> I have notice that H. S. M. Coxeter loves to present projective geometry
> synthetically, and he generally avoids the analytic approach as much as he
> can.
When I was a student (40 years ago) I always did problems by synthetic
geometry if I could (rigid-body dynamics offered some surprising
examples). When it worked at all it was usually faster, neater and more
memorable. Of courses there are many problems that won't come out that
way, and then one needs one's algebra and calculus.
Was it Feynman or Wheeler who said there were about 6 ways to do any
problem in quantum electrodynamics, and that they were all equaivalent
mathematically, but one should always do any problem all 6 ways, because
they all told you something different physically?
John Harper, School of Mathematical and Computing Sciences,
Victoria University, Wellington, New Zealand
e-mail john.harper@vuw.ac.nz phone (+64)(4)471 5341 fax (+64)(4)495 5045