The foundation of the classical geometry of unit vector fields was proposed by A.Voss at the end of the nineteenth century. The theory includes the Gaussian and the mean curvature of a vector field and their generalizations (see [1] for details).

Recently, the geometry of vector fields has been considered from another point of view. Let $T_{1}M$  be a unit tangent bundle of $M$  endowed with the Sasaki metric [15] . If $\xi$  is a unit vector field on $M$  , then one may consider $\xi$  as a mapping $\xi :M\to T_{1}M$  . The image $\xi \left(M\right)$  is a submanifold in $T_{1}M$  with metric induced from $T_{1}M$  and one may apply the methods from the study of the geometry of submanifolds to determine geometrical characteristics of a unit vector field. A unit vector field $\xi$  is said to be minimal if $\xi \left(M\right)$  is a minimal submanifold in $T_{1}M$  . A unit vector field on $S^3$  tangent to the fibers of the Hopf fibration $S^3\stackrel{S^1}{⟶}{S}^2$  is a unique unit vector field with globally minimal volume [11] . This result fails in higher dimensions. A lower volume is achieved by a vector field with one singular point, namely the inverse image under stereographic projection inverse image of a parallel vector field on $E^n$  [14] . The lowest volume is reached for the North-South vector field with two singular points [6] .

A local approach to minimality of unit vector fields was developed in [7] .

A number of examples of locally minimal unit vector fields was found [2 – 4, 7 – 10, 12 – 14, 16 – 18] on various manifolds. In [19] the author presented an explicit expression for the second fundamental form of $\xi \left(M\right)$  and found some examples of vector fields with constant mean curvature. This expression is the key to solving a problem about totally geodesic vector fields on a given Riemannian manifold. Originally, the problem of a full description of all totally geodesic submanifolds in the tangent (sphere) bundle of spaces of constant curvature was posed by A.Borisenko in [5] . The totally geodesic vector fields form a special class of such submanifolds. In [20] this problem was solved in the case of 2-manifolds of constant curvature. In [22] an example of a totally geodesic unit vector field was found on a surface of revolution with non-constant but sign-preserving Gaussian curvature.

In this paper, we completely determine the Riemannian 2-manifolds which admit a unit vector field $\xi$  such that $\xi \left(M\right)$  is a totally geodesic submanifold in $T_{1}M$  . Moreover, we explicitly determine the vector field.

Under some restrictions, we find an isometric immersion of the metric into Euclidean 3-space which gives a surface with the necessary properties.

1 The main result

Let $\xi$  be a unit vector field on a Riemannian manifold $(M^n,g)$  . Then $\xi$  can be considered as a mapping $\xi :M^n\to T_1{M}^n$  . In this way one can use geometrical properties of the submanifold $\xi \left(M^n\right)$  to determine the geometrical characteristics of the vector field.

Definition 1.1 A unit vector field on Riemannian manifold $M^n$  is said to be totally geodesic, if the submanifold $\xi \left(M^n\right)\subset T_1{M}^n$  is totally geodesic in the unit tangent bundle with the Sasaki metric.

Definition 1.2 A point $q\in M^n$  is said to be stationary for the vector field $\xi$  if ${\nabla }_X\xi {|}_q=0$  for all $X\in T_q{M}^n$  .

If stationary points fills a domain $D\subset M^n$  , then locally $M^n=M^{n-k}\times E^k$  , where $E^k$  is a Euclidean factor of dimension $k\ge 1$  . In the case $n=2$  , the manifold is then flat in $D$  . If the manifold is of sign-preserving Gaussian curvature, then we can always restrict our considerations to the domain with no stationary points of a given unit vector field. The main result of the paper is the following theorem.

Theorem 1.1 Let $M^2$  be a Riemannian manifold with sign-preserving Gaussian curvature $K$  . Then , on some open subset U of M, there exists a unit totally geodesic vector field $\xi$  if and only if

• (a) the metric $g$  on $U$  is locally of the form $$d{s}^2=d{u}^2+{sin}^2\alpha \left(u\right)d{v}^2,$$  where $\alpha \left(u\right)$  solves the differential equation $\frac{d\alpha }{du}=1-\frac{a+1}{cos\alpha };$ 
• (b) the totally geodesic unit vector field $\xi$  is of the form $$\xi =cos(av+{\omega }_0){\partial }_u+\frac{sin(av+{\omega }_0)}{sin\alpha \left(u\right)}{\partial }_v,$$  where $a,{\omega }_0=const$  .

Remark. The Gaussian curvature $K$  of the metric is

$$\begin{array}{c}K=\frac{d\alpha }{du}.\end{array}$$

(1)

Therefore, $\alpha \left(u\right)$  is the total curvature of the manifold along the meridian of the metric. The vector field is parallel along meridians and bends along parallels with constant angle speed $a$  with respect to the coordinate frame.

Proof. Let $\xi$  be a given unit vector field on Riemannian manifold $M^n$  .

For dimension reasons, the kernel of the linear operator ${\nabla }_X\xi :T{M}^n\to {\xi }^{\perp }$  is not empty. Therefore, there is a non-zero vector field $e_0$  such that ${\nabla }_{e_0}\xi =0$  .

In the case $n=2$  , the field $e_0$  can be found explicitly. Denote by $\eta$  a unit vector field on $M^2$  which is orthogonal to $\xi$  . Set $${\nabla }_{\xi }\xi =k\eta ,{\nabla }_{\eta }\eta =\mathbb{\varkappa }\xi ,$$  where $k$  and $\mathbb{\varkappa }$  are the signed geodesic curvatures of the integral trajectories of the fields $\xi$  and $\eta$  respectively. Introduce an orthonormal frame $$e_0=\frac{\mathbb{\varkappa }}{\lambda }\xi +\frac{k}{\lambda }\eta ,e_1=\frac{k}{\lambda }\xi -\frac{\mathbb{\varkappa }}{\lambda }\eta ,\lambda =\sqrt{k^2+{\mathbb{\varkappa }}^2}.$$  The fields $e_0$  and $e_1$  are correctly defined on an open subset $U\subset M^2$  where the field $\xi$  has no stationary points, i.e., points where $\lambda =0$  . Restrict ourselves to this open part. It is elementary to check that

$$\begin{array}{c}{\nabla }_{e_0}\xi =0,{\nabla }_{e_1}\xi =\lambda \eta .\end{array}$$

(2)

Denote by $\omega$  the angle function between $\xi$  and $e_0$  . Then

$$\begin{array}{c}k=\lambda sin\omega ,\mathbb{\varkappa }=\lambda cos\omega \end{array}$$

(3)

and we can set

$$\begin{array}{c}\begin{array}{c}\xi =cos\omega e_0+sin\omega e_1,\\ \eta =sin\omega e_0-cos\omega e_1.\end{array}\end{array}$$

(4)

Denote by $\mu$  and $\sigma$  the signed geodesic curvatures of the integral curves of the fields $e_0$  and $e_1$  respectively. Then $${\nabla }_{e_0}{e}_0=\mu e_1,{\nabla }_{e_1}{e}_1=\sigma e_0.$$  In these terms, the second fundamental form of the submanifold $\xi \left(M\right)\subset T_{1}M$  can be expressed as [20] 

$$\begin{array}{c}\Omega =\left[\begin{array}{cc}-\mu \frac{\lambda }{\sqrt{1+{\lambda }^2}}& \frac{1}{2}\left(\sigma \lambda +\frac{1-{\lambda }^2}{1+{\lambda }^2}{e}_0\left(\lambda \right)\right)\\ \frac{1}{2}\left(\sigma \lambda +\frac{1-{\lambda }^2}{1+{\lambda }^2}{e}_0\left(\lambda \right)\right)& e_1\left(\frac{\lambda }{\sqrt{1+{\lambda }^2}}\right)\end{array}\right].\end{array}$$

(5)

Set $$cos(\alpha /2)=\frac{1}{\sqrt{1+{\lambda }^2}}.$$  Then we have

$$\begin{array}{c}\frac{\lambda }{\sqrt{1+{\lambda }^2}}=sin(\alpha /2),\frac{1-{\lambda }^2}{1+{\lambda }^2}=cos\alpha ,\end{array}$$

$$\begin{array}{c}{e}_0\left(\lambda \right)=\frac{e_0\left(\alpha \right)}{2{cos}^2(\alpha /2)},e_1\left(\frac{\lambda }{\sqrt{1+{\lambda }^2}}\right)=\frac{1}{2}cos(\alpha /2)e_1\left(\alpha \right).\end{array}$$

After these simplifications $$\Omega =\frac{1}{2}\left[\begin{array}{cc}-2\mu sin(\alpha /2)& \frac{\sigma sin\alpha +e_0\left(\alpha \right)cos\alpha }{2{cos}^2(\alpha /2)}\\ \frac{\sigma sin\alpha +e_0\left(\alpha \right)cos\alpha }{2{cos}^2(\alpha /2)}& cos(\alpha /2)e_1\left(\alpha \right)\end{array}\right].$$  Set $\Omega \equiv 0$  . Then $\mu \equiv 0$  , since $sin(\alpha /2)\equiv 0$  implies $\lambda \equiv 0$  , which contradicts the hypothesis. Therefore, a if totally geodesic vector field exists, then the integral trajectories of the field $e_0$  are geodesics.

Since $cos(\alpha /2)\ne 0$  , then

$$\begin{array}{c}{e}_1\left(\alpha \right)\equiv 0.\end{array}$$

(6)

Introduce a local semi-geodesic coordinate system $(u,v)$  such that $${\partial }_u=e_0,{\partial }_v=f(u,v)e_1,$$  where $f(u,v)$  is some non-zero function. Then the line element of $M^2$  can be written as $$d{s}^2=d{u}^2+f^{2}d{v}^2$$  The condition  6 implies ${\partial }_v\alpha =0$  , which means that $\alpha =\alpha \left(u\right)$  .

Consider now the last condition $$\sigma sin\alpha +e_0\left(\alpha \right)cos\alpha =0.$$  If $cos\alpha \equiv 0$  , then $sin\alpha \equiv 1$  and hence $\sigma \equiv 0$  . This means that $e_0$  is a parallel vector field on $M^2$  and hence $K=0$  again. Set $$\sigma tan\alpha +e_0\left(\alpha \right)=0.$$  With respect to the chosen semi-geodesic coordinate system, $\sigma =-{\partial }_{u}f/f$  and we come to the following relation $$\frac{{\partial }_{u}f}{f}=cot\alpha {\partial }_u\alpha .$$  Because of  6 , we have $\alpha =\alpha \left(u\right)$  and the equation above has an evident solution $$f(u,v)=C\left(v\right)sin\alpha ,$$  where $C\left(v\right)\ne 0$  is a constant of integration. Making a $v$  parameter change one can always set $C\left(v\right)\equiv 1$  . Therefore, the line element of a 2-manifold $M$  which admits a totally geodesic vector unit field is necessarily of the form

$$\begin{array}{c}d{s}^2=d{u}^2+{sin}^2\alpha \left(u\right)d{v}^2.\end{array}$$

(7)

Turn now to the vector field. A direct computation yields

$$\begin{array}{c}{\nabla }_{e_0}\xi ={\nabla }_{e_0}(cos\omega e_0+sin\omega e_1)=(-e_0(\omega )-\mu )\eta ,\end{array}$$

$$\begin{array}{c}{\nabla }_{e_1}\xi ={\nabla }_{e_1}(cos\omega e_0+sin\omega e_1)=(-e_1(\omega )+\sigma )\eta .\end{array}$$

Since $\mu =0$  and ${\nabla }_{e_0}\xi =0$  , we see that ${\partial }_u\omega =0$  and hence $\omega =\omega \left(v\right)$  . The second equality means, that $$-e_1\left(\omega \right)+\sigma =tan(\alpha /2).$$  With respect to a chosen coordinate system, we have $$\sigma =-cot\alpha {\partial }_u\alpha$$  and hence $${\partial }_v\omega =sin\alpha (\sigma -tan(\alpha /2)=-cos\alpha {\partial }_u\alpha -2{sin}^2(\alpha /2)$$  The right hand side does not depend on the $v$  parameter and therefore ${\partial }_{vv}^2\omega =0$  which means that $$\omega =av+{\omega }_0,(a,{\omega }_0=const).$$  As a consequence, we come to the following differential equation for the function $\alpha \left(u\right)$  :

$$cos\alpha {\partial }_u\alpha +2{sin}^2(\alpha /2)=-a$$  or equivalently

$$\begin{array}{c}\frac{d\alpha }{du}=1-\frac{a+1}{cos\alpha }.\end{array}$$

(8)

The proof is complete.

Remark. A direct computation shows that if $\alpha$  is a solution of  8 , then Gaussian curvature of the metric  7 takes the form  1 . Since it is supposed that $K$  is sign-preserving, the relation  1 allows to choose $\alpha$  as a new parameter on $u$  -curves. With respect to the parameter $\alpha$  we have $$du=\frac{d\alpha }{K}=-\frac{cos\alpha }{a+1-cos\alpha }d\alpha$$  and the line element  7 takes the form

$$\begin{array}{c}d{s}^2={\left(\frac{cos\alpha }{a+1-cos\alpha }\right)}^{2}d{\alpha }^2+{sin}^2\alpha d{v}^2.\end{array}$$

(9)

Remark. If $\xi$  is a unit vector field on the Riemannian manifold $M^n$  , then the induced metric on $\xi \left(M^n\right)$  is $d{\tilde{s}}^2=g_{ik}d{u}^{i}d{u}^k+<{\nabla }_i\xi ,{\nabla }_k\xi >d{u}^{i}d{u}^k.$  If $\xi$  is a totally geodesic vector field on $M^2$  , then the metric of $M^2$  has the standard form  7 and ${\nabla }_{{\partial }_u}\xi ={\nabla }_{e_0}\xi =0,{\nabla }_{{\partial }_v}\xi =sin\alpha {\nabla }_{e_1}\xi =sin\alpha \lambda \eta =2{sin}^2(\alpha /2)\eta .$  Thus, we have $$d{\tilde{s}}^2=d{u}^2+{sin}^2\alpha d{v}^2+4{sin}^4(\alpha /2)d{v}^2=d{u}^2+4{sin}^2(\alpha /2)d{v}^2.$$  Taking into account  1 we can easily find the Gaussian curvature of the totally geodesic submanifold $\xi \left(M^2\right)$  , namely $$\tilde{K}=\frac{1}{4}K(K-2cot(\alpha /2\left)K_{\alpha }^{\prime }\right),$$  where $K\left(\alpha \right)$  is the Gaussian curvature of $M^2$  given by relations  1 and  8 .

The equations  8 and  1 completely determine the class of Riemannian 2-dimensional manifolds admitting a totally geodesic unit vector field.

Proposition 1.1 Let $M^2$  be a Riemannian manifold with a line element of the form $$d{s}^2=d{u}^2+{sin}^2\alpha \left(u\right)d{v}^2.$$  Denote by $K$  the Gaussian curvature of $M^2$  . Then $K=\frac{d\alpha }{du}$  if and only if the function $\alpha \left(u\right)$  satisfies $$\frac{d\alpha }{du}=1+\frac{m}{cos\alpha }(m=const).$$ 

Proof. The sufficient part is already proved. Suppose now that $$\frac{d\alpha }{du}=K(\ne 0).$$  Then we have $${\alpha }^{\prime }=K=-\frac{{\partial }_{uu}(sin\alpha )}{sin\alpha }=({\alpha }^{\prime }{)}^2-cot\alpha {\alpha }^{\prime \prime }.$$  Therefore, ${\alpha }^{\prime \prime }=-{\alpha }^{\prime }(1-{\alpha }^{\prime })tan\alpha ,$  or $\frac{{\alpha }^{\prime \prime }}{{\alpha }^{\prime }-1}={\alpha }^{\prime }tan\alpha ,$  or $$(ln|{\alpha }^{\prime }-1|{)}^{\prime }=-(ln|cos\alpha |{)}^{\prime }.$$  Evidently, now $|{\alpha }^{\prime }-1|=\frac{\left|m\right|}{|cos\alpha |}$  where $m=const$  is a constant of integration. Finally, $\frac{d\alpha }{du}=1+\frac{m}{cos\alpha }.$ 

Corollary 1.1 Let $M^2$  be a Riemannian manifold of constant curvature $c\ne 0$  . Then $M^2$  admits a totally geodesic unit vector field if and only if $c=1$  . This vector field is parallel along meridians and moves along parallels with unit angle speed.

Proof. If $K=c=const$  , then  1 can be satisfied if and only if $c=1,a=-1$  .

The equation  1 implies an elementary non-existence result.

Corollary 1.2 Let $M^2$  be a Riemannian manifold with Gaussian curvature $K$  . Then $M^2$  does not admit a totally geodesic unit vector field $\xi$  with angle speed $a$  if $|K-1|<|a+1|$  .

Proof. Indeed, one can easily see that $cos\alpha =\frac{a+1}{1-K}.$  If $|a+1|>|K-1|$  , then we come to a contradiction.

2 Integral tra jectories of the totally geodesic vector field

The integral trajectories of the totally geodesic vector field $\xi$  can be found easily as follows. Let $\gamma =\left\{u\right(s),v(s\left)\right\}$  be an integral trajectory. Since $$\xi =cos\omega e_0+sin\omega e_1=cos\omega {\partial }_u+\frac{sin\omega }{sin\alpha }{\partial }_v,$$  we can set $$\frac{du}{ds}=cos\omega ,\frac{dv}{ds}=\frac{sin\omega }{sin\alpha }$$  and then $$\frac{du}{dv}=cot\omega sin\alpha .$$  Since $\alpha =\alpha \left(u\right)$  and $\omega =av+{\omega }_0$  , we come to the equation with separable variables $$\frac{du}{sin\alpha }=cot\omega dv.$$  Using  8 , we can find $$\frac{du}{d\alpha }=\frac{cos\alpha }{-a-1+cos\alpha }$$  and make a parameter change in the left hand side of the equation above.

Then we come to the equation $$\frac{cos\alpha d\alpha }{sin\alpha (-a-1+cos\alpha )}=cot\omega dv.$$  Taking primitives, we have

$$\begin{array}{cc}tan(\alpha /2)sin(av+{\omega }_0)=c(a+(a+2\left){tan}^2\right(\alpha /2){)}^{\frac{a+1}{a+2}}& \text{for}a\ne 0,-2,\end{array}$$

$$\begin{array}{cc}tan(\alpha /2)sin(-2v+{\omega }_0)=c{e}^{\frac{1}{2}{tan}^2(\alpha /2)}& \text{for}a=-2,\end{array}$$

$$\begin{array}{cc}\frac{1}{2}tan{\omega }_0\left(\frac{1}{1-cos\alpha }+ln|tan(\alpha /2\left)\right|\right)=v-c& \text{for}a=0.\end{array}$$

Taking into account  3 , we remark that $tan(\alpha /2)sin\omega =k$  and ${tan}^2(\alpha /2)=k^2+{\mathbb{\varkappa }}^2$  . Therefore, we have an intrinsic equation on the integral curves of the totally geodesic vector field

$$\begin{array}{cc}k=c{\left[a+(a+2)(k^2+{\mathbb{\varkappa }}^2)\right]}^{\frac{a+1}{a+2}}& \text{for}a\ne 0,-2,\end{array}$$

$$\begin{array}{cc}k=c{e}^{\frac{1}{2}(k^2+{\mathbb{\varkappa }}^2)}& \text{for}a=-2,\end{array}$$

$$\begin{array}{cc}k=sin{\omega }_{0}exp\left[2cot{\omega }_0(v-c)-\frac{1}{2}\frac{1+k^2+{\mathbb{\varkappa }}^2}{k^2+{\mathbb{\varkappa }}^2}\right]& \text{for}a=0,\end{array}$$

where $c$  is a constant of integration.

Moreover, in any case

$$\xi \left(k\right)=cos\omega {\partial }_u[tan(\alpha /2)sin\omega ]+\frac{sin\omega }{sin\alpha }{\partial }_v[tan(\alpha /2)sin\omega ]=\frac{cos\omega sin\omega {\alpha }_u^{\prime }}{2{cos}^2(\alpha /2)}+\frac{asin\omega cos\omega tan(\alpha /2)}{sin\alpha }=\frac{cos\omega sin\omega }{2{cos}^2(\alpha /2)}({\alpha }_u^{\prime }+a).$$ 

The equation  8 yields $$\xi \left(k\right)=\frac{(a+1)cos\omega sin\omega }{2{cos}^2(\alpha /2)}\left(1-\frac{1}{cos\alpha }\right).$$  Thus, if $a=-1$  , then the integral trajectories of the field $\xi$  form a family of circles. The metric of $M^2$  is $$d{s}^2=d{u}^2+{sin}^{2}ud{v}^2$$  and we are dealing with the unit sphere parameterized by $$r=\left\{sinucosv,sinusinv,cosu\right\}.$$  These circles satisfy

$$\begin{array}{c}tan(u/2)sinv=c.\end{array}$$

(10)

Let $(\rho ,\phi )$  be polar coordinates in a Cartesian plane which passes through the center of the sphere such that $(0,0,1)$  is the north pole on the sphere.

Then the parameters $(\rho ,\phi )$  and $(u,v)$  are connected via stereographic projection from the south pole as

$$\begin{array}{c}\rho =tan(u/2),\end{array}$$

$$\begin{array}{c}\phi =v.\end{array}$$

Therefore, the equation  10 defines a family of parallel straight lines on the Cartesian plane. The family of integral curves of a totally geodesic vector field on the unit sphere can be obtained as inverse images under stereogrphic projection of this family.

An explicit equation of this family is $$r\left(v\right)=\left\{\frac{2csinvcosv}{c^2+{sin}^{2}v},\frac{2c{sin}^{2}v}{c^2+{sin}^{2}v},-\frac{c^2-{sin}^{2}v}{c^2+{sin}^{2}v}\right\}$$  where $c$  is the geodesic curvature of the corresponding circle. All of these circles pass through the south pole $(0,0,-1)$  when $v=0,\pi$  . We can find this by using the expression $tan(u/2)=c/sinv$  and trigonometric expressions for $sinu$  and $cosu$  via $tan(u/2)$  .

The unit sphere is not the unique surface that realizes the metric  9 . Let $(x,y,z)$  be standard Cartesian coordinates in $E^3$  . We can find an isometric immersion of the metric  9 into $E^3$  in a class of a surfaces of revolution. To do this, set

$$\begin{array}{c}x\left(\alpha \right)=sin\alpha ,\end{array}$$

$$\begin{array}{c}(x_{\alpha }^{\prime }{)}^2+(z_{\alpha }^{\prime }{)}^2={\left(\frac{cos\alpha }{a+1-cos\alpha }\right)}^2\end{array}$$

and we easily find

$$\begin{array}{c}x\left(\alpha \right)=sin\alpha ,\end{array}$$

$$\begin{array}{c}z\left(\alpha \right)={\int }_{{\alpha }_0}^{\alpha }\frac{cost}{a+1-cost}\sqrt{1-(a+1-cost{)}^2}dt,\end{array}$$

where the interval of integration is limited by the restrictions $$\{\begin{array}{c}1+a<cos\alpha <2+a,\\ -2<a<-1,\end{array}\text{or}\{\begin{array}{c}a<cos\alpha <1+a,\\ -1<a<0.\end{array}$$  The restrictions mean that if $|a+1|\ge 1$  , then the metric  9 does not admit an isometric immersion into $E^3$  in a class of surfaces of revolution. References

1. Aminov Yu. The geometry of vector fields. Gordon&Breach Publ., 2000.
2. Boeckx E., Vanhecke L. Harmonic and minimal radial vector fields. Acta Math. Hungar. 90 (2001), 317-331.
3. Boeckx E., Vanhecke L. Harmonic and minimal vector fields on tangent and unit tangent bundles. Differential Geom. Appl. 13 (2000), 77-93.
4. Boeckx E., Vanhecke L. Isoparametric functions and harmonic and minimal unit vector fields, Contemp. Math. 288 (2001), 20–31.
5. Borisenko A., Yampolsky A. Riemannian geometry of bundles. Uspehi Mat. Nauk, 26/6 (1991), 51-95; Engl. transl.: Russian Math. Surveys, 46/6 (1991), 55–106.
6. Brito F., Chacon P., Naveira A. On the volume of vector fields on spaces of constant sectional curvature, Preprint, 2001.
7. Gil-Medrano O.,Llinares-Fuster E. Minimal unit vector fields, Tôhoku Math. J. 54 (2002), 71 – 84.
8. Gil-Medrano O.,Llinares-Fuster E. Second variation of volume and energy of vector fields. Stability of Hopf vector fields, Math. Ann. 320 (2001), 531-545.
9. Gil-Medrano O. Relationship between volume and energy of unit vector fields, Diff. Geom. Appl. 15 (2001), 137-152.
10. Gil-Medrano O., González-Dávila J.C., Vanhecke L. Harmonic and minimal invariant unit vector fields on homogeneous Riemannian manifolds, Houston J. Math. 27 (2001), 377-409.
11. Gluck H., Ziller W. On the volume of a unit vector field on the three-sphere. Comm. Math. Helv. 61 (1986), 177-192.
12. González-Dávila J.C., Vanhecke L. Examples of minimal unit vector fields. Ann Global Anal. Geom. 18 (2000), 385-404.
13. González-Dávila J.C., Vanhecke L. Minimal and harmonic characteristic vector fields on three-dimensional contact mentic manifolds. J. Geom. 72 (2001), 65-76.
14. Pedersen S. L. Volumes of vector fields on spheres, Trans. Amer. Math. Soc. 336 (1993), 69-78.
15. Sasaki S. On the differential geometry of tangent bundles of Riemannian manifolds. Tôhoku Math. J. 10 (1958), 338-354.
16. Tsukada K.,Vanhecke L. Minimality and harmonicity for Hopf vector fields, Illinois J. Geom. 45 (2001), 441 – 451.
17. Tsukada K.,Vanhecke L. Invariant minimal unit vector fields on Lie groups, Period. Math. Hungar. 40 (2000), 123-133.
18. Tsukada K.,Vanhecke L. Minimal and harmonic vector fields on $G_2\left(C^{m+2}\right)$  and its dual space. Monatsh. Math. 130 (2000), 143-154.
19. Yampolsky A. On the mean curvature of a unit vector field, Math. Publ. Debrecen, 60/1-2 (2002), 131 – 155.
20. Yampolsky A. On the intrinsic geometry of a unit vector field, Comment. Math. Univ. Carol. 43/2 (2002), 299-317.
21. Yampolsky A. A totally geodesic property of Hopf vector fields, Acta Math. Hungar. 101/1-2 (2003), 93-112.
22. Yampolsky A. On extrinsic geometry of unit normal vector field of Riemannian hyperfoliation, Math. Publ. Debrecen 63/4 (2003), 555 – 567.

Department of Geometry, Faculty of Mechanics and Mathematics, Kharkiv National University, Svobody Sq. 4, 61077, Kharkiv, Ukraine. e-mail: yamp@univer.kharkov.ua