I received the following info via the TankNet group (isn't the web wonderful?):
"By Rexford 26 March 2001, 06:43
WW II tests of Cal. .50 AP M2 are included in TM9-1907, which indicates that 350 Brinell Hardness Number plate was used. This is important because armor hardness effects can strongly influence small projectiles like MG bullets.
WW II data in TM9-1907 is based on Navy Ballistic Limit, 50% of hits result in a success, where success is judged on basis of amount of projectile that makes it through armor. Navy Ballistic Limit is based on complete passage of round through armor.
U.S. also used following limits during WW II: Army Ballistic Limit (success is a light passing crack or flake) and Protection Limit (less stringent than complete passage). Army Limit used for armor acceptance testing, and is close to WW II German ballistic limit for armor acceptance.
If 7/8" armor is listed penetration in TM9-1907, it means that half the hits meet success criteria at stated velocity and range.
WW II Cal. .50 AP M2 against 350 Brinell Hardness armor at 200 yards and 0 degree impact:
2935 fps muzzle velocity 1.08"
2835 fps muzzle velocity 0.95"
Against face-hardened armor at 200 yards, 2935 fps muzzle velocity bullet penetrates 0.92", about 15% less."
In practice, armour penetration results could be affected by a number of factors. In aircraft in particular, passing through the fuselage skin could disturb the path of the bullet so it didn't hit the armour point first, which would reduce the penetration considerably. I have read tests of firing rifle-calibre AP at WW2 bomber fuselages (from behind) in which the bullets failed to penetrate through to vital areas, even without any armour!
Tony Williams
Author: "Rapid Fire: The development of automatic cannon, heavy machine guns and their ammunition for armies, navies and air forces"
Details on my military gun and ammunition website:
http://www.users.globalnet.co.uk/~autogun/index.htm 
