In tracking down some issues Arthur Moraes had running jimmy 4.2, I discovered a bug in the manual online. I incorrectly said that pbar was herwig particle ID 75. It is not. It is particle ID 91. 75 is the neutron (and 93 is the antineutron).
The effect of this is that is you run herwig/jimmy with PART1='P' and PART2='PBAR', changing JMRAD(73) and JMRAD(75) has no effect. This is because Jimmy assumes that p, pbar have the same radius, and takes the radius from PART2, which in this case would be JMRAD(PART2) = JMRAD(91), i.e. no change.
If you have PART1='PBAR' and PART2='P', then changing JMRAD(73) and JMRAD(75) would have the desired effect, since the radius used would be JMRAD(PART2) = JMRAD(73), which has been changed.
Hopefully this has not affected you, since it either behaves correctly, or changing the radius has no effect - in which case you would have noticed, I guess!
It will not affect LHC people of course, and it also does not affect anyone who used HZSteer since this checks that the proton and neutron and antiparticles all have the same radius. But if you were setting this by hand then you could have suffered from this bug, I am afraid. Apologies.