Page 203 -
P. 203

180    9. Elliptic and Hypergeometric Functions

                                   
(n) = (n − 1)!.

           Hence 
(s) is a generalization or analytic continuation of the factorial. The ana-
        lytic function 
(s) defined for Re(s)> 0 can be continued as a meromorphic func-
        tion in the entire plane C with poles on the set −N of negatives of natural numbers.
        Using (5.2), we derive the relation

                                         
(s + n)
                          
(s) =
                                s(s + 1)(s + 2) ··· (s + n − 1)
        for Re(s)> 0, but the right-hand side is a meromorphic function for Re(s)> −n,
        and it defines a meromorphic continuation of 
(s) to the half plane Re(s)> −n.
        This works on all half planes Re(s)> −n and hence on all of C.
                                            2
        (5.3) Remark. By change of variable x = t , dx = 2tdt,and dx/x = 2(dt/t),we
        have the following integral formula for the gamma function:

                                         ∞       dt
                                        (
                                            2s −1
                                
(s) = 2   t e     .
                                         0        t
        (5.4) Proposition. For a, b > 0 we have
                         (  π/2                    
(a)
(b)
                                2a−1     2b−1
                        2     cos   θ · sin  θ dθ =        .
                          0                        
(a + b)
        Proof. We calculate
                             ∞   ∞              "          #
                           (   (
                                                     2
                                        η
               
(a)
(b) = 4        ξ 2a−1 2b−1  exp − ξ + η 2  dξ dη
                            0   0
                             ∞         2 dr
                           (                 (  π/2
                                    e
                        = 2    r 2a+2b −r  · 2    cos 2a−1  θ sin 2b−1  θ dθ
                            0            r    0
        using the transformation dξ dη = rdrdθ to polar coordinates. This proves the
        proposition.
                                                 π/2
           As a special case a = b = 1/2, we have 2  dθ = 
(1/2)
(1/2)/ 
(1) and
                      √                        0
        hence 
(1/2) =  π. The gamma function prolongs the factorial, and the binomial
        coefficient is prolonged by the formula
                              s    s(s − 1) ··· (s − n + 1)

                                 =                    .
                              n             n!
        The corresponding binomial series is

                                               s
                                       s          n
                                (1 + x) =        x
                                              n
                                          0≤n
        which converges for |x| < 1and Re(s)> 0.
   198   199   200   201   202   203   204   205   206   207   208