I just re-read Maciej Cegłowski’s excellent essay on the discovery and loss of the remedy to scurvy.
Incidentally, I’ve seen scurvy used as a case study of how theory colours evidence (or directs the interpretation of it) before in Lukas Vermeer’s talk Creating data: a boat filled with sauerkraut.
There are loads of quotable passages, but rather than try to condense a great read into a few snippets, I’ll share a somewhat ancillary tidbit:
[T]echnological progress in one area can lead to surprising regressions. I mentioned how the advent of steam travel made it possible to accidentally replace an effective antiscorbutic with an ineffective one. An even starker example was the rash of cases of infantile scurvy that afflicted upper class families in the late 19th century. This outbreak was the direct result of another technological development, the pasteurization of cow’s milk. The procedure made milk vastly safer for infants to drink, but also destroyed vitamin C. For poorer children, who tended to be breast-fed and quickly weaned onto adult foods, this was not an issue, but the wealthy infants fed a special diet of cooked cereals and milk were at grave risk.
The essay contained two words new to me: antiscorbutic — having the effect of preventing or curing scurvy; and austral — of or coming from the south.