Gaspar Correia’s Lendas da
Índia (mid-1500’s). National Archives (DGARQ), Lisbon (reproduced
from Correia 1975). North is to the left, right is south.[1]
(Please click on image for larger version.)
The drawing above comes from a book written by Gaspar Correia, who is believed to have travelled out to India in 1512, and to have subsequently returned to Portugal in 1529. He left Portugal once more for India, living out the balance of his life there until he died in 1563. It is not clear when the drawing was done. Some authorities believe that the drawing probably pre-dates 1550.
The most striking object in the drawing is the tall tower or keep that stands over the centre of the fort. This tower is essentially a Medieval keep of a form commonly used in Portugal and indeed Spain as well. It is designed to resist weapons like bows and arrows, but would not be suitable to withstand cannon fire.
This suggests that at the time when the tower was built the Portuguese believed that the possibility of an attack by cannon on the fort was low.
Cannon had been present in significant numbers at sieges in Europe since the 1450's. Although cannon had been in use in small numbers since the 1360's, the first really effective use of cannon against castles had been in France when in 1449 Jean and Gaspard Bureau had used a siege train to reduce sixty English held castles in Normandy within a single year of campaigning, bringing to an end the English colonies in France.
At the time that the Portuguese fort at Cannanore was being conceived, the techniques for the use of cannon to defeat castles had developed very rapidly, especially in Italy after 1494 when King Charles VIII of France who had equipped himself with a state of the art, highly mobile artillery train, and had been able to rapidly defeat the Italian armies.
This dramatic series of campaigns, which changed the balance of power in Italy led to a fundamental rethinking of the designs of fortifications, first in Italy where huge new forts surrounded with earthen banks and angled bastions beginning to be built, and then further afield.
Italian architect Antonio da Sangallo was one of the first of this new breed of military engineers. Between 1492 and 1495 he added octagonal bastions to Castel Sant' Angelo in Rome to provide protection against cannon, whose stone balls could destroy high stone walls with ease. By 1500 many fortresses and towns across Italy, France and Spanish controlled Naples were receiving similarly designed new walls and bastions. [2]
The fort drawn by Correira appears at first glance to be medieval in style, however on closer inspection of the drawing the fort can be seen to have a number of strikingly new features, linked to the development of forts designed specifically for artillery.
For example, the drum shaped tower in the north east corner is surprisingly squat for a medieval tower.
Indeed it could almost be taken for a circular bastion of the type that was often used by German cities when they first adapted their town defences to cannon.
It is interesting to try to compare the fort at Cannanore with other forts in Portugal or in its overseas possessions.
By 1415 the Portuguese had largely defeated the Moors in Portugal, and as a newly assertive country had acheived a balance of power with Spain.
Most of the wars fought between Portugal and Spain, before 1500 had taken place when artillery was of little real importance in warfare. So although there are lots of later artillery forts along the Iberian border, these cannot tell us much about the design of Cannanore's defences.
However, when the Portuguese had first arrived in India they had already had nearly 75 years of colonial experience behind them gained in North Africa and West Africa. So I began extending the area of my search to Morocco, where I have begun to find examples of forts that are very similar in style to that shown in Correia's drawing of Cannanore.
It has proved possible to find several forts in Africa with details showing similarity to Cannanore.
The fort with the greatest similarity is Castelo do Mar at the former Portuguese colony called Safim located in Morocco, and which is called Safi today.
This colony is especially significant for our purposes because the Portuguese control of Safi was of a very limited duration between 1508, and 1541. This enables us to assign a sequence of dates to the features at Safi, and because very similar features are also present at Cananore, we can make an attempt to date the sequence the development of the Portuguese fort at Cananore.
Figure 2. Castello do Mar, Safi, Morocco, showing a very similar
keep to the one shown by Correia at Cananore. [3]
A number of modern authors including André Teixeira [4] and Martin Elbl suggest that the Portuguese typically built their forts in Africa and India over a number of phases, with the first stage being the very rapid construction of a rudimentary fort to secure their position, followed by subsequent upgrades and expansions.
This pattern of development appears to be borne out by the development of the Cannanore settlement and fort.
The very first Portuguese "fort" at Cannanore was built in November 1502. Da Gama had failed establish a trading post in Calicut, which was the best location for trade on the Malabar Coast at that time. Forced by Moor hostility to avoid Calicut he tried trading with Cochin next, where he was able to obtain pepper, before returning to Cannanore, which was also hostile to Calicut.
In order to lighten his ships before the long journey back to Lisbon, he decided to negotiate with the Kolattiri Raja to leave many of his ships cannon at Cannanore. Agreement was reached that the cannon could be buried, and the store would be surrounded by a security palisade and a door fitted with a padlock. The key of which was keep at night by the Raja. Two hundred Portuguese were left behind when Da Gama left on the 28th December 1502. It doesn't appear that the enclosure was used for accommodation, and one must presume that the Portuguese went to live in the nearby Mopila town along the beach to the east.
It is not clear where this first stockade was, and it may not have been located on the site of the current fort at Cannanore.
In September 1503 Don Francisco de Albuquerque arrived with a new fleet at Cochin, and soon began to construct a fort out of coconut tree trunks and earth to protect his settlement and his local ally, the Rajah of Cochin better against the serious hostility of the Zamorin. This fort was the first fort built by the Portuguese in India.
The next major fleet to arrive was that led by Don Francisco de Almeyda which arrived off the coast of India in September 1505 in eight ships. He had set out from Lisbon with twenty two ships and 1500 troops on board, but many ships were lost or delayed along the way. Almeyda had been issued with instructions dated 25th March 1505 specifically tasking him with building four forts at Anjediva Island, Cannanore, Cochin and Quilon.
He left a party at Anjediva on the 13th of September to start work a fort there, and then left for Cannanore, arriving on the 23rd October 1505, where he disembarked Lorenzo de Brito with 150 men and two ships to commence work on the fort, which he named St. Angelo.
It is not clear what the first fort that was besieged by the Raja of Cannanore and successfully defended by the Portuguese from 27th April 1507 for a period of four months looked like, or it's exact extent. [5]
However, there is a vital clue arising from events during the siege.
During the course of this first siege it had been discovered that the original design of the fort had been deeply flawed. The only place that the water supply for the fort could be obtained was from a well located outside the walls of the fort on the side occupied by the attacking forces.
I believe that this original wall was probably aligned along the line of the wall shown to the south (right) of the keep or tall tower in Figure 3 below.
Figure 3. The suggested extent of "fort" built by Lorenzo de Brito
in 1505.
When Lorenzo de Brito was building the fort, it was as an ally of the local ruler, and it is quite possible that it was being built at a trading post, rather than as a full scale fort. The most important thing was to be ready to defend themselves as quickly as possible, and at this first period the most likely enemy was the Zamorin.
By walling off the headland they protected themselves with the minimum of effort. The most likely route that the Zamorin would arrive by was from would be by sea from the south.
In the second stage in the development of forts proposed by Elbl the forts were consolidated and permanent defences were built. I believe this second stage included the construction of the keep and its associated wall immediately to the north of the keep. The tower may have been built to dominate both the entrance door as well as the well which was now securely inside the new wall.[6]
It is not possible to date the keep or tower at Cannanore exactly, but it was almost certainly built immediately after the 1507 siege.
It is interesting to compare the keep at Cannanore with the keep constructed at Castelo Do Mar at Safim in Morocco on the western seaboard of North Africa in about 1512, which is very similar in construction, and this enables us to picture what the tower at Cananore must have looked like, and how it functioned.
Figure 4. Safi Fort from Google Earth showing both the keep tower
and early drum towers fitted for artillery along the southern
elevation.
One significant difference between the circular towers at Safim, which were built shortly after 1508, and those at Cannanore are their respective diameters and construction. The towers at Safim appear to be taller, and to have rooftop positions, and to have been of a smaller in diameter than those built at Cannanore. This suggests that the Castelo do Mar towers pre-date the Cananore ones.
I believe that the towers were constructed because although the defenders intended using handguns or small cannon to defend the fort, that these were still small hand held, or tripod mounted weapons. Notice how the towers in Figure 5 have embrasures set into their walls from which cannon can fire both away from the walls, but also in some cases along the base of the walls.
Figure 5. Towers along the southern face of Castelo do Mar, Safim.
Note the letter box apertures installed for handguns or small cannons. [7]
The towers at Safim are approximately 9 metres across the top of the fighting platform, and were probably equipped with breech loading cannon like the one below.
Figure 6.
Wrought iron breech-loading swivel gun with reinforcing bands around the barrel, cast in Portugal, circa 16th century. [8]
It is probable that the earliest cannon used in the forts at Safim and at Cannanore were taken from ships, and that they were therefore relatively small.
Figure 6. The area of second phase construction after 1507, marked in red.
Note the polygonal gun tower at the shore.
Because it is not possible to attribute exact dates to the various drawings and maps that survive showing Cannanore, it is not possible to prove an exact sequence to the maps.
However, a valuable clue to the dates of these plans exists, when Figure 6 and Figure 7 are compared. Figure 7, fails to include the polygonal tower. Therefore it is probably earlier than figure 6.
Figure 7. Another edition of Antonio Bocarro map from the Livro das fortalezas,
showing Cananore at an earlier period without the polygonal tower. The earlier largely Muslim trading settlement belonging to the Bibi's of Arrakal can be seen across the bay.
The Polygonal Tower is again significant, because it offers another parallel with a polygonal tower at Safi, which helps to date the Correira drawing.
Safi has two fortified areas, the Castelo do Mar on the seashore, and a larger walled town on top of a nearby hill. The Portuguese were able to capture this Arab town, and they proceeded to fortify it in turn.
Figure 8. A photo showing the Polygonal Tower on the southern face
of the walls surrounding Safim town [9]
This polygonal tower at Safim must have been built before the town was lost by the Portuguese in 1541, and it is very likely that it was built to reinforce the earlier walls that had earlier circular towers. The Moroccan forces had by the 1520's acquired considerable numbers of cannon, and the weaknesses of circular and square towers with their concealed zones were becoming apparent.
Figure 9. A diagram from "Firearms and Fortifications" by Simon Pepper and
Nicholas Adams, illustrating the way that both round towers and square towers
have areas (hatched) at their bases which are unable to be covered by fire
from their adjacent walls. [10]
It is possible that this polygonal tower was an attempt by the Portuguese to address the weaknesses of the round towers with their covered zones, that they had previously been building. The covered areas at the bases of these towers would have enabled besiegers to get into cover at the base of towers, where they could shelter from fire and from where they could start to undermine the bases of the towers. The use of polygonal towers, perhaps demonstrates the comparative isolation of Portugal from events in Italy where the angled bastion was being developed at around this time, and which would prove to be the most effective form of artillery defence for the next two centuries.
From Figure 7 above it is possible to see that the polygonal tower was added after the outer wall had previously been built with it's three semi circular towers. So it is likely to date not much later than 1540 to 1545, but it was built after the three cannon bastions were added. When were these three semi-circular bastions and the outer wall built?
Again, it is not entirely possible to be sure.
Elbl writes that the Portuguese forts in Morocco first came under serious attack by cannon in the 1520's when the Sa'dians began to acquire cannon with which to attack the Portuguese. By 1526-9 the threat had become serious enough for the Portuguese to begin to employ Italian military engineers in Morocco to remodel some of their forts.
However, there are clues from Morocco and elsewhere in India like at Shirgaon, which is a small coastal town north of Vasai, that has another Portuguese fort that appears to date from about this same period. Unlike the fort at Cannanore, however Shirgaon was not upgraded at a later date, so that many more of the original early features survive to this day.
Figure 10. Shirgaon Fort from Google Earth.
Shirgaon is interesting because like Cannanore, it also experienced a series of re-constructions in its early period as it evolved from a manor house like structure with two large walled compounds (gardens?), as shown in Figure 11 below, into a fort.
Figure 11. Shirgaon from Livro das fortalezas, Antonio Bocarro,
showing the first phase.
It is quite possible that when the first buildings at Shirgaon were constructed the Portuguese had not gained full control of the region from it's existing rulers, as was also the case at Cananore in 1505.
The first settlements at these coastal settlements were only there at the sufferance of the local ruler's, and therefore could not be seen to present too overtly hostile a style.
After a period of time, and as the Portuguese grew more dominant, there was a matching and equal increase in hostility on the part of the previous rulers and their supporters, towards the Portuguese presence.
It was not until 1541 that nearby Vasai fell to the Portuguese. While it is not possible to assign firm dates to the various phases at Shirgaon, it is probable that the country estate in Figure 11 pre-dates 1520, while the fort shown in Figure 10, which replaced it probably dates from after about 1520.
The fort at Shirgaon was probably originally built as a keep to which was subsequently added a curtain wall fitted with four matching circular corner towers. These towers are approximately 9.9 metres across the upper fighting platform.
The fort then experienced two subsequent re-builds, as first a polygonal tower was added. I believe that this polygonal tower was probably added to Shirgaon at around about the same time as the towers at Safim, and Cannanore.
It is not possible to precisely date these additions, however they were probably added in the later 1530 to 1545 period. The final phase at Shirgaon was the addition of an angled bastion, facing out towards the creek. This creek is now silted up, but it was probably an enclosed anchorage open to the sea at one end when the bastion was originally added.
I am not able to date this bastion, however it very possibly dates from the period when the Portuguese began to loose control of the coast line. This may link construction of this bastion to the arrival of the Dutch and English off the coast in about 1600, with their much more powerful armament.
Figure 12. The polygonal bastion at Shirgaon. Photo by S. Patel.
At some point between about 1510 and 1540, the settlement at Cananore began to attract greater numbers of settlers, who all had to be accommodated within the new fort, or the surrounding area.
It is not clear if these were new Portuguese arrivals, or if they were the offspring of the many relationships that were formed between the Portuguese and local women.
At other later European settlements in India like Madras or Tellicherry, suburbs or shanty towns soon began to form around the European forts made up of displaced persons or migrant workers drawn to the new towns by their relative freedoms from the reach of the local rulers, as well as the economic draw provided by trade, so they may have included Indian families. This may account for the fact that the wall between the older section of the fort, and the new suburb remained in place. The wall and the old tower would have allowed the Portuguese garrison to overawe the inhabitants of the new suburb in the event that trouble broke out in the new town.
Over time a substantial suburb grew up to the north of the fort, and this had then to be defended by additional walls.
Figure 13. The area of third phase construction after
1510, marked in red.
These new walls at Cannanore appear to have been designed to be defended by artillery from the very beginning.
They differ however in one highly significant way from the earlier towers at Safim and Shirgaon. The towers are open to the rear, and the cannon are designed to be placed at ground level.
These developments suggest that these towers are designed specifically from the start for small wheeled cannon, and not handguns. These cannon were also expected to be able to fire more rapidly than had been the case before.
Figure 14. A bronze cannon on an early land carriage. [11]
One significant issue with towers like the ones at Safim would be that when handguns or cannon were fired inside them, the gun room would rapidly fill with smoke. Here at Cannanore, this was avoided by leaving them open to air. The lack of 360 degree cover would have also allowed the Portuguese to clear these walls, in the event that they had been taken over in a successful assault, by firing at them from the old tower.
Accounts of the return to Cannanore in late February or early March 1509 of the Portuguese Viceroy Almeyda describe how he hung at Cannanore, Turkish prisoners that he had taken in Gujarat on the 3rd of February 1509 when he had defeated the combined Egyptian Fleet manned by Ottoman Turks, and that of the Zamorin from Calicut. He then went on to blow other Turks from guns, showering the bits over the Moorish town across the bay.[12]
The Portuguese used the fort extensively throughout the 1520's as a supporting subsidiary base in support of their settlements at Goa and Cochin, and as a base from which to attack the Zamorin at Calicut. 1524 Vasco de Gama who had made the first Portuguese voyage to India returned as Viceroy of India. On his way from Goa to Cochin he spent three days at Cannanore during which time he forced the Kolattiri Raja to had over Bala Hassan, who the Portuguese considered to be a pirate. Hassan was thrown into a dungeon located in the Fort.
The local seafarers suffered enormously from the Portuguese attacks, and their previous trade was greatly reduced as many of their ships were attacked and plundered. The local Rajah's however evolved a system of beacons on the local headlands, so that coast watchers could spot the Portuguese shipping, and thereby warn local coastal shipping of its presence. The local vessels were often able to run for the inlets and coastal channels, so that the Portuguese were unable to capture them.
Throughout the 16th Century a bitter and protracted war was fought all along the West Coast of India predominantly between the Muslim traders and the Portuguese. In 1564 the Muslims were able to defeat the Hindu dynasty from Vijayanagar, and this enabled them to move more extensively down to the coast from Inland. Cannanore was besieged, but managed to hold out. Once the Muslims had retreated the Portuguese are said to have cut down forty thousand coconut trees to punish the local inhabitants.
During 1580 the Portuguese King Henry I died and the Spanish gained control of Portugal. The great local rival and threat to the Portuguese was the Zamorin of Calicut. His kingdom was exhausted by the constant warfare and disruption of the trade to his port, and by 1584 had reached an accommodation with the new Viceroy Mascarenhas.
In many ways this represented the peak of prosperity for the Luso-Spanish settlements, as with 20 years first the Dutch and then the English started to arrive on the coast. As the Netherlands was at war with Spain, and would remain so until 1648, the former Portuguese settlements were seen as legitimate targets of Dutch aggression.
The Dutch established themselves at Vingorla in 1655, and attempted to capture Goa in 1660, but were unsuccessful. In 1661 they then attacked the Portuguese at Cochin, and found allies in the Paliat Achan, and the Raja of Cochin, as well as amongst the Jews resident in Cochin. On the 8th of January 1663, Cochin fell to the Dutch, and shortly afterwards the fort at Cannanore also surrendered to the Dutch.
[1] I am grateful for Zoltán Biedermann who brought this illustration to my attention. Biedermann published a very interesting comparison between Kannur Fort and that in Columbo, called "Colombo versus Cannanore:
Contrasting Structures of Two Colonial Port Cities (1500-1700)" in the Journal of the Economic and
Social History of the Orient 52 (2009) 413-459.
[2] See Simon Pepper and Nicholas Adams, "Firearms Fortifications, Military Architecture and Siege Warfare in Sixteenth Century Siena" published Chicago, 1986, and Christopher Duffy, Siege Warfare, The Fortress in the Early Modern World 1494-1660. Published London, 1979.
[3] Photo by Horto P.
[4] André Teixeira "Fortalazas, Estado Português da India" published Lisbon, 2008.
[5] William Logan, Malabar Manual, volume 1. Page 313.
[6] Martin Elbl writing in "City walls: the urban enceinte in global perspective," page 354 edited by James D. Tracy.
[7] Photo by Rui Ornelas from http://farm1.static.flickr.com/104/261641968_c22e223396_z.jpg?zz=1
[8] Photo by elakramine.
[9] http://silverhawkauthor.com/artillery-preserved-in-portugal_403.html, Photos taken at the
Museu Militar de Lisboa (Portuguese Army Military Museum of Lisbon), Portugal by Harold Skaarup.
[10] Simon Pepper and Nicholas Adams, page 4.
[11] Photos taken at the
Museu Militar de Lisboa (Portuguese Army Military Museum of Lisbon), Portugal by Harold Skaarup.
[12] William Logan, Malabar Manual, Vol. I, page 315.
