A Surprising Solution

In Exercise 1.3 of SICP, we are asked to define a procedure that take three numbers as input and returns the sum of the squares of the 2 larger numbers. We already have a procedure, sum-of-squares, that returns the sum of the squares of its two inputs, so a natural solution is something of the sort

(define sum-of-squares-of-max
  (lambda (a b c)
  (sum-of-squares
  (if (< a b) b a)
  (if (< (if (< a b) a b) c) c (if (< a b) a b)))))

or, perhaps, the less opaque

(define sum-of-squares-max
  (lambda (a b c)
  (cond ((and (<= a b) (<= a c)) (sum-of-squares b c)) ;;a smallest
          ((and (<= b a) (<= b c)) (sum-of-squares a c));;b smallest
          (else (sum-of-squares a b)))))                ;;c smallest

Joe Marshall over at Abstract Heresies offers this surprising and elegant solution

(define sum-of-squares-max
  (lambda (a b c)
  (if (and (<= a b) (<= a c))                         ;;a smallest
        (sum-of-squares b c)
      (sum-of-squares-max b c a))))