What Was Inside the Skull? Reading a Brain From the Space It Left

An endocast is a model of the space inside the braincase, produced either by nature, where sediment filled the cavity and hardened, or by a researcher, historically by making a physical cast and now almost always by computed tomography scanning followed by digital reconstruction. The cast records the shape of the cavity rather than the brain itself, and how closely those correspond varies enormously between animals. In birds and mammals the brain fills the cavity closely, so the cast is a good approximation. In crocodiles and lizards it fills perhaps half, with the rest occupied by membranes, blood vessels and fluid, so a cast substantially overstates the brain. Dinosaurs sit between, with the proportion differing between groups, and estimating it is the first and largest source of uncertainty in anything that follows.

Certain features are reliable and each supports a specific inference:

• •Olfactory bulbs at the front, whose size relative to the rest indicates how much of the brain was devoted to smell
• •Optic lobes, whose prominence indicates reliance on vision
• •The inner ear, preserved as a cavity in the same bone, which records the range of hearing and the sensitivity
• •The semicircular canals, whose orientation indicates the habitual angle at which the head was held
• •Cranial nerve openings, which indicate what the nerves served and how large they were
• •Overall volume relative to body mass, which is a crude measure and the one most frequently overinterpreted

The semicircular canals of the inner ear detect rotation and sit in a consistent orientation relative to the horizon when an animal holds its head in the alert resting position, which is established across a wide range of living species. That relationship allows the habitual head posture of an extinct animal to be reconstructed from the canals alone, independently of any assumption about the neck, and the method has been applied widely. Results have supported reconstructions where the snout is angled downward in several groups and have corrected some museum mounts. The approach has limits, since the relationship in living animals has more scatter than early applications assumed, and since posture varies with activity. It remains one of the better examples of a specific measurable feature answering a question that would otherwise be guesswork.

A persistent popular claim holds that some dinosaurs had a second brain in the hip, which is worth addressing because it is repeated constantly and is wrong. The observation behind it is real, since several groups including stegosaurs have a conspicuous enlargement of the spinal canal in the hip region, considerably larger than the braincase in some specimens. The interpretation as a supplementary brain was proposed in the nineteenth century and entered popular culture, including a widely reprinted comic verse. The modern explanation is that the space housed a structure found in living birds, a body of glycogen-rich tissue whose function is not fully settled but which is certainly not nervous control, and that comparable enlargements in birds serve that structure. Nerve tissue does not work that way in any case, since a second processing centre would need to communicate with the first and the arrangement confers no advantage.

Comparing brain size between species requires accounting for body size, since larger animals have larger brains for reasons that have nothing to do with capability, and the standard approach uses a ratio between actual brain size and the size expected for an animal of that mass. Applying that to dinosaurs runs into several problems at once. Body mass estimates for extinct animals carry large uncertainties, and the ratio is sensitive to them. The proportion of the cavity actually occupied by brain is unknown and differs between groups. And the relationship between the resulting number and anything behavioural is weak even among living animals, where species with similar ratios differ enormously in what they do. Published figures giving such ratios for extinct animals should therefore be read as rough orderings rather than measurements, and conclusions about intelligence drawn from them are not supported by the method.

A cast of the braincase cavity records the space rather than the brain, and how much of that space the brain filled varies from nearly all in birds to about half in crocodiles. Olfactory bulbs, optic lobes and the inner ear give reliable readings about senses. Semicircular canal orientation reconstructs habitual head posture independently of the neck. Brain size ratios rest on uncertain body masses and support very little.