Now I should note that the initial response in Italy to the shocking appearance of effective siege artillery was not to immediately devise an almost entirely new system of fortifications from first principles, but rather – as you might imagine – to hastily retrofit old fortresses. But […] we’re going to focus on the eventual new system of fortresses which emerge, with the first mature examples appearing around the first decades of the 1500s in Italy. This system of European gunpowder fort that spreads throughout much of Europe and into the by-this-point expanding European imperial holdings abroad (albeit more unevenly there) goes by a few names: “bastion” fort (functional, for reasons we’ll get to in a moment), “star fort” (marvelously descriptive), and the trace italienne or “the Italian line”. since that was where it was from.

Since the goal remains preventing an enemy from entering a place, be that a city or a fortress, the first step has to be to develop a wall that can’t simply be demolished by artillery in a good afternoon or two. The solution that is come upon ends up looking a lot like those Chinese rammed earth walls: earthworks are very good at absorbing the impact of cannon balls (which, remember, are at this point just that: stone and metal balls; they do not explode yet): small air pockets absorb some of the energy of impact and dirt doesn’t shatter, it just displaces (and not very far: again, no high explosive shells, so nothing to blow up the earthwork). Facing an earthwork mound with stonework lets the earth absorb the impacts while giving your wall a good, climb-resistant face.

So you have your form: a stonework or brick-faced wall that is backed up by essentially a thick earthen berm like the Roman agger. Now you want to make sure incoming cannon balls aren’t striking it dead on: you want to literally play the angles. Inclining the wall slightly makes its construction easier and the end result more stable (because earthworks tend not to stand straight up) and gives you an non-perpendicular angle of impact from cannon when they’re firing at very short range (and thus at very low trajectory), which is when they are most dangerous since that’s when they’ll have the most energy in impact. Ideally, you’ll want more angles than this, but we’ll get to that in a moment.

Because we now have a problem: escalade. Remember escalade?

Earthworks need to be wide at the base to support a meaningful amount of height, tall-and-thin isn’t an option. Which means that in building these cannon resistant walls, for a given amount of labor and resources and a given wall circuit, we’re going to end up with substantially lower walls. We can enhance their relative height with a ditch several out in front (and we will), but that doesn’t change the fact that our walls are lower and also that they now incline backwards slightly, which makes them easier to scale or get ladders on. But obviously we can’t achieved much if we’ve rendered our walls safe from bombardment only to have them taken by escalade. We need some way to stop people just climbing over the wall.

The solution here is firepower. Whereas a castle was designed under the assumption the enemy would reach the foot of the wall (and then have their escalade defeated), if our defenders can develop enough fire, both against approaching enemies and also against any enemy that reaches the wall, they can prohibit escalade. And good news: gunpowder has, by this point, delivered much more lethal anti-personnel weapons, in the form of lighter cannon but also in the form of muskets and arquebuses. At close range, those weapons were powerful enough to defeat any shield or armor a man could carry, meaning that enemies at close range trying to approach the wall, set up ladders and scale would be extremely vulnerable: in practice, if you could get enough muskets and small cannon firing at them, they wouldn’t even be able to make the attempt.

But the old projecting tower of the castle, you will recall, was designed to allow only a handful of defenders fire down any given section of wall; we still want that good enfilade fire effect, but we need a lot more space to get enough muskets up there to develop that fire. The solution: the bastion. A bastion was an often diamond or triangular-shaped projection from the wall of the fort, which provided a longer stretch of protected wall which could fire down the length of the curtain wall. It consists of two “flanks” which meet the curtain wall and are perpendicular to it, allowing fire along the wall; the “faces” (also two) then face outward, away from the fort to direct fire at distant besiegers. When places at the corners of forts, this setup tends to produce outward-spiked diamonds, while a bastion set along a flat face of curtain wall tends to resemble an irregular pentagon (“home plate”) shape [Wiki]. The added benefit for these angles? From the enemy siege lines, they present an oblique profile to enemy artillery, making the bastions quite hard to batter down with cannon, since shots will tend to ricochet off of the slanted line.

In the simplest trace italienne forts [Wiki], this is all you will need: four or five thick-and-low curtain walls to make the shape, plus a bastion at each corner (also thick-and-low, sometimes hollow, sometimes all at the height of the wall-walk), with a dry moat (read: big ditch) running the perimeter to slow down attackers, increase the effective height of the wall and shield the base of the curtain wall from artillery fire.

But why stay simple, there’s so much more we can do! First of all, our enemy, we assume, have cannon. Probably lots of cannon. And while our walls are now cannon resistant, they’re not cannon immune; pound on them long enough and there will be a breach. Of course collapsing a bastion is both hard (because it is angled) and doesn’t produce a breach, but the curtain walls both have to run perpendicular to the enemy’s firing position (because they have to enclose something) and if breached will allow access to the fort. We have to protect them! Of course one option is to protect them with fire, which is why our bastions have faces; note above how while the flanks of the bastions are designed for small arms, the faces are built with cannon in mind: this is for counter-battery fire against a besieger, to silence his cannon and protect the curtain wall. But our besieger wouldn’t be here if they didn’t think they could decisively outshoot our defensive guns.

But we can protect the curtain further, and further complicate the attack with outworks [Wiki], effectively little mini-bastions projecting off of the main wall which both provide advanced firing positions (which do not provide access to the fort and so which can be safely abandoned if necessary) and physically obstruct the curtain wall itself from enemy fire. The most basic of these was a ravelin (also called a “demi-lune”), which was essentially a “flying” bastion – a triangular earthwork set out from the walls. Ravelins are almost always hollow (that is, the walls only face away from the fort), so that if attackers were to seize a ravelin, they’d have no cover from fire coming from the main bastions and the curtain wall.

And now, unlike the Modern Major-General, you know what is meant by a ravelin … but are you still, in matters vegetable, animal and mineral, the very model of a modern Major-General?

But we can take this even further (can you tell I just love these damn forts?). A big part of our defense is developing fire from our bastions with our own cannon to force back enemy artillery. But our bastions are potentially vulnerable themselves; our ravelins cover their flanks, but the bastion faces could be battered down. We need some way to prevent the enemy from aiming effective fire at the base of our bastion. The solution? A crownwork. Essentially a super-ravelin, the crownwork contains a full bastion at its center (but lower than our main bastion, so we can fire over it), along with two half-bastions (called, wait for it, “demi-bastions”) to provide a ton of enfilade fire along the curtain wall, physically shielding our bastion from fire and giving us a forward fighting position we can use to protect our big guns up in the bastion. A smaller version of the crownwork, called a hornwork can also be used: this is just the two half-bastions with the full bastion removed, often used to shield ravelins (so you have a hornwork shielding a ravelin shielding the curtain wall shielding the fort). For good measure, we can connect these outworks to the main fort with removable little wooden bridges so we can easily move from the main fort out to the outworks, but if the enemy takes an outwork, we can quickly cut it off and – because the outworks are all made hollow – shoot down the attackers who cannot take cover within the hollow shape.

An ideal form of a bastion fortress to show each kind of common work and outwork.
Drawing by Francis Lima via Wikimedia Commons.

We can also do some work with the moat. By adding an earthwork directly in front of it, which arcs slightly uphill, called a glacis, we can both put the enemy at an angle where shots from our wall will run parallel to the ground, thus exposing the attackers further as they advance, and create a position for our own troops to come out of the fort and fire from further forward, by having them crouch in the moat behind the glacis. Indeed, having prepared, covered forward positions (which are designed to be entirely open to the fort) for firing from at defenders is extremely handy, so we could even put such firing positions – set up in these same, carefully mathematically calculated angle shapes, but much lower to the ground – out in front of the glacis; these get all sorts of names: a counterguard or couvreface if they’re a simple triangle-shape, a redan if they have something closer to a shallow bastion shape, and a flèche if they have a sharper, more pronounced face. Thus as an enemy advances, defending skirmishers can first fire from the redans and flèches, before falling back to fire from the glacis while the main garrison fires over their heads into the enemy from the bastions and outworks themselves.

A diagram showing a glacis supporting a pair of bastions, one hollow, one not.
Diagram by Arch via Wikimedia Commons.

At the same time, a bastion fortress complex might connect multiple complete circuits. In some cases, an entire bastion fort might be placed within the first, merely elevated above it (the term for this is a “cavalier“) so that both could fire, one over the other. Alternately, when entire cities were enclosed in these fortification systems (and that was common along the fracture zones between the emerging European great powers), something as large as a city might require an extensive fortress system, with bastions and outworks running the whole perimeter of the city, sometimes with nearly complete bastion fortresses placed within the network as citadels.

Fort Saint-Nicolas, which dominates the Old Port of Marseille. The fort forms part of a system with the low outwork you see here and also an older refitted castle, Fort Saint-Jean, on the other side of the harbor.
Photo via Wikimedia Commons.

All of this geometry needed to be carefully laid out to ensure that all lines of approach were covered with as much fire as possible and that there were no blindspots along the wall. That in turn meant that the designers of these fortresses needed to be careful with their layout: the spacing, angles and lines all needed to be right, which required quite a lot of math and geometry to manage. Combined with the increasing importance of ballistics for calculating artillery trajectories, this led to an increasing emphasis on mathematics in the “science of warfare”, to the point that some military theorists began to argue (particularly as one pushes into the Enlightenment with its emphasis on the power of reason, logic and empirical investigation to answer all questions) that military affairs could be reduced to pure calculation, a “hard science” as it were, a point which Clausewitz (drink!) goes out of his way to dismiss (as does Ardant du Picq in Battle Studies, but at substantially greater length). But it isn’t hard to see how, in the heady centuries between 1500 and 1800 how the rapid way that science had revolutionized war and reduced activities once governed by tradition and habit to exercises in geometry, one might look forward and assume that trend would continue until the whole affair of war could be reduced to a set of theorems and postulates. It cannot be, of course – the problem is the human element (though the military training of those centuries worked hard to try to turn men into “mechanical soldiers” who could be expected to perform their role with the same neat mathmatical precision of a trace italienne ravelin). Nevertheless this tension – between the science of war and its art – was not new (it dates back at least as far as Hellenistic military manuals) nor is it yet settled.

An aerial view of the Bourtange Fortress in Groningen, Netherlands. Built in 1593, the fort has been restored to its 1750s configuration, seen here.
Photo by Dack9 via Wikimedia Commons.

But coming back to our fancy forts, of course such fortresses required larger and larger garrisons to fire all of the muskets and cannon that their firepower oriented defense plans required. Fortunately for the fortress designers, state capacity in Europe was rising rapidly and so larger and larger armies were ready to hand. That causes all sorts of other knock on effects we’re not directly concerned with here (but see the bibliography at the top). For us, the more immediate problem is, well, now we’ve built one of these things … how on earth does one besiege it?

Bret Devereaux, “Collections: Fortification, Part IV: French Guns and Italian Lines”, A Collection of Unmitigated Pedantry, 2021-12-17.