Answer: A pound of potatoes, because feathers are measured in the Troy system.
(Correct? Perhaps iron and gold would be better.)
Here's the best explanation of this I can find, in a quotation in my OED,
sv 'pound':
1749 Reynardson in Phil. Trans. XLVI. 59 At the same Time [1696] and Place,
the Standard Troy Weights were compared with the Standard
Avoirdepois,..which fixes the Pound Avoirdepois at 7000 such Grains, as the
Troy Pound weighs 5760.
David Fowler
At 11:18 pm -0400 3/6/99, Everdell@aol.com wrote:
>Pat Ballew writes:
>
><< My dictionary defines the word "as" as an ancient Roman unit of weight
>equal to about one Troy pound. It also defines "libra" as an ancient unit of
>weight correspoinding to one pound. I know that the as was also a coin at
>some point in the Roman history. My question is, can anyone tell me if one
>preceeded the other or were they in use at the same time. >>
>
>My Oxford English Dict. informs me that the "as" was a coin of 12 (Roman)
>ounces or one (Roman) pound of copper until the 1st Punic War, when inflation
>reduced it to 2 ounces. It went to 1 ounce during the Second Punic War, and
>to 1/2 ounce by the Lex Papiria in 191 BC, which is a few years after
>Carthage lost that second war. The same source gives fewer facts about
>"libra": only the same weight, 12 ounces, the vague date, "medieval Latin,"
>and a first English citation from 1398. On the other hand, that might well
>be sufficient to answer half the question by demonstrating that the libra
>must have been introduced at 12 ounces at a time after the as had been
>devalued to 1/24 of a pound (though who could measure it in the Middle Ages
>so far had standards deteriorated in weights and measures). Whether the
>coins or the denominations were ever used together, most likely in medieval
>Europe, needs more than such half-baked desktop research to decide.
>
>-Bill Everdell, Brooklyn