(To expand on Adrain and Mica's posts slightly):
The individual Schaller pegs come with a complete tab for the mounting screw. This tab is about 7/32 long, and has a single thru-hole for the screw.
Mounting two of these pegs in the A and D positions--at the same distance from the nut--on the cone peghead would force these two tabs to overlap. Alembic grinds the end of the tab off, leaving 1/2 of the thru-hole. This forms a U shaped opening at the end of the shortened tab.
Placing two of these modified pegs together forms a complete thru-hole, cradled by these two 1/2 tabs. A single screw is then used to hold both pegs in place. This is what you're seeing in Dave's photo, above.
Actually, the pegs are attached to the peghead by the shaft ferrules that slide over the string posts and then screw into the peg body. (This is why the ferrules have a hex head.)
The smaller screw through the tab is an anti-rotation feature that prevents the entire peg from spinning in the hole when you tighten the string.
Alembic's manufacturing tolerances are precise enough so that it's often extremely difficult to see the joint between the ground-off ends of the tabs. Over time, however, it's possible for one or both pegs to rotate very slightly (as in, .010 or so) and thus open up that joint very slightly. It shouldn't be a problem as long as the anti-rotation screw itself is still secure in the peghead. The U channel of the ground-off tab is still captive to the screw shaft, so the peg is still prevented from rotating.
To close the joint gap, simply loosen _one_ of the peg ferrules _very slightly_, (it doesn't matter which one) and rotate that peg to close the gap. Then re-tighten the ferrule. You probably won't need to tighten the anti-rotation screw, but check it anyway.
Btw., this mounting arrangement (and the required modification to the peg) isn't unique to Alembic. Any bass builder using a symetrical 2/2 peg arrangement with a straight string path head geometry runs into it, and generally solves it the same way. The string-to-string distance is such that the clearance between the inner edges of the string posts dictates the position of the peg bodies.
This is true of small-body Schallers, Gotohs and any of their clones, but not for pegs like the Fender-style pegs, which are simply too large to be used in that configuration. If you use the larger pegs, you have to go to a fan arrangement for the strings, from the nut to the peg.
There are several schools of thought on the peghead fan: one school says that a having the strings run straight over the nut to the peg provides maximum coupling of the string to the nut, while the other school says that the thrust differential between a straight path and a fanned path are negligible, especially since the string is already changing aspect by the peghead angle.
Note that various Alembic pegheads use different spreads; in general the original feeling was that with a neck-through it didn't make enough difference to matter, but with a set neck or a bolt-on it was more of an issue. Overall, though, it's usually more of an aesthetic consideration, though there are builders who will get quite passionate over the issue. (For instance, amongst boutique 6-string builders, it's a factor in the ongoing Strat vs. Les Paul battle.)
One final consideration: Lateral movement of the nut. A set of strings does not necessarily exert the same tension from string-to-string. In some cases, with a fan path, the lateral tension through the nut is greater on one side than on the other, because the strings on that side are under greater tension. If the nut was floating, it'd slide towards the tighter side, to equalize the tension.
Many builders don't mechanically mount the nut to the neck, but rather use a spot of glue to hold the nut in place. This is usually enough to prevent side-to-side shifting. However, this lateral tension can also show up in the neck-to-body joint of a set neck or bolt-on, causing instability. This is usually trivial, but can sometimes become a sonic issue, leading to dead spots on the neck, truss rods that are harder to adjust than necessary, or, in extreme cases with sloppy designs, shifting of the neck in the pocket.
Maintaining a straight-across path all the way to the peg eliminates much of the problem. What residual lateral stresses there are are negligible on the truss rod, and shouldn't affect a well-designed neck-to-body joint at all.
Until you start getting into 6 or 8-string basses. There, the lateral tension is much more pronounced, so the peghead engineering is that much more critical.
Alembic has totally avoided the problem by physically bolting the nut to the neck. This coupling, along with the laminated neck construction and dual truss rods, essentially eliminates lateral tension as a factor in even extreme multi-string designs. Yes, there is still a side-to-side tension differential, but the support beam structure with the fixed end-point coupling reduces the ratio of lateral to linear stress to a very trivial number, which can be disregarded when doing structural analysis of the string support and harmonic nodal performance.
nic
