196 5. Further Determinant Theory
∑
r,s†
crcsErs
=−∑
r,s†
(c2
r+c2
s)Ers=−
1
3∑
r,s,tE
rst,rst
, (5.4.25)∑
r,scrcsErs
=∑
rc2
rErr
−1
3
∑
r,s,tE
rst,rst
, (5.4.26)∑
r,s(c2
r+c2
s)Ers
=2∑
rc2
rErr
+1
3
∑
r,s,tE
rst,rst
, (5.4.27)∑
r,s†
c2
rErs
=1
6
∑
r,s,tE
rst,rst
+∑
r,scrErs,rs
, (5.4.28)∑
r,s†
c2
sErs
=1
6
∑
r,s,tE
rst,rst
−∑
r,scrErs,rs. (5.4.29)
To prove (5.4.20), apply the second equation of (5.4.4) and (5.4.16).
Epr,ps=∂Ers∂xp.
Multiply by (cr−cs) and sum overrands:
∑
r,s(cr−cs)Epr,ps=∂
∂xp∑
r,s(cr−cs)Ers=
∂
∂xp∑
r,sErs,rs=
∑
r,sEprs,prs,which is equivalent to (5.4.20). The application of the fifth equation in
(5.4.4) with the modification (i, p, r, q, s)→(u, r, t, s, t) to (5.4.20) yields
(5.4.21).
To prove (5.4.22), sum (5.4.11) overiandj, change the dummy variablesas indicated
∑i,j(c2
i−c2
j)Eij
=F−Gwhere, referring to (5.4.6) and (5.4.7),
F=
∑
i,sE
is
∑
r,jcrErj
+
∑
r,jE
rj
∑
i,scsEis
=
∑
iE
ii
∑
r,j
(j→s)crErj
+∑
i,s
(i→r)csEis