This includes topics that are no longer covered in calc classes, but
used to be.
Mostly they're no longer relevant -- or rather, they're cute but not
useful.

----------------------------------------
This draws heavily from:
Steven Krantz's "How to teach mathematics", p.256 (Uhl's appendix)
----------------------------------------
In particular, things we've seen dropped are:

- (manual) numerical methods
  Before calculators, people needed to do calculations by hand, so a
  proficiency with complicated manipulations was important.

  This is an area where calculators have really replaced knowing math.
  [That is, many of these computations had no intrinsic value;
   they are -just- techniques/algorithms, w/o much deeper meaning.]
  [A similar point in elementary math is that we don't teach ``casting
   out 9s'' to check sums. We actually should: it's not a valuable
   computational skill, but there's important math behind it,
   namely modular arithmetic.]
  I'd argue that symbolic calculus systems have done the same for the
  trig reduction formulae in integration.
  Note also that what computations we -still- do need justification:
  - simple computations, like adding with carrying and subtracting
    with borrowing, are useful for small quantities, like calculating
    change or estimating small sums
  - multiplication and long division are not useful -per se-,
    but b/c of their deeper meaning: long division -> the euclidean
    algorithm and polynomial division
    Long multiplication? well, kinda similarly!

  Today, -some- knowledge of numerical issues matter, especially:
  - stability
    ...and how many digits of accuracy one can get
  - problem cases
    EG, when a function has derivative 0 or infinity,
    many formulae don't work very well.
    (EG, computing arc-cosine for values close to 0)

- elementary real algebraic geometry
  EG:
  - detailed study of conics
  - special (often algebraic) plane curves
  - Descarte's rule of signs

  This has disappeared for 3 main reasons:
  - more general functions
    for many applications, general smooth (or analytic, or continuous)
    functions are just as useful as real polynomials, if not more so.
    
  - complex numbers
    Much algebra gets -much- clearer over the complex numbers;
    if you respect them and like them (which you should),
    there's not much general value in studying real polynomials,
    though they do have some nice (but not terribly useful) properties

  - less astronomy
    Astronomers love conics.

Note that the lessening importance of astronomy lead to both less
study of conics and less study of spherical geometry.