People show their wonderful creations, Thought I would show a failure.
Had some success boring bullet molds on the lathe, Thought I would try a cherry,
Made a cherry, heat treated, hand relieved and honed.
Used the method suggested by Deadeye Bly of equal spacers to position the blocks which should work.
Blocks 12L14. Sides machined flat with respect to the pins. Cut slowly with plenty of oil.
Got about .005” offset between the block cavities. The mill is better than that.
Tried twice, same result. Now must make new blocks and recheck settings.
Suggestions appreciated.
Chuck

I try to forget my failures - I could do several pages in a post of what I do that doesn't turn out    so Chuck you have nothing on me.  If you're not making mistakes you're not doing much and you're not learning.  I like the cherry design and the mold looks great.

Your mill is probably far better than what I use most of the time but I'm guessing there's a bit of lash when you change directions and it's pulling out that lash...?  If you haven't already, place an indicator using a mag base - reach as far out on the table and check your lash, with indicator in place you can probably move the end of the table several thousandths , in and out? (see attached - mine moves about .060" and the gib is all the way in - I'd have to weld and fit an extension on mine)   Unless it's as worn as mine, there's adjustments you can make to the gib that will improve this mismatch - if this is your problem.  You can snug the gibs up for cutting the mold and then back them off for normal use.  Sometimes just keeping moderate tension on the table lock is enough. 
With your failure tho, all is not lost! Once you fix the problem, you can go to the next larger bullet.   
Greg

Chuck,
Let me first say that your doing this and your cutter look good. Some years ago I talked to a guy who was selling out on his mold business. He said that the hardest thing was getting the halves to line up because the cutter of course wants to flex and climb because of the cutter force. Which is exactly what I see in your results. Tighten up the gibs and make sure you take the backlash out of the lead screws. 
Now You might be able to save the mold. I suggest, at least consider this, pull the alignment pin then line the cavities up using a gauge pin with a honeymoon fit on the groove and then re- drill the alignment pins. Worth a try at seeing if it would work.
I have never had much success with cherry cutting so I went back to lathe boring all my molds. Grooved bullets are harder but can be done. If you want to discuss how I do it then PM me your phone number and I would be happy to help.

Just a thought,  For some reason that I don't even know I have always locked the spindle on a mill full up unless I would be making a plunge cut.  I have always been afraid of lowering the spindle and then trusting that little lock lever to keep things in place.  There has to be some clearance between the spindle and the housing and I wonder if that lets the cutter walk as others have suggested.  Thanks for sharing the process and good luck!  David

I think some of the mis-alignment is due to cutter flex and some due to backlash from changing directions of the cut. there has to be some clearance of the mill gibs to allow movement. Tighten all table and spindle locks fully except the one moving and snug it up as much as possible. Go slowly with a light cut stopping the table to clear the chips and allow the cutter to release any spring action. This method with a space between the blocks is not perfect but it does allow the clearance of the chips.

Chuck,
Are you cutting one side of the blocks at a time? That's what the set up looks like to me.

If so, I would say that mainly, tool deflection is (maybe a little machine movement too) causing your problem. I wouldn't cut slow, I'd cut as fast as that type of tool steel, will allow, reducing the feed and chip load on the cutter, that way. If it's 32 cal and 0-1, I'd run it at 800 rpm or a little less. That will reduce the chip load and give a little better finish. Also, mark the cutter with Dykem and try a cut on some scrap, to make sure the relief is good on all the cutting edges.

Edit:
Another thing that might help is, it rough out the driving band area of cavity with a 1/4" EM.

Frank

When I said go slowly I didn't mean run the cutter slow but to take a light feed. Roughing it out undersize may also help.

Chuck,
Could you explain why you don't like self centering devises? 

I bought mine new, about 30 years ago. It was made to do shafting and tubing. I put 1/4" steel, against the jaws, then the 1/2" aluminum jaws, capture it. There are two, heavy die springs, to take up any back lash and make for a smooth closure. I used two 1/2" dowel pins (one pressed and one slip fit) as additional alignment, to the two 1 1/4" shafts that the vise, itself runs on.

When I set it up, I put a 1/4" parallel between the jaws and indicate the 1 1/4" wide area that holds the mold blocks. I pick up the edge of the aluminum jaw and dial over to where I want the cavity to be.

When I get close to the jaws closing, I put a .010 feeler gauge, between the blocks and then open them and clear the chips and deburr the cavity. I replace that with a .005 gauge and do the same, before closing on the blocks to finish the cut.

2 jaw chucks, aren't expensive on Ebay and you can make a smaller version, using a 8" or 10" chuck. Use the dowels and die springs but, with the springs mounted on the dowels.

If you get a Buck (adjustable) type chuck, you can also use it on the lathe, to cut cavity's.

Frank