In the history of the Roman Republic and Empire there were many sieges, and over the years the Romans developed their own methods for executing such actions. From those that are reasonably well documented today it appears that the Romans were particularly fond of simply assaulting a besieged town rather than waiting to demolish the walls or starve the occupants into surrender, but there are also plenty of examples where machines were employed in the attempt to overcome those walls. Some of these were adopted from the Greeks, but the onager seems to have been an entirely Roman invention. The first reasonably solid evidence for them comes from the third century CE, but there are vague references to one-arm machines for several centuries before that which might be the same machine, so their origins are unclear. The name ‘onager’ means ‘wild ass’, and this is commonly believed to be because of the violent kick of the machine, although other theories also exist. It has been made several times before in this hobby, but all of these models have suffered serious accuracy problems, so we will start by considering this latest offering to see if it is any better.

The first impression is positive, because unlike previous models, this one has its projectile in a sling (rather than a ‘spoon’) which is correct for the Roman period. Another old mistake avoided here is with the wheels – a machine this size would likely crush any such wheels, and anyway the weight would make it almost impossible to move, so they were always built on site as required. The general form looks fine, with the arm inserted into a central torsion-spring, pulled back by levers, and then released to strike a large padded frame, releasing the stone toward the target. This matches the illustration on the box artwork, and all modern reconstructions, except for one crucial detail. The ropes that pull back the arm, giving it the necessary power when released, are always behind the arm. This is obvious, as otherwise it could not pull the arm at all, yet on this model the winch is precisely underneath the attachment point of the arm when the arm is at no more than about thirty degrees to the vertical, meaning it would be impossible for the winch to pull the arm any further back than that. If this were the case with a real onager then it would be virtually useless, and there would be little force in the release, but of course the Romans were excellent engineers, and so never built a machine like this. We cannot know, but it looks to us like the sculptor saw the general design, but then shortened the whole machine, completely forgetting that by doing so they were largely destroying the basic physics that made the machine work. Now of course this is not a working model anyway, yet this error is really glaring and looks bad even on a static model.

In more general terms the model suffers from the same problem that afflicts all models made in this relatively soft plastic – the parts are quite soft and imprecise, making it an unsatisfying kit to put together. We found that all the holes had to be drilled out before they would accept the intended pegs (you could also trim down the pegs, but we find it easier to drill the holes out), and the relatively small amount of flash on the parts only adds to the vague fit. The central arm is also used as part of the sprue, so has many more connections that need to be trimmed, and it does not actually insert into the spring. Instead it just about reaches the spring, so you have to come up with your own way of fixing it, which is not good enough. The location of the parts means it can only be put together as shown above, which means the arm is in an impossible position as we have said. The lever to the right of the winch is inserted, and we assume the handle to the left is the lock (and therefore release trigger) for the ratchet (the other lever is in the hands of one of the crew). In total this is a poor model from a technical point of view even without the fundamental design flaw.

This machine stands 35 mm to the top of the frame, and covers a base of 75 mm by 41 mm. It is assumed that such machines were made in many sizes, and this one seems as good as any, but to serve such a machine would have probably needed twice the four men this set provides. The poses we do get include two men working the winch, one of which holds the second lever. One man seems at first glance to be pulling on a lever while the other is not engaging with the machine, but to turn the winch it would be necessary for force to be applied equally to both sides, so the man 'pulling' the lever cannot be doing so. Instead, he must be either inserting or withdrawing it from the winch, and since the arm is already fully pulled back and the ammunition loaded, he must be removing the lever. The third pose is of a man holding a stone, and is fine except that we did wonder exactly how heavy such a stone would be (it is 5 mm, or 36 cm, in diameter), and would one man be able to hold it with such apparent ease? Lastly we find what looks like the man in charge, and he at least is an easy pose to accept. As a token crew this is not too bad (a comment equally applicable to many artillery sets of course), but in reality such a machine would have required many more men.

Most of the crew wear segmented armour, with only the commander wearing a mail cuirass, so these have the feel of first or second century Romans, which is echoed in their helmets. This tends to be the most popular period of the Roman Empire, so while we cannot be sure if the machine existed at this time, it is reasonable to guess that it might, and so easy to understand the choice. The clothing and armour all look authentic, and there was nothing particularly distinctive about legionaries that worked such machines, so these are correctly attired. Their sculpting is in the style we have become used to from this manufacturer, with nice, well-detailed figures that are a bit chunky but quite serviceable. The second man has a strap for the sword on his right side, but has no sword or scabbard, so just a strap leading nowhere.

The missing scabbard is clearly a problem during manufacture of the mould, as surely no sculptor would make such a mistake, but on our review example we found a much more serious problem – one of the sprues had not been properly filled. The one shown above was okay, but the second in the box was short-filled to the extent that the machine could not be put together, which is a fundamental problem that should have stopped the product from being sent to market like that.

Despite the better research apparently done when making this set, the basic physics errors in its operation badly detract from the overall impression, and the frustrations in building it do not make for an enjoyable experience for the modeller either, even if your copy of this set is properly filled. The figures are better, and a machine this size is certainly eye-catching on any diorama or table-top game, but for the purist this is another unsatisfactory attempt to model what must have been quite a fearsome weapon for anyone hoping to shelter behind their walls.