Battle of Graham’s Town continued…

PART  II of III

The battle

The geography of the battlefield is such that on the eastern side of what was the edge of the embryo village of Graham’s Town, a gentle slope leads to a small stream known as the Blaauwkrantz River. On the other side of the stream is a plain where Willshire deployed his forces.  2 km to the east of the stream the plain gives way to a ridge, the most prominent feature of which is a knoll, now known as Makana’s Kop. It was along the ridge and the Kop where the amaNdlambe forces massed before the attack. This was the only set-piece battle of the East Cape Frontier Wars before 1850 and the only occasion on which the amaNdlambe abandoned their traditional bush-fighting in favour of open warfare (Peires, 1981 p143).

Table 1 (in Part I) refers. The forces defending the village were under the overall command of Lt-Col Willshire, OC of the 38^th Regiment of Foot. He was regarded as a strict, but fair, and highly capable officer known to his men as ‘Tiger Tom’.

As best can be discerned from Willshire’s account of his troop deployment, he pushed the 38^th Light Company across the river to the point where he expected the thrust of the amaNdlambe attack to be. See Figure 1 (Map) and Figure 2 (Oblique photograph). The Cape Regiment (CR) was extended “along and below the point of a gentle slope from a plain about 800 yards from the town to cover two guns” (i.e. artillery pieces) taken across the river and placed on the open plain in rear of and above the mounted infantry and the 38^th.

Figure 1 A conjectural map of the battle based on Lt Col Willshire’s 1846 article.

Part of the Royal African Corps (RAC) were sent across the river “to remain in support of the guns and extended troops”. Willshire  states that he “therefore left five pieces of artillery at the end of town” so that “as soon as we descended  [i.e. retreated] from the plain into the ravine [donga] to re-cross, those guns would have the [enemy] open to them all across the plain if they followed us”.  The CR were kept “in reserve for those guns in the event of any attack being made on the town from another point.” About 2 km downstream from the position occupied by the 38^th were the East Barracks, home of the CR, on the right bank of the river. This was where Fort England is now located and 60 RAC troops were sent to defend them.

Willshire estimated that the amaNdlambe force was not in excess of 6 000 warriors, although most writers of secondary sources place it much higher. Stretch (1876), for example, gives a figure of 9 000. In his report to Earl Bathurst, the Secretary of State for War and the Colonies, Somerset inflated the figure to at least 10 000, almost certainly as leverage for obtaining more British troops for the colony from a parsimonious treasury (Somerset, 1819 p193; Scott, 1973, p118). Fraser stated that anything above 5 000 was an exaggeration.

The warriors were each armed with 8-10 throwing spears which, if well thrown, had a maximum range of a little over 60 metres in the hands of a skilled thrower according to Tylden, (1952, p136). The hafts could be broken should the need for close in-fighting arise (Barrow, 1806, p414). In addition, many of the warriors had a body-length oval ox-hide shield which a skilled operator could also use as a weapon, but which provided no protection against bullets.  Both Willshire and Fraser claim that several of the warriors had muskets which, if they did, would probably have been obtained from traders or deserters. Willshire also makes a passing reference to ‘a deserter’ assisting them, which Somerset translates into “the plan was formed and directed by certain deserters of the [Royal] African Corps” (Somerset, 1819, p201). The amaNdlambe and their allies were under the overall command of either Makana or Mdushane, the eldest son of Ndlambe, which would have been the tradition. Historical records, such as they are, are ambiguous about this. Sources also vary as to who was leading which part of the battle.

The weaponry available to the British forces consisted of the muzzle-loading, smooth-bore flint-lock musket known as the ‘Brown Bess’, which had a 15 inch bayonet for hand-to-hand fighting.  Its effective range was about 70m (Tylden, 1952, p136) and a well-trained soldier could fire two to three shots a minute, though not for continuous periods. Willshire is obscure about the artillery which he had available. His descriptions have been variously read as between three and seven pieces. Tylden, (1952, p136) suggests five, possibly a combination of 3-pdrs and 6-pdrs, under the command of an officer of the Royal Artillery.  It is nowhere stated who actually manned the guns during the battle.

When the amaNdlambe advanced down the hill onto the plains below at 13h30, they were organised into three divisions: about 1 000 warriors had been sent to attack the East Barracks and the other two divisions launched a frontal assault on the eastern side of town itself, apparently where Willshire had expected them to.  (See Figure 2 – an oblique photograph of the battlefield.)

Figure 2 An oblique aerial photograph showing the slopes down which the amaNdlambe attacked and retreated.

Willshire then directed his troops to advance and open fire on those in front to induce the mass of warriors to move down to their support and get them within the range of the two artillery pieces placed across the river. Amidst war cries they “rushed down to the troops, a short distance, in masses and then spread into clouds covering the hill as they ran.”  Willshire describes this charge in terms such as “a most determined and well arranged attack” and that their determination “to do as much mischief as possible was wonderful”. The amaNdlambe came to within 30-35 yards of the troops who only then retaliated with disciplined volley fire from the muskets and possibly case-shot or canister from the artillery.

Willshire records that he saw “immense numbers” trying to outflank him on only one occasion, to the right of the 38^th, and that he moved the RAC from reserve, bringing them forward into line with the 38^th and the CR’s mounted infantry from where they opened “a well-directed fire and completely stopped [them] from proceeding though they would not retreat till I ordered the advance to sound, when the soldiers cheered, and strange to say, [the amaNdlambe] began retreating pursued by the troops: but they ran so excessively fast that the men were not able to keep up with them”. Not wishing to be outflanked and wary of an amaNdlambe reserve rushing to get in their rear, Willshire sounded the retreat and brought his troops back to where the guns were.

There is no indication of any close-quarter fighting taking place in the direct attack on the town, the disciplined volleys of musket fire and possibly the two cannons keeping the amaNdlambe at a distance. These warriors never got to use their spears, although there were unconfirmed reports that, after the battle, many were found with their hafts broken in anticipation of hand to hand combat. According to Fraser (1952 p139) many dead warriors were also found with their full complement of spears still clutched in their right hands. The reports nevertheless suggest that the defenders had been hard pressed.

Willshire does not give any details of the fighting at the East Barracks, but Fraser states that the fighting was fierce, some warriors even getting into the barracks square which suggests that there would have been some hand-to-hand combat. Neither Willshire nor Fraser mention a hunter named Boesak possibly intervening in the battle, as alleged by Stretch and some subsequent writers. Independent sources however suggest that a small incursion by a group of hunters, towards the end of the battle cannot be entirely ruled out (Pringle 1834).

At around 15h00 the rank and file of the amaNdlambe began to waver and lose their determination, and then to retreat, thus turning the tide of battle. Initially this was in the main attack while “the firing still continued at the barracks”.  Soon after, it seems that the attackers at the East Barracks also withdrew. By 15h30, says Willshire, the amaNdlambe “were beaten in every direction and retreated.” Whether this was by command or by recognition that with mounting losses they were making no headway, or by a spontaneous sense of defeat, the records make no comment and we have no idea. There was no pursuit due to a lack of horses as there was a severe outbreak of horse sickness at the time.

There are disagreements about the circumstances of the withdrawal. Stretch, and those who copy him, claim that it was a rout.  Neither Willshire nor Fraser concur with this. Quite the contrary, it appears that it was sufficiently orderly for the amaNdlambe to make off with 1 000 head of cattle, mainly those belonging to the CR soldiers, a point bemoaned by Fraser, an officer in the regiment (Fraser,1819, p140; Malherbe, 2012, p78). Moreover Willshire, presumably expecting a counter attack, only withdrew into the town at dusk where he “placed the troops and guns at the necessary points for its defence, and who remained at their arms all night”. No further attack came although the British forces were on edge for days afterwards.

 Casualties

Among the defenders of Graham’s Town, casualties were minimal, probably all from the East Barracks. All accounts agree with the official figure which gives three killed (two Cape Regiment men and one from the Royal African Corps) and five wounded. Eight others, including a woman and child, five soldiers and a herdsman were killed in the vicinity of Graham’s Town on the day of the battle (Fraser, p141).

 The amaNdlambe casualties of the battle are difficult to determine with any accuracy as the accounts vary widely, particularly those given for the number killed. Willshire (in Somerset, 1819) reports between 700 and 800 killed and an unknown number wounded – numbers which Somerset also settles on.  Many of the wounded were carried away by their comrades, and some would have died later from their wounds (Fraser, p139). Stretch (1876, p301), without giving any sources, suggests a figure of 2 000 killed. There is in fact no indication of how many would have been killed in the fighting at the East Barracks or how many would have succumbed to the sustained musket volleys and artillery fire at the main point of battle.  There is also no direct evidence of how effective the cannon were and it is possible that disciplined musket fire was more effective.

PART III to follow.

 

 

4. The Battle of Graham’s Town, 22nd April 1819 (Part I)

The Battle of Graham’s Town was one of the most decisive battles in South African history with both short- and long-term consequences flowing from it. Grahamstown, as the modern city is called, is located at 33°18’15.19″S and 26°31’57.94″E in the Eastern Cape Province. See also the first entry in this blog titled ‘An overview of the East Cape Frontier Wars’

The substance of this entry was originally published in the Military History Journal 18 (3) 12-17 (December 2018) and is reproduced here with permission and with some editing.  The entry on the battle is in three parts: Part I outlines briefly, the major literary sources on the battle, the circumstances leading up to it and the immediate prelude to the commencement of the battle. Part II describes the battle itself. It includes a map of the positions and movements of the two armies, and a summary of the casualties. Part III evaluates the battle, the performance of the role players and the consequences flowing from the event.  All references in the text are given at the end of Part III.

PART  I of III

Introduction

The 22^nd April 2019 was the 200^th anniversary of the Battle of Graham’s Town, then a small hamlet of probably no more than 30 simple buildings. In the major encounter of the Fifth Frontier War (1818-1819) a small British force defended themselves against an army of the amaNdlambe clan and some minor allies, 18 times its size.  This was arguably one of the most decisive battles in South African history, along with Blaauwberg and Blood River/Ncome, as it altered the course of events and political geography during the 19^th century.

Two major factors impinge on any description and understanding of this battle: the paucity of reliable source material, particularly primary sources; and the considerable amount of writing (over 40 publications) which constitutes the secondary sources available. Much of the latter is contradictory, over-imaginative, or very close to fiction. This account of the battle has accordingly drawn only on primary or substantiated secondary sources.  Strictly speaking there are only two primary sources on the battle itself, both written by the British commanding officer Lt.Col . Thomas Willshire viz. his 1819 despatch to the Governor, Lord Charles Somerset, and his 1846 article in the Graham’s Town Journal, which are virtually identical.

The much-used and often-quoted 1876 article by Charles Lennox Stretch is neither a primary source nor reliable. He relied on the hearsay of others, some of which was false information. He was neither present at the battle nor ever personally claimed he was, although he was in Graham’s Town a few days later.  Thus, despite Willshire’s laconic style and paucity of detail, this article places greatest value on his account. It is backed up by a letter, a hybrid yet authentic source, written by Major George Fraser to Colonel John Graham the day after the battle. All quotations are from Willshire’s 1846 article, page 2, unless stated otherwise.

Circumstances leading to the battle

The detailed reasons for the amaNdlambe attack on the British military headquarters at Graham’s Town are beyond the scope of this article, but they boil down to long-simmering tensions on the eastern Cape frontier (Cory, 1910; Scott, 1973; Milton 1983; Peires, 2001). These involved extensive cattle theft, raids and counter raids across the border, at that time the Great Fish River, as well as internecine struggles within the amaXhosa polity. This animosity was primarily between Ngqika, the young senior chief of the amaRharabe (a major branch of the amaXhosa people) and his aged uncle, Ndlambe, who had acted as regent until Ngqika became of age. The antagonism between them was both political and personal, involving power and jealousy. While the majority of the clan appears to have supported Ndlambe, the governor, Lord Charles Somerset, had, short-sightedly in retrospect, allied the British with Ngqika.

After a major battle at Amalinde in October 1818, when Ngqika’s forces had suffered a major military defeat at the hands of Ndlambe (Herbst & Kopke, 2006), Ngqika appealed to his British allies for help. This was forthcoming in the form of a major British-amaNgqika raid into amaNdlambe territory in early December 1818, which left many of the amaXhosa east of the Fish River impoverished. The amaNdlambe were quick to retaliate and invaded the colony in late December 1818. This precipitated the Fifth Frontier War, resulting in widespread destruction of property and loss of life. The British and colonial authorities then began to build up forces for a major punitive expedition, a point not lost on the amaNdlambe.

Implicated in all this was a politically astute mystic, Nxele, also known as Makana.   Although a commoner, he had a wide following and was an influential councillor to Ndlambe. He was also well known to both missionaries and the military, among whom his agile mind and debating skills were recognised. The details of his involvement and the role he played in planning the attack on Graham’s Town are not clear although, as a war doctor and mystic, it would probably have been significant. Most writers, and certainly the British, believed he was the mastermind behind the attack and had orchestrated both it and the  general invasion which preceded it.

Prelude to the battle

The attack on Graham’s Town began to take form at mid-morning on 22^nd April 1819. The amaNdlambe had in the preceding days managed to assemble an army of several thousands of warriors within 10km of the settlement without the British being aware of it. At about 10h30 Lt.Col. Willshire, newly arrived on the frontier, was inspecting the Cape Regiment (CR) troop of mounted infantry when he received a report of cattle being taken nearby.

Accompanied by 25 members of the troop, he went to investigate, pursuing the perpetrators some distance before coming across a body of 200-300 warriors who then retreated. Following them, he realised that they were attempting to lead him into a trap and surround him. Willshire attempted to retreat, but was surprised to come upon the entire amaNdlambe army which he estimated at 6 000. He concluded that they intended to attack Graham’s Town, and after sending a message to his second-in-command, Capt. Trappes, warning him of an impending attack, he and his party tried unsuccessfully to delay the advancing force as they made their escape.

As Willshire reached the town at about 11h45, the amaNdlambe army appeared on the hills and ridges to the east of it, about 2 km distant. Although they may have had the opportunity of immediately surprising and overwhelming the town, the attack was delayed for nearly two hours while, according to tradition, various rituals were conducted and final arrangements made. This delay allowed the defenders time to deploy their forces and prepare for the attack.

The village was defended by 333 armed men, the composition of which is shown in Table 1.

___________________________________________________________________________

 

Table 1   Forces defending Grahamstown: 22^nd April 1819

Royal African Corps  *                                   135                 

 Cape Regiment **                                         121                   

 38^th Regt of Foot – Light Company †         45                    

 Armed men unattached §                           32          

TOTAL                                                              333‡   
(after Somerset, 1819: Official report)

*   The Royal African Corps was a ‘Condemned’ or ‘Penal’ regiment composed of generally hardened criminals under a severe disciplinary regime. It reportedly fought with distinction in the battle.  Its commanding officer in Graham’s Town was Lieutenant Cartwright.
** The Cape Regiment, was a Khoi unit of light infantry incorporated into the British Army after 1806 because of their skill as soldiers. They were renowned as good shots and 39 of the 121 were mounted infantry. The unit was often incorrectly referred to as the ‘Cape Corps’, the ‘Cape Light Infantry’ and the ‘Cape Cavalry’.  Their barracks in Graham’s Town, where many also had their families were known as the East Barracks or Witrugkamp. It was one of the two major centres of fighting during the battle. Its commanding officer was Major Fraser.
†   The 38^th Regiment of Foot were regular infantry of the line with a good reputation. Col. Willshire was their Commanding officer.

§   These are variously described as civilians and artificers. Presumably there were gunners among them.

‡    In addition there were also an unknown number of women and children in the village (Malherbe, 2012, p83).

__________________________________________________________________________

PART II to follow.