From Division of Labor: The Nigerian lottery scam c. 1907?
How to win safely at baccarat has been discovered by Sylvain Crollet and his friend, Mme. Noel. The mathematical formula is X equals G divided by 3 plus P. X is thought to be the amount to be won. Unfortunately, we cannot determine P or G. However, a second lady was satisfied, and gave Sylvain 240 [pounds] to which to apply the formula.
The latter was such a success that Sylvain was able to pay the lady 12 pounds a month interest. She immediately intrusted to him 500 pounds more, but this time the equation seems to have failed. He paid her 16 pounds the first money, 16 pounds more two or three months later, then nothing at all. In court Sylvain, prosecuted by the lady, explained that his equation should have succeeded, but bad luck pursued him.
His mathematical calculations had not foreseen the element of chance. Ostend baccarat tables had completely upset his theorem. Anyhow, he had been unfortunate, not dishonest. But the court thought differently. Sylvain has got a year, and Mme. Noel, six months, while both are jointly sentenced to refund 660 pounds to the deluded lady, plus 20 pounds damages.
The single "equation" has two unknowns and therefore P and G cannot be determined uniquely. The fact that an innumerate lady gave money towards such a system seems funny but reminds us that promises of unearned riches are not unique to the Internet age.
The defense that it was "bad luck" and the element of "chance" that "upset the theorem" gets a chuckle from me.
Although DoL's post doesn't say it I am guessing this is from their ongoing New York Times "This day in the Times series".