I have just painted the Pilum and it is drying in the oven,time permitting, I will be able to post some pictures by the end of the week.
The following is a description of the Pilum:
The javelin or pilum consists of a long iron head with a small point, and a wooden shaft. On the most common type, the bottom of the head widens into a flat tang, which is riveted into the widened top of the wood shaft. The second type has a socketed head, and a third type, less well-known, has a spike tang. In the first century AD, some tanged pila are shown with a spherical weight, presumed to be lead, behind the joint block. Apparently the weapon had become lighter over the centuries, and the weight was added to increase its "punch".
Pilum heads are 14" to 30" long, with pyramidal or barbed points c. 2" long. The iron shanks are about 1/4" thick (round) below the point, swelling to c. 3/8" or 1/2" square at the base. The tang is an inch or more wide, and can be rectangular or slightly flaired. (Before forging the tang, it's a good idea to fold up c. 6" of the shank, then flatten, to strengthen the tang and allow more width.) There is no evidence that the points were specially hardened--they penetrate simply due to their shape. Likewise, the iron shank cannot really be described as "soft", it bends on impact because it is thin.
The wood shaft is made all in one piece, generally ash, though oak or hickory are also acceptable. Overall it is 4 to 5 feet long, making the complete weapon 5-1/2 to 7 feet in length. Most of the length (or all of it, for a socketed pilum) is round in section, about 7/8" to 1-1/8" in diameter. At the top of the top of the shaft is the tapered, square-section "joint block", 5" to 8" long. It is slotted to receive the tang, and capped with an iron ferrule or collett which is secured by 2 little iron wedges. (Since the ferrule is also tapered, it works best to allow a little of the wood to project above it, to be splayed out by the wedges.) Two or three rivets hold the tang in place; a socketed head needs only a small nail.
The pilum is a very dangerous weapon. The small point could penetrate a shield and wound the man behind it, or even pierce armor. It has often been said that the pilum was designed to get stuck in an enemy's shield, forcing him to discard it, but of course its primary function was to kill. However, the shield-disabling capabilities of the weapon would be a very visible and important side effect! A shield with a pilum stuck firmly in it would be very cumbersome, due to the leverage of the shaft. An oncoming Roman would be able knock the shaft aside to pull the victim's shield out of place, or simply step on it (if it's close enough to the ground) to rip the shield out of the hand. If the pilum is loose in the hole it has made, the buttspike will dig in and act like a doorstop, possibly very suddenly. There would only be a few seconds from the time the pila hit to the moment when the Romans arrive with swords and shields ready, not long enough to set a shield down and yank out a pilum. One can well imagine the effect of hundreds of pila crashing into a line of barbarians who are just starting to charge. As well as those wounded and killed outright, many men will suddenly be tripping over pila and shields, or trying to stop and back up to remove pila from their shields (or bodies!). This will disrupt the entire formation--the pilum is a charge-breaker.
Finally, no matter what the javelin hit, its iron shank was supposed to bend, if only a little, so that an enemy could not throw it back. When the Romans were finished winning the battle they could gather their pila and straighten them.
Back in the early Republic, c. 5th to 4th century BC, the pilum was made in "heavy" and "light" versions. The light one seems to have been the socketed style, with a long narrow iron shank and a small point, with a socket at the bottom to connect to the wooden shaft. The heavy version generally had a shorter, stouter iron shank with a barbed head, widening at the base into a large flat tang which was solidly riveted into block at the top of the wooden shaft. By about the 2nd century BC or so, the tanged variety also has a version with a longer, slimmer iron shank like the light pilum, though it seems the overall construction was still "heavy". The general concept was to throw the light pila first, probably at a range of about 30 yards, then the heavy ones just before the final charge. The men farther back in the ranks may have held onto theirs at first, and moved up to the front as the men who started there got tired and moved back to rest.
Gaius Marius is credited with a design change about 100 BC. He found that the iron shank was not bending very often, so that the enemy were able to throw the pila back at the Romans. So he had one of the two iron rivets that held the parts together replaced with a wooden peg which would break or shear off on impact, causing the head to flop and making it unusable. After the battle it was a simple matter to replace those pegs. One problem is that on many of the surviving pilum heads from this general era, the edges of the tang are bent to form flanges which essentially wrap around the wooden junction block. So they aren't going to flop if one rivet is missing! But of course few of these can be dated with certainty, and there do seem to be pilum heads with simple flat tangs which would function as the story says.
Cheers
Roc