Cavalry Operations: Moving an army is difficult and dangerous

[posted by Gavin Robinson, 7:00 am, 7 October 2013]

[Edit May 2016: I no longer think these cavalry operations posts are the best thing I’ve written anywhere, but they are the best blog posts I’ve ever written. I was just sketching out a hypothesis here which still needs throroughly testing against evidence (which I’m going to do in my next project, especially if my current funding application is successful).]

At last I’ve written the series of posts on cavalry operations that I’ve been promising for a long time. There are lots of details I haven’t gone into, especially to do with geography and fodder, but the main point I’ll be trying to make is that the number of cavalry available has a big influence on where, how and whether armies can move. This series will be four posts, all about the First Civil War in England, mostly in the South. I’ll be concentrating on the Earl of Essex’s army because that’s the one I know best, it’s been under researched and often misunderstood, and its campaigns give some great examples of how important cavalry were. Along the way, I’ll keep challenging the myth that Essex’s cavalry were useless ‘decayed serving men and tapsters’. The basic facts of the movements of armies in the civil war aren’t disputed much, or at least I won’t be disputing them the way I have with the received wisdom about tactics. I’m mostly relying on Wanklyn and Jones for these facts, and a few other secondary works for more details of certain campaigns. These posts will try to explain why armies moved the way that they did, and how cavalry or lack of it could limit their options. A lot of this is hypothetical and can’t be strongly proved using the traditional method of picking anecdotes from narrative sources (but what can?), but it works for me.


There are lots of boring practical things that get in the way of moving an army. Although I’m not going to consider them in much detail in this series, geography and the weather always limit the movements of armies. Roads and bridges dictate where they can march. Campaigning is harder in winter than in summer. Assuming that these things are in your favour, moving can still be a trial. In this period, infantry have to march on foot, which is slow. They will usually have to march in long, narrow columns in order to fit on the road. Artillery guns are drawn by horses, and like infantry will march in long columns which are slow even if it isn’t very muddy. Then there are wagons to carry ammunition, tools, and anything else that the soldiers can’t carry themselves. This can include food, but methods of food supply varied and so did the amount of transport needed for it. There were at least two estimates for the establishment of the New Model Army’s artillery train in 1645. I think they probably used the higher one, but even the lower one needed 670 draught horses to pull 30 field guns and 87 wagons. This included 22 wagons to carry musket balls and match for the infantry, and 14 wagons for tools and materials for various artificers. (TNA: PRO, SP 28/144/10, ff. 238-9. The higher estimate is in SP 28/145, unfol.)

Loading the wagons and getting them moving is a tedious but predictable routine provided that the army owns all the transport it needs. English Civil War armies often didn’t. Parliamentary field armies usually had at least some state-owned draft horses, and these could be supplemented with horses hired from private contractors, sometimes on a long-term basis, but there was often a shortfall. This was made up by borrowing horses and wagons from civilians in the areas where the armies campaigned. In 1644, the Earl of Essex’s artillery train had 558 state-owned horses, 179 hired from long-term contractors, and 50 teams (about 300 horses) temporarily impressed (TNA: PRO, SP 28/146, f. 83). Commissaries often had to go out and requisition transport before the army could set off. As well as delaying operations, this could throw away surprise. The scouts reporting to Sir Samuel Luke, the Earl of Essex’s scoutmaster in 1643, could tell that the king’s forces were about to move whenever they took up horses and carts from the countryside (Luke, Journal, ii, 89, 101, 120, 122, 126, 140, 167).


Early-modern armies were generally slow but cavalry were an exception. Horses can move faster than infantry, even when carrying a rider and equipment. We should be careful not to overstate this. Cavalry couldn’t go galloping around the countryside all day every day, because horses get tired. But before railways in the 19th century and motor vehicles in the 20th, horses were the fastest way to move on land. This meant there was always a risk that a long, slow column of infantry, artillery and wagons could be caught by enemy cavalry. Armies on the march are also at risk of being shot at, but this risk varies according to the technology available at the time. Both risks also depend on the type of terrain that the army is marching through. Generally more cover means less risk of cavalry attack and more risk of missile attack. Infantry were usually safe from cavalry attack as long as they stayed in the correct formation, but were vulnerable if they didn’t. This is perfectly illustrated by Sir John Byron’s account of the end of the battle of Roundway Down, after the parliamentary cavalry had run off, leaving their infantry unsupported (Young, ‘Roundway‘, p. 131):

in the meantime my Ld. Wilmot charged their foot with the horse he had with him, but could not break them, and in the charge Dudley Smith was slain, & Lt. Col. Weston, hurt & many others, but when they saw my horse rallied together again before them, & the Lieut. Gen. continuing still in the rear of them, and that the Cornish foot began to sally out of the town, they thought it not fit to stay any longer, they began first gently to march off, their officers marching before them… With that I advanced towards them with those troops I had rallied, & shot at them with the cannon I had formerly taken, their officers thought it not fit to stay any longer, but such as had horses rid away as fast as they could, & too fast for us to overtake them, & the rest blew up their powder & threw down their arms & betook themselves to their heels, our horse fell in amongst them & killed 600 of them, & hurt many more, and took 800 prisoners & all their colours

Infantry on the march were not in the correct formation. If they had time to form up, this would still slow down their march. In this period, most muskets were matchlocks. The matches would probably not be kept lit all the time because it was dangerous, inconvenient and used up a finite resource.

There is a way of protecting armies from this threat, but it doesn’t get much attention from academic military historians, at leas for this period. Early modern military history tends to be split between two views of how wars were decided. The traditional view, recently revived by Wanklyn and Jones for the English Civil War, was that wars were decided by decisive battles, and that these battles were decisive because of the superior ability of Great Captains and/or their soldiers. For the English Civil War, this usually means Oliver Cromwell and his Ironsides. This kind of history does narrate campaigns, but tends to treat them like moves in a board game, without paying much attention to logistics. Wanklyn is good on geography and makes vague allusions to ‘cavalry support’ being necessary on the march, but hasn’t really got to the bottom of how operations worked. The other view, associated with the War and Society school, is that battles are not important because wars are determined by economic resources. These two opinions also happen to coincide roughly with Whig and Marxist views of history. It might be going too far to say that there was a debate between them, because they’re more like two sets of unquestioned prejudices which casually dismiss each other. I tried to critically analyse and synthesise them in my book, making the point that resources are both as important and as contingent as battles, but perversely for a book ostensibly about horses, I failed to emphasise how uniquely important horses were to military operations. I think this is because the battles versus resources split has had an insidious influence on my intellectual development. I always wanted to study cavalry charges, but Frank Tallett and John Childs pushed me into studying horse supply instead. I can now see that this was a false dichotomy: going either way, or even trying to study both at same time, provides no obvious way into the operational level. Even when I recognised that there was a big gap in my knowledge, I kept getting pulled in other directions by what people wanted to read about and debate. Recently I read Tom Crawshaw’s new PhD thesis about military finance, which disproved the main argument of my War in History article (that the finances of Essex’s army improved in the spring of 1644, and that this was continuity with the New Model Army). Suddenly I realised that I was underselling horses by trying to use them as a case study of supply and finance in general, prefaced by a few platitudes about how horses were needed for this or that. With these blog posts, I’m trying to redress the balance by emphasising what Erik Lund has always said: horses are very special.

The cavalry screen protects a marching army from threats. I didn’t fully appreciate the importance of this until I read Erik’s blog, especially this post about cavalry and armoured divisions. The screen consists of cavalry units moving around and ahead of a marching army to deter attacks and fight off any attacks that do happen. When it works properly, the infantry, artillery and wagons can move safely and with minimal delays. When it doesn’t work properly, the army is in danger.

Case study: the Gloucester expedition

In August 1643, the king’s main army besieged the Parliamentary garrison at Gloucester. This city was important because it was at the intersection of the River Severn and the land route between Oxford and South Wales, so it could disrupt communications and resource gathering in an area otherwise held by the King’s forces. On 16 August, parliamentary leaders agreed that the Earl of Essex should take his army to relieve the garrison (Wanklyn and Jones, Military History, 115). This involved a long march across hostile territory. The expedition required extra transport to take supplies to the garrison on top of the field army’s own needs. Civilians in Buckinghamshire had to provide many of the horses and carts, some of which never came back.

An operation like this depended on a good cavalry screen. It’s not certain how many cavalry Essex’s army had at this time. Earlier in the year, the regular cavalry were reported to be down to 2,500, but they were reinforced by the London militia and other local forces for the Gloucester expedition. Wanklyn and Jones estimate that this brought the cavalry to between 3,000 and 4,000, and Day suggests up to 4,500 (Wanklyn and Jones, 116; Day, Gloucester, 93. Wanklyn and Jones claim that the King had twice as many cavalry around Gloucester but I find that implausible). Essex’s relief force set off from Aylesbury on 30 August (Day, 90). The royalist cavalry under Wilmot and Rupert skirmished with  Essex’s army throughout its advance, but failed to stop it from getting to Gloucester (Day, 90-91, 94-5, 100, 103-5). The King had to raise the siege on 5 September, allowing the relief column to deliver supplies (Wanklyn and Jones, 117). Essex then had to get his army home again. He got a head start when he set off on 15 September, but Rupert’s cavalry caught up with him at Aldbourne Chase on 18 September (Wanklyn and Jones, 118). Essex’s cavalry screen mostly fought off this attack, but some cavaliers got through and did some damage to the baggage. Essex had to commit some infantry regiments to the rearguard to make sure that the rest of the army could get away safely. Sergeant Henry Foster, who went on the campaign with the London Trained Band infantry, witnessed some of this action (Thomason Tracts, E.69[15], sig B2r):

Then Collonell Meldrums and Colonell Harvies troops drew up in a body & gave the enemy a very fierce charge, which was performed with as brave courage and valour as ever men did, and then wheeled about to a Regiment of our foot that stood in the reer of them, the enemy pursued them in their retreat, skirmishing one at another all the way:

I don’t see this as a case of Essex’s cavalry being inferior. I’d say that both sides were fairly evenly matched, so that neither could overwhelm the other. Rupert failed to overrun and destroy the main column, but Essex’s cavalry screen couldn’t quite guarantee that the enemy would always be kept away. Essex decided to take a safer but longer route south of the River Kennet. Because of this, the King’s main army got to Newbury first and blocked the road to London (Wanklyn and Jones, 118).

At the first Battle of Newbury, the tactical performance of the parliamentary cavalry was mixed. They successfully repulsed two royalist charges by standing and firing, but were routed themselves by the third and apparently played no further part in the battle. The Parliamentary infantry held their ground despite cavalry charges and artillery bombardment, but failed to break through. Luckily, the King’s army decided to withdraw during the night because of obscure and controversial reasons.

I hope I’ve shown that battles and sieges are only part of the story. The way the campaign went was heavily influenced by the operations of cavalry screens. The relief of Gloucester and the first Battle of Newbury only happened like they did because neither side had overwhelming superiority of cavalry for the whole campaign.


Comments are closed.