A quick visual study of the box and bullets revealed they are swaged bullets with a cone shaped depression in the base. Other bullets, illustrations, and descriptions I have seen have always shown a shallow dish in the base. This cone is deeper than the dished bases I have seen.

The paper patch is nearly perfectly consistent on all 25 bullets with some variation on the amount of overlap on the base which is a Hyde style (meaning it does not have a tail and in fact has a hole over the hollow base cavity). All patches are tight; these bullets could be fired today.

Box indicates 1 to 19 bullet alloy - about 10bhn.

Here's the pic from Aaron:

Measurements

Weight 370grs Nominal - more on actual weights later.
Length 1.245", except for some measuring 1.238" - more on that later.
Diameter over the two wraps of PP .4010". Bore rider.
Diameter just ahead of the PP .388". This exposed nose appears to be tapering. Did not unwrap any to check the cylindrical section.
Paper thickness appears to be .002" from careful micrometer measurement at the base. Given that, the cylindrical diameter would be .401" - .008" = .393". Did not unwrap for a certain measurement.

Each of the 25 bullets were weighed and the spreadsheet kicked out these numbers:

Sample Size = 25
Mean = 371.1grs
Minimum = 369.9grs
Maximum = 371.8grs
Extreme Spread = 1.9grs
Standard Deviation = .656grs
3 Sigma = 1.968grs
Mean - 3 Sigma = 369.1grs
Mean + 3 Sigma = 373.1grs
Percent Variation (3 Sigma / Mean) = 0.53%

Since swaged bullets should be more consistent than cast bullets, these numbers were surprisingly bad; 3 sigma predicting plus or minus 1.97grs to encompass 99.7% of all bullets from this process. Sorting the individual bullet weights into a graph that mimics a Bell curve, something appeared - let me explain what the chart represents first. The horizontal axis is marked with individual bullet weights in .1gr increments. The vertical axis are numbers ranging from 0 to 7. These represent the numbers of bullets in the 25ct sample that weighed that amount. For example, there were 6 bullets that weighed 371.5grs.

If a process is under control and working well, there will be a single bell shaped distribution of the product - preferably a tall and narrow bell, as this indicates a tight range as represented by a low standard deviation. Here we have TWO bell shaped groups! Remember the length measurements? The left most (lighter) group measured .008" shorter, consistently. Best guess is this box contained bullets from two different swaging dies? Interesting... never considered the tolerance variation between dies impacting production quality. These would respond well to sorting, however, as the numbers are not all over the place, but cluster together in two groups.

Measurement Data

For comparison, and to show cast bullets can be consistent, here is data from a 206ct batch of Lyman 457125 520gr bullets.

Sample Size = 206
Mean = 519.0grs
Minimum = 518.1
Maximum = 519.8
Extreme Spread = 1.7grs
Standard Deviation = .348
3 Sigma = 1.044grs
Mean - 3 Sigma = 518.0
Mean + 3 Sigma = 520.1
Percent Variation (3 Sigma / Mean) = .20%

Here they are stacked on paper with the weights marked below. One grouping, physically illustrating how the chart above works. It is easy to do if you want to see if your casting process is consistent and working well.

Again, thanks to AaronB for making these available!

This bimodal distribution certainly argues for two different swaging dies, or two different something at least. I'm wondering if the bullets were swaged from slugs of cut lead wire, and it was variation in the weight of the lead slug prior to going into the die that made the difference.

It might have been two different operators at the blank cutter, or (possibly) two different lots of lead wire (with one being just a hair thicker than the other).

Two different blanks, one 2 grains heavier than the other, might not be noticeably different to the swager operator.

-AaronB

PS: Now what? Are you going to load and shoot them?

and it is unusual to see such a clear difference. Definitely food for thought.

Shoot them? I couldn't do that... no, these are for study and admiring. Perhaps some future owner might.

John K.

that level of scrutiny of swaged bullets! Now, air pellets, some shooters get down right persnickety. Crosman used to do 1 die lots in 1250 or so pellet count, if I recall correctly. The theory being that when all pellets were from one die they'd be more consistent. BUT their usual production is from multiple dies all going into one batch from which they then fill the tins. Which helps explain why their pellets often aren't the most accurate, among other factors of course. Your findings definitely make me curious as to what kind of equipment the factory was using to produce those.

"Fill one out for the collectors." Maybe?

That box of bullets came from an estate sale.

variation would have a large percentage impact overall and wreak havoc with velocities and groups.

In a sense, perhaps air rifle shooting is comparable to long range BPCR shooting. The performance of each are dependent on consistency of velocity to get predictable round groups at range. This is why I started studying cast bullets.