It was in April 1986
when Bednorz and Mueller of the IBM Zü\;rich laboratories sent a pape
r about &ldquo\;possible high-Tc superconductivity&rdquo\; to *Zeitschr
ift fü\;r Physik* B. \; The resulting bombshell changed conden
sed-matter physics forever. \; Experimenters and theorists developed m
ethods to measure and calculate in ways that were much improved over prior
years. However\, despite 30 years of intense study\, the description of t
hese materials remains incomplete. I&rsquo\;ll discuss the \; discover
y of the high Tc cuprates from the perspective of a participant. \; I&
rsquo\;ll then turn to what infrared spectroscopy can tell us about their
properties. Measurements for a number of cuprate families of optical refle
ctance over a wide spectral range (far-infrared to ultraviolet) have been
analyzed using Kramers-Kronig analysis to obtain the optical conductivity\
, *s*(*w*)\, and (by integration of the real part of the con
ductivity) the spectral weight of low- and mid-energy excitations. For the
Kramers-Kronig analysis to give reliable results\, accurate high-frequenc
y extrapolations\, based on x-ray atomic scattering functions\, were used.
When the optical conductivities of the normal and superconducting states
are compared\, a transfer of spectral weight from finite frequencies to th
e zero-frequency delta-function conductivity of the superconductor is seen
. The strength of this delta function gives the superfluid density\, *r
**s*. There are two ways to measure *r**s*\, usin
g either the low energy spectral weight or by examination of the imaginary
part\, *s*2(*w*)\; both estimates show that 98% of the *
ab*-plane superfluid density comes from low energy scales\, below abou
t 0.15 eV. Moreover\, there is a notable difference between clean metallic
superconductors and the cuprates. \; In the former\, the superfluid d
ensity is essentially equal to the conduction electron density. The cuprat
es\, in contrast\, have only about 20% of the *ab*-plane low-energy
spectral weight in the superfluid. The rest remains in finite-frequency\,
midinfrared absorption. In underdoped materials the superfluid fraction i
s even smaller. The consequences of this observation for the electronic st
ructure will be addressed.