So I went out to the range today to conduct my experiment. A few days back I mentioned I segregated the brass that gave high and low velocities in an effort to see if differences in the brass could account for the terrible muzzle velocity results. As I started shooting over the chronometer, the same old random results started to appear. Boy was I disappointed!

The one thing in the back of my mind was that the powder load on the shells was about 85% of capacity. Maybe the distribution of the powder in the shell when fired had an effect. So I started tapping the cartridge on the bench to settle the powder, inserted it into the chamber with the rifle held vertically, then slowly lowered it to horizontal and made the shot. I was stunned. The velocities settled down and I got several very acceptable strings with low muzzle velocity variability.

As a final touch I had loaded some with a larger powder charge that completely filled the case. I fired all of these without any specialized handling and got nice, consistent muzzle velocity data (2087 fps average, spread 48, standard deviation 11). Looks like I have solved the mystery, and had a lot of fun in the process.