The Burgess Shale preserves a diversity of body plans that later collapsed

Gould centers his argument on the Burgess Shale, a fossil bed in the Canadian Rockies exceptional for preserving soft-bodied animals from roughly 505 million years ago in remarkable detail, shortly after the Cambrian explosion of animal diversity. Detailed reanalysis of these fossils, largely by researchers at Cambridge, revealed an anatomical range wider than modern classification initially assumed: arthropods alone displayed far more distinct body-plan variations than the handful of designs modern arthropod groups now represent, alongside genuinely strange creatures that don't fit comfortably into any living phylum.

Gould's key observation is that most of this early diversity did not survive into the modern world. The vast majority of these body plans, whatever their initial success, simply vanished from the fossil record without leaving descendants, leaving today's animal kingdom populated by only a narrow subset of what once existed.

This sets up his central puzzle: why did such a wide field of early anatomical experimentation narrow so drastically, and was that narrowing driven by clear superiority of the survivors, or by something closer to chance? Takeaway: evolutionary history began with more anatomical possibility than survived into the world we recognize today.