It was found [3, 4, 11-14] that many problems in the universal algebraic geometry, such as geometric equivalence, geometric similarity, isomorphism and equivalence of categories of algebraic sets and varieties depend on structure of $Aut\left({\Theta }^0\right)$  and Aut(End( $W$  )), where ${\Theta }^0$  is the category of algebras $W=W\left(X\right)$  with the finite $X$  that are free in $\Theta$  , and $W$  is a free algebra in $\Theta$  .

Structure of $Aut(Com-P{)}^0$  (i.e, the classical case) was described in [2].

Here was introduced variant of the concept of quasiinner automorphism basic concept for description of Aut(End( $W$  )) and $\text{Aut}\left({\Theta }^0\right)$  . Later groups $\text{Aut}\left({\Theta }^0\right)$  were described for categories of free Lie algebras and free associative algebras [1, 5, 8, 9]and for categories of free modules and free Lie modules [5, 6]. Some other definition of quasiinner automorphism was used in mentioned papers. Also are described $Aut\left({\Theta }^0\right)$  for varieties of groups, semigroups and some another varieties. For some varieties is described Aut(End( $W$  )), but these problems are more difficult. In particular, this problem was not solved even in the classical case of free commutative algebras, i.e. the structure of Aut(End $P[x_1,\dots ,x_n]$  )) is not describe if $n>2$  . In all solved cases groups Aut(End( $W$  )) are generated by semi-inner automorphisms and mirror automorphism. In [3] this problem in classical case was solved, if the field $P$  is algebraically closed.

I want to note that the proof in [3] does not use the description of the group $Aut\left(P\right[X\left]\right)$  ). It is very important, because structure of the group $Aut\left(P\right[x_1,\dots ,x_n\left]\right)$  is not described, when $n>2$  , but counterexamples show that structures of $Aut\left(P\right[x_1,\dots ,x_n\left]\right)$  and $Aut\left(P\right[x,y\left]\right)$  are principally different.

Now in the presented paper the group Aut(End( $W_n$  )) is described for commutative and associative algebras over arbitrary infinite field.

2 Definitions

Let us recall some definitions for variety of commutative algebras $\text{Com}-P$  (see [4,13] for general case).

Definition 2.1 Let $P\left[X\right]=P[x_1,\dots ,x_n]$  be a free commutative algebra over a field $P$  with a finite set of generators and $\tau \in \text{Aut}\left(\text{End}\right(P\left[X\right]\left)\right)$  . It is known $\text{[4]}$  that there exists a bijection $\mu :P\left[X\right]\mapsto P\left[X\right]$  such that for every $s\in \text{End}\left(P\right[X\left]\right)$  $$s^{\tau }=\mu s{\mu }^{-1}.$$  Such representation of $\tau$  is called a representation of $\tau$  as quasiinner automorphism generated by $\mu$  .

Of course, arbitrary set of substitutions does not generate an automorphism, But there are two kinds of bijections $\mu$  in commutative case and three in the associative case that generate an automorphism.

1. $\mu =\overline{\alpha }$  , where $\overline{\alpha }$  is the natural extension of an automorphism $\alpha \in \text{Aut}\left(P\right[X\left]\right)$  on $W$  , i.e. $\overline{\alpha }\left(\sum a_k{u}_k\right)=\sum \alpha \left(a_k\right)u_k$  . We shall write $\overline{\alpha }=\alpha$  and call it ”automorphism of field”.

2. $\mu =\eta \in \text{Aut}\left(P\right[X\left]\right)$  .

Definition 2.2 The automorphism $\tau$  generated by an automorphism $\mu \in \text{Aut}\left(P\right[X\left]\right)$  is called inner automorphism.

Definition 2.3 Note, that every $\alpha \in \text{Aut}\left(P\right)$  belongs to normalizer of subgroup $\text{Aut}\left(P\right[X\left]\right)$  in the group of all bijections of $P\left[X\right]$  . So, every product of elements of $\text{Aut}\left(P\right)$  and $\text{Aut}\left(P\right[X\left]\right)$  my be represented in the form $\mu =\alpha \eta$  , $\alpha \in \text{Aut}P$  , $\eta \in \text{Aut}\left(P\right[X\left]\right)$  . Automorphism $\tau \in \text{Aut}\left(\text{End}\right(P\left[X\right]\left)\right)$  generated by $\mu$  is called semiinner.

3. Now we describe the third type of $\mu$  for associative algebras. Let $S=S\left(X\right)$  be a free semi-group. For every $u=x_{i_1}{x}_{i_2}\dots x_{i_n}$  in $S$  , take $\overline{u}=x_{i_n}\dots x_{i_2}{x}_{i_1}$  . Then $u\to \overline{u}$  is an antiautomorphism of the semigroup S.

Definition 2.4 Let $W=W\left(X\right)$  be a free associative algebra. For every its element $\omega ={\lambda }_0+{\lambda }_1{u}_1+\cdots +{\lambda }_k{u}_k$  denote $\beta \left(\omega \right)=\overline{\omega }={\lambda }_0+{\lambda }_1\overline{u_1}+\cdots +{\lambda }_k\overline{u_k}$  . The transition $\beta :W\left(X\right)\to W\left(X\right)$  is an antiautomorphism of the algebra $W$  , and it generates the automorphism $\delta \in \text{Aut}\left(\text{End}\right(W\left(X\right)\left)\right)$  , $s^{\delta }=\beta s{\beta }^{-1}$  . We call $\beta$  mirror antiautomorphism of $W$  and $\delta$  mirror automorphism of $\text{Aut}\left({\Theta }^0\right)$  .

Here $\delta$  is not inner and is not semi-inner, but is quasi-inner. Note that

• 1. $\delta$  belongs to the normalizer of subgroup $\text{Sinn}\left(\text{End}\right(W\left(X\right)\left)\right)$  in $\text{Aut}\left(\text{End}\right(W\left(X\right)\left)\right)$  that consists of all semi-inner automorphisms;
• 2. every antiautomorphism of $W$  is a product of the mirror antiautomorphism $\beta$  and an automorphism of $W$  .

Let now $W\left(X\right)$  be a free finitely generated algebra in arbitrary variety $\Theta$  .

Definition 2.5 Bijection $\mu :W\left(X\right)\mapsto W\left(X\right)$  is called central, if for every $s\in \text{End}\left(W\right(X\left)\right)$  $$\mu s=s\mu .$$  Algebra $W\left(X\right)$  is called central, if every its central bijection is identical.

Theorem 1 Let $W\left(X\right)$  be a free central algebra and $\mu$  some bijection of $W\left(X\right)$  generating an automorphism $\tau \in \text{Aut}\left(\text{End}\right(W\left(X\right)\left)\right)$  . Then $\mu$  transforms every base of algebra $W\left(X\right)$  to a base of algebra.

Proof. [see 4] For free algebras in a variety $\Theta$  of algebras over field $P$  (Com-P, Ass-P, and others) we shall correct the definition of central algebra. Clearly, in every such algebra $W\left(X\right)$  we have a bijection $\mu \left(u\right)=au+b$  , where $a,b\in P,$  $a\ne 0$  (linear bijection). This bijection commutate with every endomorphism of algebra $W\left(X\right)$  and, obviously, transforms every base of $W\left(X\right)$  to a base.

Definition 2.6 Finitely generated algebra $W\left(X\right)$  over field $P$  is called almost central, if every its central bijection is linear.

Theorem 2 Free finitely generated commutative, associative are almost central.

Proof. Let $\mu$  be a central bijection of $W\left(X\right)$  , $\mu \left(x_i\right)=r_i(x_1,x_2,\dots ,x_n)\in W\left(X\right)$  , and $s\in \text{End}\left(W\right(X\left)\right)$  , $s\left(x_1\right)=x_1$  , $s\left(x_i\right)=0$  for $i=1$  . Then $$\mu s\left(x_1\right)=\mu \left(x_1\right)=r_1(x_1,x_2,\dots ,x_n),$$  $$s\mu \left(x_1\right)=s\left(r_1\right(x_1,x_2,\dots ,x_n\left)\right)=r_1(x_1,0,\dots ,0)=r\left(x_1\right).$$  So $r_1(x_1,x_2,\dots ,x_n)=r\left(x_1\right)$  is a polynomial of one variable.

For arbitrary $u\in W\left(X\right)$  take $s\in \text{End}\left(W\right(X\left)\right)$  $s\left(x_1\right)=u$  , $s\left(x_i\right)=x_i$  for $i>1$  . Then $$\mu \left(u\right)=\mu s\left(x_1\right)=s\mu \left(x_1\right)=s\left(r\right(x_1\left)\right)=r\left(u\right),$$  and so $\mu \left(u\right)=r\left(u\right)=a_0+a_{1}u+\cdots +a_k{u}^k$  . Because $\mu$  is bijection, so $r$  is linear, Q.E.D.. Corollary. From the theorems 1 and 2 we have that in mentioned algebras every bijection $\mu$  , which generate an automorphism of $\text{End}\left(W\right(X\left)\right)$  transform every base of algebra $W\left(X\right)$  to a base of this algebra.

3 Linearity of automorphism generating bijection

Let $W=W\left(X\right)$  be a free commutative, associative or Lee algebra over the field $P$  with a finite set of generators, and $\tau$  be an automorphism of $\text{End}\left(W\right(X\left)\right)$  generated by bijection $\mu$  . We call $s\in \text{End}\left(W\right(X\left)\right)$  a constant endomorphism if $\text{Im}\left(s\right)=P$  . Clearly, if $s$  is constant then $s^{\tau }$  also is constant.

Denote the set of all constant endomorphisms by $Const$  .

Let $a\in P$  and $s$  be a constant endomorphism such that $s\left(x_1\right)=a$  . Then $\mu \left(a\right)=\mu \left(s\right(x_1\left)\right)=s^{\tau }\left(\mu \right(x_1\left)\right)\in P$  . So a restriction of $\mu$  on $P$  is a bijection:

$\mu :P\mapsto P$  .

Let $\mu \left(0\right)=a$  and $\mu \left(1\right)=b$  . We shall consider the linear bijection $l:W\left(X\right)\mapsto W\left(X\right)$  , $l\left(u\right)=cu+d$  such that $l\left(a\right)=0$  and $l\left(b\right)=1$  . Since linear bijection generates identical automorphism of the semigroup $\text{End}\left(W\right(X\left)\right)$  , so bijections $\mu$  and ${\mu }^{\prime }=l\mu$  generate equal automorphisms. Note that ${\mu }^{\prime }\left(0\right)=0$  and ${\mu }^{\prime }\left(1\right)=1$  . We denote $\mu ={\mu }^{\prime }$  and assume that $\mu \left(0\right)=0$  and $\mu \left(1\right)=1$  .

Since $\mu$  transforms base of algebra $W\left(X\right)$  to base, so the endomorphism $\eta :$  $\eta \left(x_i\right)=\mu \left(x_i\right)$  is an automorphism of $W\left(X\right)$  . Now we shall consider the bijection ${\mu }^{\prime }={\eta }^{-1}\mu$  . We have for ${\mu }^{\prime }$  that ${\mu }^{\prime }\left(x_i\right)=x_i$  .

Now we shall consider the Galois correspondence between ideals in $W\left(X\right)$  and algebraic sets in affine space over $W\left(X\right)$  . We recall that every ideal $T$  in $W\left(X\right)$  defines an algebraic set of points in affine space over $W\left(X\right)$  .

$$T^{\prime }=T_{W\left(X\right)}^{\prime }=A=\{\eta \in \text{End}(W\left(X\right)\left)\right|T\subset \text{Ker}\eta \}.$$  A set of points $A\subset \text{End}\left(W\right(X\left)\right)$  determines ideal in $W\left(X\right)$  :

$$A^{\prime }=A_W^{\prime }={\bigcap }_{\eta \in A}\text{Ker}\eta .$$  An element $u\in W\left(X\right)$  is called basic, if it belongs to some base of algebra $W\left(X\right)$  .

Lemma 3.1 If $u$  is an basic element of W(X), then $${\langle u\rangle }^{\prime \prime }=\langle u\rangle$$ 

Proof. Clearly, $$A={\langle u_1\rangle }^{\prime }=\{\eta \in \text{End}W|\eta \left(u_1\right)=0\}$$  and $$A^{\prime }={\langle u_1\rangle }^{\prime \prime }={\bigcap }_{\eta \in A}\text{Ker}\eta ={\bigcap }_{\eta \left(u_1\right)=0}\text{Ker}\eta .$$  Let $f(u_1,u_2,\dots ,u_n)\in {\langle u_1\rangle }^{\prime \prime }$  and $\eta :\eta \left(u_1\right)=0,\eta \left(u_i\right)=u_i$  for $i>1$  . Then $0=\eta \left(f\right(u_1,u_2,\dots ,u_n\left)\right)=f(0,u_2,\dots ,u_n)$  , so $f\in \langle u_1\rangle$  , Q.E.D..

Lemma 3.2 Let $\mu$  be an automorphism generating bijection of $W$  such that $\mu \left(0\right)=0,\mu \left(1\right)=1$  and $\mu \left(x_i\right)=x_i$  . Then $$\langle \mu \left(\alpha x_1\right)\rangle =\langle \mu \left(x_1\right)\rangle =\langle x_1\rangle$$ 

Proof. At first note that for any endomorphism $s\in \text{End}\left(W\right)$  $$s\left(x_1\right)=0\iff s^{\tau }\left(x_1\right)=0.$$  Really, if $s\left(x_1\right)=0$  , then $s^{\tau }\left(x_1\right)=s^{\tau }\mu \left(x_1\right)=\mu s\left(x_1\right)=0$  ; and if $s^{\tau }\left(x_1\right)=0$  , then $0=s^{\tau }\left(x_1\right)=s^{\tau }\mu \left(x_1\right)=\mu s\left(x_1\right)=0$  , i.e. $s\left(x_1\right)=0$  .

Let $\mu \left(\alpha x_1\right)=r_1(x_1,\dots ,x_n)$  . Take $s\in \text{End}\left(W\right)$  such that $s^{\tau }\left(x_1\right)=0$  and $s^{\tau }\left(x_i\right)=x_i$  for $i>1$  . Then $0=\mu \left(0\right)=\mu s\left(\alpha x_1\right)=s^{\tau }\mu \left(\alpha x_1\right)=s^{\tau }{r}_1(x_1,\dots ,x_n)=r_1(0,x_1,\dots ,x_n)$  , i.e. $\mu \left(\alpha x_1\right)\in {\langle x_1\rangle }^{\prime \prime }=\langle x_1\rangle$  .

Now let us prove, that $x_1\in \langle \mu \left(\alpha x_1\right)\rangle$  . Note that $$\langle \mu \left(\alpha x_1\right)\rangle =\langle \mu \left(\alpha x_1\right){\rangle }^{\prime \prime }={\bigcap }_{{\eta }^{\tau }:{\eta }^{\tau }\left(\mu \right(\alpha x_1\left)\right)=0}\text{Ker}{\eta }^{\tau }.$$  So we must prove that, if ${\eta }^{\tau }\left(\mu \right(\alpha x_1\left)\right)=0$  then ${\eta }^{\tau }\left(x_1\right)=0$  . Really, $$0={\eta }^{\tau }\left(\mu \right(\alpha x_1\left)\right)=\mu \eta \left(\alpha x_1\right)=\mu \left(\alpha \eta \right(x_1\left)\right)\Rightarrow \alpha \eta \left(x_1\right)=0\Rightarrow {\eta }^{\tau }\left(x_1\right)=0.$$  Q.E.D.

Lemma 3.3 Let $\mu$  be an automorphism generating bijection of $W$  such that $\mu \left(0\right)=0,\mu \left(1\right)=1$  and $\mu \left(x_i\right)=x_i$  .Then for every $a\in P$  and $u\in P\left[X\right]$  $$\mu \left(au\right)=\mu \left(a\right)\mu \left(u\right).$$ 

Proof. Since $\langle \mu \left(a{x}_1\right)\rangle =\langle x_1\rangle$  , so $\mu \left(a{x}_1\right)=b{x}_1=b\mu \left(x_1\right)$  , where $w\in P\left[X\right]$  .

Substitution $s\left(x_1\right)=1$  gives that $b=\mu \left(a\right)$  , i.e. $\mu \left(a{x}_1\right)=\mu \left(a\right)\mu \left(x_1\right)$  .

In associative case we consider only the case, where the field $P$  is of characteristic 0. I must note, that $\mu \left(a{x}_1\right)$  together with $x_2,\dots .x_n$  is a base of algebra, and so $\mu \left(a{x}_1\right)=b{x}_1+f(x_2,\dots ,x_n)$  (see [7]). Because $\mu \left(a{x}_1\right)\in \langle x_1\rangle$  , so $\mu \left(a{x}_1\right)=b{x}_1$  .

Lemma 3.4 With the preceding notation $$\mu (u+v)=\mu \left(u\right)+\mu \left(v\right).$$ 

Proof. Let $\mu (x_1+x_2)-\mu \left(x_1\right)-\mu \left(x_2\right)=f(x_1,\dots ,x_n)=g\left(\mu \right(x_1),\dots ,\mu (x_n\left)\right)$  , then $\mu (u_1+u_2)-\mu \left(u_1\right)-\mu \left(u_2\right)=g\left(\mu \right(u_1),\dots ,\mu (u_n\left)\right)$  . Substitution $s:s\left(u_1\right)=u_1,s\left(u_2\right)=u_2,s\left(u_i\right)=0$  for $i>2$  gives identity $$\mu (u_1+u_2)-\mu \left(u_1\right)-\mu \left(u_n\right)=\overline{g}\left(\mu \right(u_1),\mu (u_2\left)\right).$$  Clearly, $\overline{g}$  is homogenous of degree 1 of $u_1$  and $u_2$  . Therefore $\overline{g}(u_1,u_2)=a{u}_1+b{u}_2$  and $$\mu (u+v)=\mu \left(u\right)+\mu \left(v\right)+a\mu \left(u\right)+b\mu \left(v\right).$$  Substitutions $u=0$  and $v=0$  gives that $a=b=0$  , Q.E.D.. Corollary. Note that restriction $\mu$  on $P$  is an automorphism of $P$  and denote it $\eta$  . Let ${\mu }^{\prime }={\eta }^{-1}\mu$  , then ${\mu }^{\prime }$  is identical on $P$  . So for $\mu ={\mu }^{\prime }$  we have identity

$$\begin{array}{c}\mu \left(au\right)=a\mu \left(u\right),\end{array}$$

(1)

for every $a\in P$  and $u\in W\left(X\right)$  .

4 Description of $\text{Aut}\left(\text{End}\right(W\left(X\right)\left)\right)$ 

Let consider at first the commutative case.

Lemma 4.1 Let $\mu$  be a bijection of $P\left[X\right]$  generating an automorphism $\tau \in \text{Aut}\left(\text{End}\right(P\left[X\right]\left)\right)$  such that $\mu$  is identical on $P$  and $\mu \left(x_i\right)=x_i$  . Then for every $u,v\in P\left[X\right]$  it holds that $$\mu (u\cdot v)=\mu \left(u\right)\cdot \mu \left(v\right)$$ 

Proof. Let $s$  be a constant endomorphism of $P\left[X\right]$  .Then $s^{\tau }$  also is constant and so $$s^{\tau }\left(\mu \right(x_1{x}_2)-\mu (x_1\left)\mu \right(x_2\left)\right)=s^{\tau }\mu \left(x_1{x}_2\right)-s^{\tau }\mu \left(x_1\right)\cdot s^{\tau }\mu \left(x_2\right)=$$  $$\mu s\left(x_1{x}_2\right)-\mu s\left(x_1\right)\cdot \mu s\left(x_2\right)=\mu \left(a_1{a}_2\right)-\mu \left(a_1\right)\cdot \mu \left(a_2\right)=0,$$  $$\mu \left(x_1{x}_2\right)-\mu \left(x_1\right)\mu \left(x_2\right)\in {\bigcap }_{s^{\tau }\in Const}\text{Ker}{s}^{\tau }=0,$$  Q.E.D.. Now we can formulate the main result for commutative case.

Theorem 3 Every automorphism of $\text{End}\left(P\right[X\left]\right)$  is semiinner.

Proof. It follows from lemma 3.4, lemma 4.1 and identity (1).

Now let consider associative case.

Lemma 4.2 Let $\mu$  be a bijection of $W\left(X\right)$  generating an automorphism $\tau \in \text{Aut}\left(\text{End}\right(W\left(X\right)\left)\right)$  such that ${\mu }^{\prime }$  is identical on $P$  and $\mu \left(x_i\right)=x_i$  . Then for every $u,v\in W\left(X\right)$  it holds that $$\mu (u\cdot v)=\mu \left(u\right)\cdot \mu \left(v\right),$$  or for every $u,v\in W\left(X\right)$  it holds that $$\mu (u\cdot v)=\mu \left(v\right)\cdot \mu \left(u\right),$$ 

Proof. Let $\mu \left(x_1{x}_2\right)-\mu \left(x_1\right)\mu \left(x_1\right)=f(x_1,\dots ,x_n)=g\left(\mu \right(x_1),\dots ,\mu (x_n\left)\right)=\overline{g}\left(\mu \right(x_1),\mu (x_2\left)\right)$  . Note that $\overline{g}$  is 1-homogenous of $x_1$  and 1-homogenous of $x_2$  .

Following the proof of lemma 3.4 we get identity $$\mu \left(uv\right)=a\mu \left(u\right)\mu \left(v\right)+b\mu \left(v\right)\mu \left(u\right).$$  Substitution $v=1$  gives equality $a+b=1$  . We shall denote $\mu \left(u\right)=\overline{u}$  and calculate $\mu \left(xxy\right)$  .

$$\mu \left(xxy\right)=a\overline{xx}\cdot \overline{y}+b\overline{y}\cdot \overline{xx}=$$  $$a(a+b)\overline{x}\cdot \overline{x}\cdot \overline{y}+b(a+b)\overline{y}\cdot \overline{x}\cdot \overline{x}=$$  $$a\overline{x}\cdot \overline{x}\cdot \overline{y}+b\overline{y}\cdot \overline{x}\cdot \overline{x},$$  $$\mu \left(xxy\right)=a\overline{x}\cdot \overline{xy}+b\overline{xy}\cdot \overline{x}=$$  $$a^2\overline{x}\cdot \overline{x}\cdot \overline{y}+2ab\overline{x}\cdot \overline{y}\cdot \overline{x}+b^2\overline{y}\cdot \overline{x}\cdot \overline{xy}.$$  So $a=a^2,b=b^2,ab=0,a+b=1$  and we have two solutions $a=1,b=0$  or $a=0,b=1$  , Q.E.D.. Now we can formulate the main result for associative algebras.

Theorem 4 The group $\text{Aut}\left(\text{End}\right(W\left(X\right)\left)\right)$  is generated by semiinner automorphisms and mirror automorphism.

Proof. It follows from lemma 3.4, lemma 4.2 and identity (1).

Acknowledgments The author is happy to thank professor B. Plotkin for stimulating discussions of the results.

References

• 1. A. Berzins, The automorphisms of End $K\left[x\right]$  , Proc. Latvian Acad. Sci., Section B, 2003, vol. 57, no. 3/4, pp. 78-81
• 2. A. Berzins, Geometric equivalence of algebras, Int. J. Alg. Comput., 2001, vol. 11, no. 4, pp. 447-456.
• 3. A. Berzins, The group of automorphisms of semigroup End( $P\left[X\right]$  ,(to appear)
• 4. A. Berzins, B. Plotkin and E. Plotkin, Algebraic geometry in Varieties with the Given Algebra of constants, J. Math. Sci., New York, 2000, vol. 102, no. 3, pp. 4039-4070
• 5. R. Lipyanski, B. Plotkin Automorphisms of categories of free modules and free Lie algebras, ArXiv math.RA/0502212 (10 Feb 2005) pp. 14.
• 6. Y. Katsov, R. Lipyanski, B. Plotkin, Automorphisms of categories of free modules, free semimodules and free Lie modules, (to appear)
• 7. L. Makar-Limanov Algebraically closed skew field, Journal of Algebra, Vol. 93, 1985, pp. 117-135.
• 8. G. Mashevitzky, Automorphisms of categories of free associative algebras., Preprint.
• 9. G. Mashevitzky, B. Plotkin, E. Plotkin, Automorphisms of categories of free Lie algebras., Journal of Algebra, Vol. 282, 2004, pp. 490-512.
• 10. B. Plotkin, Algebraic logic, varieties of algebras and algebraic varieties, Proc. Int. Alg. Conf., St. Petersburg, 1995, Walter de Gruyter, New York, London, 1996.
• 11. B. Plotkin, Varieties of algebras and algebraic varieties, Israel J. of Mathematics, 1996, vol. 96, no. 2, pp. 511-522.
• 12. B. Plotkin, Varieties of algebras and algebraic varieties. Categories of algebraic varieties, Sib. Adv. Math., 1997, vol. 7, no. 2, pp. 64-97.
• 13. B. Plotkin, Algebras with the same (algebraic) geometry, Proc. Steklov Inst. Math, Vol. 242, 2003, pp. 165-196.
• 14. B. Plotkin, G Zhitomirski, On automorphisms of categories of universal algebras, ArXiv math.CT/0411408 (18 Nov 2004). pp. 24.