In the early 1880s, Henri Poincaré discovered an equilibrium figure for uniformly-rotating fluid masses—the pear, or piriform figure—and speculated that in certain circumstances the pear splits into two unequal parts, and provides thereby a model for the origin of binary stars. The contemporary emergence of photometric and spectroscopic studies of variable stars fueled the first models of eclipsing binaries, and provided empirical support for a realist view of equilibrium figures—including the pear—in the cosmic realm. The paper reviews astrophysical interpretation of (...) the Poincaré pear and the Poincaré-Darwin fission hypothesis with respect to research on variable stars from 1885 to 1901. (shrink)