Yes, there is a speedup

[eliot@winnie.Princeton.EDU (eliot handelman)]

It turns out that I'm guilty of an oversight too shameful to mention. Here
are the correct timings -- as Bob Boyer indicated, the speedup was better
than 1000%. I compiled both functions using speed 3 & safety 0. The
machine is a microvax II.


-Eliot                                    

;;; No declarations
> (time (fib 29))
real time : 357.767 secs
run time  : 352.283 secs
514229

;;; Lots of declarations
> (time (fib 29))
real time : 31.650 secs
run time  : 31.467 secs
514229