TITLE: Mathematical objects whose name is a single letter
QUESTION [5 upvotes]: (Not research-level, but perhaps not easily answered elsewhere — you decide if MO can afford the innocent fun. If so, it should likely be “community-wiki” i.e. one object per answer.)

I am seeking stories of mathematical objects that, in context, eat out namespace because their (most usual) name literally is a letter (e.g., in calculus, $e$).

Per discussion in the comments, please rather exclude letters that are frozen out by being merely common notation ($e$ in group theory, $g$ in Riemannian geometry, whole alphabets in semisimple Lie theory), and not really the name of any single object. But include: how such (poor?) practice came about; what did or didn’t help reclaim letters (new names, new typography,...); or any good story.
Wikipedia’s disambiguation pages can suggest many ($c$, $e$, $i$, $j$, $k$, $o$, $q$, $t$, $F$, $G$, $J$, $K$, $L$, $O$, $P$, $W$, $Y$, $\mathcal O$, $\wp$, $\delta$, $\zeta$, $\eta$, $\vartheta$, $\varkappa$, $\lambda$, $\xi$, $\pi$, $\sigma$, $\tau$, $\chi$, $\mathrm B$, $\Gamma$, $\mathrm H$, $\Xi$, $\Omega$,...), but I am sure that is not all.

REPLY [5 votes]: Imagine how different would be our mathematical discourse if students struggled
with $\alpha$-$\beta$ proofs rather than $\epsilon$-$\delta$ proofs!
From an earlier MO question: "Why do we use $\epsilon$ and $\delta$?,"
an answer quoting Judith Grabiner:

$\epsilon$ corresponds to the initial letter in the word "erreur" (or "error"), and Cauchy in fact used $\epsilon$ for "error" in some of his work on probability."

As @RyanBudney summarizes,
"So it's  $\epsilon$ for error in the answer, and presumably $\delta$ is in reference to difference in the input variables."
Or, as @SimonRose says, "Or possibly that $\delta$ is just the next letter over..."

REPLY [2 votes]: Trivial example, but not to be omitted, the Greek and Latin numerals. E.g.: