1.6. ELEMENTARY MATRICES 21

with the 1 in the jth position from the left. The elementary matrix which results fromapplying this operation to the ith row of the identity matrix is of the form

E (c, i) =

r1...

cri...rn

 .

Now consider what this does to a column vector.r1...

cri...rn



v1...vi...

vn

=

v1...

cvi...

vn

 .

Denote by E (c, i) this elementary matrix which multiplies the ith row of the identity by thenonzero constant, c. Then from what was just discussed and the way matrices are multi-plied,

E (c, i)

a11 a12 · · · a1p

......

...ai1 ai2 · · · aip...

......

an1 an2 · · · anp

equals a matrix having the columns indicated below.

=

a11 a12 · · · a1p

......

...cai1 cai2 · · · caip

......

...an1 an2 · · · anp

 .

This proves the following lemma.

Lemma 1.6.4 Let E (c, i) denote the elementary matrix corresponding to the row op-eration in which the ith row is multiplied by the nonzero constant c. Thus E (c, i) involvesmultiplying the ith row of the identity matrix by c. Then

E (c, i)A = B

where B is obtained from A by multiplying the ith row of A by c.

Finally consider the third of these row operations. Letting r j be the jth row of theidentity matrix, denote by E (c× i+ j) the elementary matrix obtained from the identity