[Edit May 2016: Peter Gaunt cited this blog post in his review of my book in War in History to show how my thinking had changed, which was nice.]
Over the last three posts, I’ve shown that early modern armies couldn’t move without an adequate cavalry screen, that what was adequate depended on objectives and balance of forces, and that the balance between cavalry in field armies could be affected by small-scale raids. Now I’ll bring it all together, in a post that could be titled ‘how horses won the English Civil War’.
After the Earl of Essex captured Reading in the spring of 1643, his next objective was the King’s capital at Oxford. It took a long time for the army to get ready to move. Most historians accept that disease broke out in the infantry, but this hasn’t quite overcome the stereotype that Essex was always reluctant to advance. The army moved north of the River Thames, and got as far as Thame but no further. After a few weeks there, during which Prince Rupert raided Chinnor and won the fight at Chalgrove that I discussed last time, Essex retreated to Great Brickhill. He didn’t attempt another major offensive until the relief of Gloucester. The campaign had definitely been a failure, but the reasons for that failure have never been analysed very well. It has usually been explained away by stereotypes: either that Essex deliberately delayed because he didn’t really want to fight, or that his decayed serving men were essentially inferior to Rupert’s cavaliers.
I’m going to suggest something different: Essex’s army was operationally crippled by a shortage of cavalry and transport caused by failures of horse supply. This should be obvious because Essex and his senior officers said so at the time. The House of Lords received two letters from Essex on 11 July, one demanding 500 remounts and a further 200 per month, the other saying:
The Enemy being so strong in Horse, and this Army being neither recruited with Horses, nor Arms, nor Saddles, it is impossible to keep the Counties from being plundered, nor to fight with them, but where and when they list; we being forced, when we march, to move with the whole Army, which can be but slow Marches;
The army’s council of war sent another letter on 20 July, which was read in the Lords on 22 July:
thus much we shall be bold to say, that, if a constant Course be not held, that the Soldiers may be duly paid, and better cloathed, and that Recruits of Men, Horse, Saddles, and Arms, may likewise be provided, it will be impossible for us to answer your Expectations, or to discharge the Duties of our Place; whereof we have thought fit to give your Lordships timely Notice, that we may not hereafter have it laid unto our Charge, that we have not dealt faithfully, in concealing that which in the End (and that too soon) will be the Destruction and Overthrow of this Army, if speedy Course be not taken to supply these Wants, and prevent our further Weakness
On 31 July, the Lords heard another report from Essex:
The Number of the Horse Two Thousand Five Hundred (Three Thousand last Muster) occasioned by Loss of Horses upon hard Duty and Service, Recruits of Horse though often desired not performed
Ever since my PhD viva, I’ve been fighting against the prejudice that this was just special pleading because Essex didn’t really want to fight. Compare this with how readily historians other than Malcolm Wanklyn have accepted Cromwell’s statement that his cavalry were unable to fight after the second Battle of Newbury. I don’t think there’s any good reason to automatically disbelieve either general. I’ve come to this conclusion through years of detailed research on the practical problems of getting and keeping enough horses. My research corroborates what Essex always claimed: he really didn’t have enough horses (see my War in History article for a summary, and my book for more details). Tom Crawshaw’s PhD thesis has added even more weight to this argument by proving beyond doubt that the finances of Essex’s army were declining in 1643. This makes Essex’s motives and intentions irrelevant. Whatever he really wanted, it was impossible for his army to advance any further. Essex’s political rivals accepted that more cavalry were needed, even though they wanted someone else to command them. When the newsbook The Parliament Scout reported Essex’s problems in July 1643, it also asked why Essex didn’t raise ‘a flying Army that might have forst them to fight’ (Thomason Tracts, E.60, p. 18). It’s seems likely that the editor, John Dillingham, was a client of the Earl of Manchester and was preparing the way for a scheme that Parliament actually started to implement a few weeks later: to raise a new force of 6,500 cavalry to be commanded by Manchester (A & O, i, 215–19). This helped the Eastern Association to become very strong in cavalry, but didn’t help Essex. The defeat at Chalgrove probably helped to persuade him that he couldn’t rely on his cavalry screen because of their inferior and declining numbers, and so had to move his army further away from Oxford. His army had to be reinforced by units borrowed from local forces before he could relieve Gloucester.
There was a big change in the spring of 1644, when Essex’s army spent £9,000 buying 1,000 troop horses from the biggest horse dealers in London. The mistake I made in my earlier work was to assume that this was a sign of a general improvement in the army’s finances. Tom Crawshaw’s PhD thesis has conclusively disproved that. What it really means is that horses had become a much higher priority for spending even though money was shorter than ever. I think this shows that whoever was in charge of spending decisions had learned from the mistakes of the previous year, recognising that horses were operationally indispensable and that buying them for cash was now the only reliable way to get them. Finding the cash remained the biggest problem. The army couldn’t keep buying horses at the same rate, and the ones they did buy were only just enough to cover attrition. As we saw in part two, Essex could only muster 2,785 troopers in July 1644, including a regiment borrowed from the London militia, but this was initially enough to achieve some objectives in Dorset and Somerset because the balance of forces was quite favourable.
Everything came right with the reforms that created the New Model Army in the spring of 1645. Its planned establishment included 6,000 regular cavalry troopers. This was later increased by 600 when Edward Rossiter’s regiment was tacked on, although this unit was anomalous and not entirely in or out of the New Model Army. It wasn’t supplied with horses in the same way as the rest of the army but did happen to be attached at the battle of Naseby. The main army’s establishment seems to have been achieved quite easily by merging the cavalry of the three old armies. The horse supply situation was better than ever because the army had good finances and the right spending priorities. A constant supply of remounts came in from the same group of dealers that had supplied Essex’s army in the previous spring. At the same time, a more effective quota system ensured that all of the artillery train’s draught horses were owned by the state right from the start, and eventually replacements were bought from the dealers in the same way as the cavalry remounts. Although this was a sign that Parliament had finally achieved resource superiority over the king, it would still be a mistake to see Parliament’s victory as attritional. It’s not just about the resources that Parliament could afford to lose. It’s also about what those resources allowed Sir Thomas Fairfax to do.
The New Model Army was much more mobile than Essex’s army had been, because it had enough in-house transport and even more because it had a huge cavalry screen. This allowed Fairfax to move whenever and wherever he wanted, force battles when the odds were in his favour, and avoid them if they weren’t. The royalist decision to fight Fairfax at Naseby on 14 June 1645 is surrounded by controversy and confusion over whether the King was persuaded by Rupert and the military men, or by less experienced courtiers (Wanklyn and Jones, Military History, 244; Foard, Naseby, 191; Gentles, New Model Army, 55). I agree with Wanklyn that because Fairfax was so close and the only route for a retreat was through open country, ‘Rupert therefore had no choice. He had to fight.’ (Wanklyn, Warrior Generals, 162).
Part of the New Model Army was detached to Somerset to relieve Taunton and prevent Goring’s 3,000 cavalry from joining the King. Despite this, Fairfax had nine regimsts of regular cavalry and kept getting stronger as reinforcements sent by orders of the Comittee of Both Kingdoms arrived. John Fiennes’s Oxfordshire regiment had been operating with the New Model for several weeks despite not being administratively part of it; Cromwell turned up early on 13 June with several hundred cavalry from the Eastern Association; Rossiter’s regiment arrived from Lincolnshire on 14 June before the battle started; some other local forces were on the way but didn’t get there in time to fight (Foard, 187, 198, 200). Secondary works give various figures for the strength of both sides at Naseby (Gentles, 55; Wanklyn and Jones, 246; Foard, 199-200, 203-4). Taking their opinions into account, I reckon about 5,000 is the maximum for the King and the minimum for Fairfax. This is the smallest possible disparity,and depends on the double standard of taking the King’s figure from the work that gives the highest estimates for everyone and Fairfax’s from the work that gives the lowest estimates for everyone. Each work agrees that the royal cavalry were outnumbered, and it would only get worse as more reinforcements came to Fairfax but Goring didn’t come to the King (and if Goring had been able to link up with the Oxford army, Fairfax could have had at least another regiment of regular cavalry back from Somerset, and maybe some more local forces too). At the other extreme, it’s possible that the King’s cavalry were outnumbered by as much as two to one.
Whatever the true figures are – and they’re somewhere between the two extremes defined above – it would be very hard for the King’s army to march away safely once Fairfax was nearly on top of them, and I suspect that inadequate numbers of cavalry probably contributed to the scouting failures that allowed Fairfax to get so close without being detected. The smaller army was not more mobile. There was at least a slim chance that the King’s army could win a set-piece battle, because battles were always unpredictable. Although the New Model ended up destroying the King’s army at Naseby, at first it wasn’t quite as easy as the disparity of numbers might suggest.
Fairfax’s march into the west following Naseby was not quite as impressive as we might have been led to believe, considering that there was no-one to stop them. And an army with so many horses to feed had to keep moving even if it didn’t want to. When the army got into Somerset, the advance guard overran Goring’s rearguard at Langport, turning a planned orderly retreat into a rout. Again this was down to being able to move quickly and take away the enemy’s freedom of movement. Once their screen had been beaten in, Goring’s men were absolutely right to panic.
At last I’ve arrived at an explanation for the result of the English Civil War that doesn’t rely on economic determinism, Great Man history, or essentialism. The operational level is the missing link between battles and resources. Wanklyn and Jones were so near yet so far. Parliament did win the First Civil War because of cavalry, but because of superior numbers at the operational level, not because of superior tactics on the battlefield. This is why horse supply is so important. And that’s how I should have ended my book.