This scatterplot of the rate of change of wage rates and the percentage unemployment for the years 1861-1913 is the womb of an assumed relation that was later on baptized Phillips curve by Samuelson and Solow. Here is the non-photoshop-edited plot:"The crosses give the average values of the rate of change of money wage rates and of the unemployment rate in those years in which unemployment lay between 0 and 2, 2 and 3, 3 and 4, 4 and 5, 5 and 7, and 7 and 11 per cent respectively. <> The curve was fitted to the crosses. The form of equation chosen was

log(y + a) = log b + c log x

where y is the rate of change of wage rates and x is the percentage unemployment. The constant b and c were estimated by least squares using the values of y and x corresponding to the crosses in the four intervals between 0 and 5 per cent unemployment, the constant a being chosen by trial and error to make the curve pass as close as possible to the remaining two crosses in the intervals between 5 and 11 per cent unemployment. The equation of the fitted curve is log(y + 0.9) = 0.984 - 1.394 log x."[1]

A two-page summary of Phillips curve related stuff can be found here (New Evidence of the Old Phillips Curve, Cato Institute, 2002) Contains some Brad DeLong-Bashing ;-D.

I plotted the rate of inflation against the rate of unemployment in Austria for each year from 1962 to 2003. The message: Austrians figured out that price stability is a good thing but have no clue what to do against non-keynesian unemployment (The unemployment rate in Austria is highly persistent, it has been rising steadily during the past 30 years. see also Hysteresis)

[1] The Relation between Unemployment and the Rate of Change of Money Wage Rates in the United Kingdom, 1861-1957, A.W. Phillips, Economica, Vol. 25, Nov. 1958)