The Self-Mending Net
209
processes took place, but of course the cells couldn't penetrate the di-
vider. After the cellular activity had died down, Bernstein removed the
barriers, but there was no further change. However, when he then cut
off each damaged end, producing an even larger gap and reinjuring the
cord, the cells started from scratch and healed the defect completely.
Thus there's good reason to believe that even long-standing spinal inju-
ries can potentially be regenerated if we can extend the basic capabilities
of human cells.
One would expect to see some healing response in mammals, even if
it fell short. After all, we only need the elongation and reattachment of
fibers, which does take place in peripheral nerves. Instead the opposite
happens. The cord cells die a short distance above and below the injury.
Cysts form near the ends, and, instead of ependyma, scar tissue fills the
gap. Only after this destruction is there an abortive attempt at re-
growth. In humans this amounts to only a few millimeters of fiber
elongation many months after the injury. By then it's too late; the epen-
dymal cells and nerve fibers can't penetrate the scar.
CYSTS AND SCARS PREVENT CORD REGROWTH IN MAMMALS
Why the difference between salamanders and mammals? The reason
may lie in the cord's immediate response. In all animals the injury in-
stantly results in spinal shock, during which all neuronal activity is pro-
foundly depressed, especially in the part of the cord still connected to
the brain. Even the simplest reflexes disappear. As the shock wears off,
the cord below the injury becomes hyperactive. Its reflexes become tre-