We shouldn't forget about cute and (historically) important stem-and-leaf plot (that Tufte loves too!). You get a directly numerical overview of you data density and shape (of course if your data set is not larger then about 200 points). In R, the function stem produces your stem-and-leaf dislay (in workspace). I prefer to use gstem function from package fmsb to draw it directly in a graphic device. Below is a beaver body temperature variance (data should be in your default dataset) in a stem-by-leaf display:
require(fmsb)
gstem(beaver1$temp)
