Page 135 - Determinants and Their Applications in Mathematical Physics
P. 135

120   4. Particular Determinants

          method can be illustrated adequately by taking the particular case in which
          (n, r)=(4, 3) and φ m is an Appell polynomial so that F =1.
            Let

                             T = C 3 C 4 C 5 C 6 = T 3456 .


          Then
                              D(T)=3T 2456 ,
                            2
                           D (T)/2!=3T 1456 +6T 2356 ,
                            3
                           D (T)/3! = T 0456 +8T 1356 +10T 2346 ,
                         .......................................
                            9
                           D (T)/9! = T 0126 +8T 0135 +10T 0234 ,
                          10
                         D (T)/10!=3T 0125 +6T 0134 ,                (4.9.8)
                          11
                         D (T)/11!=3T 0124 ,
                          12
                         D (T)/12! = T 0123 ,
                                   = |φ m | 4 ,  0 ≤ m ≤ 6
                                   = constant.

          The array of coefficients is symmetric about the sixth derivative. This result
          and several others of a similar nature suggest the following conjecture.

          Conjecture.
                      D {T  (n,r) } =(nr)!|φ m | n ,  0 ≤ m ≤ 2n − 2
                        nr
                                  = constant.

            Assuming this conjecture to be valid, T (n,r)  is a polynomial of degree
          nr and not n(n + r − 1) as may be expected by examining the product of
          the elements in the secondary diagonal. Hence, the loss of degree due to
          cancellations is n(n − 1).
            Let


                       T = T  (n,r)  = C r C r+1 C r+2 ··· C r+n−1 ,


                                                           n
          where

                                                           T
                       C j = ψ r+j−1 ψ r+j ψ r+j+1 ··· ψ r+j+n−2
                                                           n
                             f (m) (x)
                       ψ m =        ,  f(x) arbitrary
                               m!
                       ψ    =(m +1)ψ m+1 .                           (4.9.9)
                        m
          Theorem 4.35.



                      T =(2n − 1+ r) C r C r+1 ··· C r+n−2 C r+n
                                                             n
                                        (n+1,r)
                        = −(2n − 1+ r)T      .
                                        n+1,n
   130   131   132   133   134   135   136   137   138   139   140