I've just un-split the crankcases. After many happy (not) hours degeasin' I finally was ready to glue th ecrank cases back together, Rather than my old faithful (red hermetite) I followed advice from a pal and used Wellseal for a less obvious job. I had torqued down the mains and was just running around the 6mm bolts at 7ft/lb, when I had gried from one bolt. It was turning with too little resitance, but not little enough to be stripping. I had a nasty feeling that it was going to be a weakened bolt and sure enough ir went
PING 
The temptation to say
ruck it and glue the sod in place was almost too much. But I did the right thing and undid everything and rand round cleaning the uber sticky Wellseal from all mating faces.
Clearly I wasn't gonna find a new bolt that lenghth so I grabbed a bit o 1/4" 316 SS off the shelf and turned the threaded bit down to 6mm, having assertained that the casing holes could accept 1/4". I grabbed a 6mm nut and bored it 1/4" and then TIGGED it to the top of the bar and with a quick faceoff in the lathe I had a new bolt. There was a bit o clenchy sphincter as I torqued it down for the second time, but it was fine proving that the bolt must have been weak.
Can you imagine my joy when I found that the fourth gear on the output shaft wont come across enough to fit the selector fork when it is engaged in the profile

I have absolutely no idea why the bast#*& wont come across. It doesn't go tight but comes to a positive mechanical stop.

I just can't work it out but I really, really don't want to split the crankcases again if at allow possible. I never removed the selection drum from the casing and as I only split the crank cases because shit fell down when the barrels where removed I didn't take the clutch off either, and I didn't take either shafts apart either. Something is clearly wrong but I'm buggered if I can work out what
Any suggestions would be greatly appreciated.
Thanks