A=00d0beacf,∣A∣=−abd=0c11=+(bf−ce),c12=−(−c d)=cd,c21=−(−ca)=ae,c22=+(−ad)=−ad,c31=+(−ab)=−ab,c32c23=−(0)=0AdjA=(bf−ce)cd−bdae−ad0−ab00 A−1=∣A∣1(adjA)=abd1bf−cecd−bdae−ad0−ab00 AT=00a0bcdef Now A−1=AT⇒−abd1bf−cecd−bdae−ad0−ab00⇒bf−cecd−bdae−ad0−ab00∴bf−ce=00a0bcdef abd2=ab,ab2d=ad,a2bd=bd...(ii) abde =abcd=abdf=0 From (ii), (abd2)⋅(ab2d)⋅(a2bd)=ab.ad.bd ⇒(abd)4−(abd)2=0 ⇒(abd)2[(abd)2−1]=0 ∵abd=0,∴abd=±1 From (iii) and (iv), e=c=f=0 From (i) and (v), bf=ae=cd=0 From (iv), (v) and (vi), it is clear that a,b, d can be any non-zero integer such that abd=±1 But it is only possible, if a=b=d=±1 Hence, there are 2 choices for each a,b and d. there fore, there are 2×2×2 choices for a,b and d. Hence number of required matrices =2×2×2=(2)3