## A condition for a Riemannian manifold to be a hyperbolic space

[Update: A counter example when $\phi>|\nabla\phi|$ is dropped and another version of this theorem by Pong Ting Kei (15-10-2011)]

In this post I will give a characterization when a complete connected Riemannian manifolds ${(M,g)}$ is isometric to the hyperbolic space. Of course this is the analogue of Obata’s theorem for the necessary and sufficient condition for a manifold to be a sphere.

Recall that ${\mathbb{R}^{n,1}=\{(x_0,x_1,\cdots, x_n)\in \mathbb{R}^{n+1}\}}$ is the Minkowski spacetime with metric ${-dx_0^2+dx_1^2+\cdots +dx_n^2}$ and the hyperbolic space is

$\displaystyle \mathbb{H}^n = \{ (x_0,\cdots, x_n)\in \mathbb{R}^{n,1}: x_0 ^2 -x_1^2-\cdots -x_n^2 =1, x_0>0\}$

with induced metric. In geodesic polar coordinates, the metric can be expressed as

$\displaystyle ds^2 + \sinh^2 s\, d\theta^2$

where ${s}$ is the distance function from the origin and ${d\theta^2}$ is the metric on the standard unit ball. It can be checked that ${\mathbb{H}^n}$ has constant sectional curvature ${-1}$, i.e. a space form.

We can now state the Obata-type theorem for hyperbolic space:

 Theorem 1 Let ${(M^n,g)}$ be a complete connected Riemannian manifold. Then it is isometric to the hyperbolic space if and only if there is a non-constant smooth function ${\phi}$ on ${M}$ such that $\displaystyle \nabla ^2 \phi = \phi g \ \ \ \ \ (1)$ and $\displaystyle \phi> |\nabla \phi|.$

The proof is quite similar to Obata’s proof, with some modification.
Proof: The necessary condition is clear. Indeed, for the position function ${X\in \mathbb{R}^{n,1}}$ of ${\mathbb{H}^n}$, note that ${X}$ can be regarded as the normal ${N}$ of the hypersurface ${\mathbb{H}^n\subset \mathbb{R}^{n,1}}$, using the fact that ${\partial_u X=u}$ for any tangent vector ${u}$ of ${\mathbb{H}^n}$ , together with ${\nabla_u v= (\partial _u v)^T= \frac{\partial ^2X}{\partial u\partial v}+\langle \frac{\partial ^2X}{\partial u\partial v}, X\rangle X}$ , it is easy to see that the second fundamental form ${h(u,v)= \langle \partial _uX,v\rangle= g(u,v)}$ and so

$\displaystyle \nabla ^2 X(u,v)= h(u,v)N= g(u,v)X.$

Thus if we take ${\phi=x_0}$, then ${ \nabla ^2 \phi = \phi g}$ and ${\phi^2 -|\nabla \phi|^2=\cosh ^2 s-\sinh ^2 s=1}$.

For the converse, suppose there is such ${\phi}$. Take any ${q}$ such that ${\nabla\phi(q)\neq0}$. Then for any arclength-parametrized geodesic ${\gamma}$ starting from this point, the equation is reduced to

$\displaystyle \phi''=\phi,$

so the solution is

$\displaystyle \phi =A\cosh s+ B \sinh s,$

where ${A= \phi(q)}$ and ${B= \nabla _{\gamma'(0)}\phi}$. Using the compound angle formula, it can also be written as

$\displaystyle \phi (s) =\sqrt{\phi(q)^2-|\nabla _{\gamma'}\phi|^2} \cosh (s+\theta)$

where ${\tanh \theta =B/A}$. [This is where we have to use the condition ${\phi> |\nabla \phi|}$. ] From this form we see that when we choose ${\gamma'(0)=\frac{\nabla\phi(q)}{|\nabla\phi(q)|}}$ and when ${s=s_0=-\tanh^{-1}(\frac{ |\nabla \phi(q)|}{\phi(q)})}$, then ${\phi(\gamma(s_0))}$ is actually the unique minimum of ${\phi}$ over ${M}$. (The uniqueness of the minimum can also be seen from the fact that $\phi$ is strictly convex in view of (1). )

Now let ${p=\phi(\gamma (s_0))}$, we can further rescale ${\phi}$ such that for any arclength-parametrized geodesic ${\gamma (s)}$ from this point

$\displaystyle \phi=\phi(s) = \cosh s. \ \ \ \ \ (2)$

It is not hard to see from (2) that the level set ${\phi^{-1}(\cosh s)}$ is exactly the geodesic sphere ${S_p(s)}$ centered at ${p}$ of radius ${s}$. Furthermore, we claim that there is only one geodesic from each point to ${p}$. For, if there are two distinct geodesics ${\gamma,\beta}$ joining ${p}$ to ${q}$, then they must be of the same length ${l}$ (why?), i.e. ${\gamma,\beta:[0,l]\rightarrow M}$ both have starting point ${p}$ and endpoint ${q}$. As ${(M,g)}$ is complete, we can extend ${\gamma}$ on ${[0, l+1]}$. (2) then shows that ${\phi(\gamma(l+1))=\cosh (l+1)}$. But then since ${\beta|_{[0,l]}\cup \gamma |_{[l, l+1]}}$ is a broken geodesic joining ${p}$ and ${\gamma (l+1)}$, the distance ${d(p, \gamma(l+1))}$ must be strictly smaller than ${l+1}$, but then (2) shows that ${\phi(\gamma (l+1))<\cosh(l+1) }$, a contradiction.

The above argument shows that ${\exp_p : T_pM \rightarrow M}$ is a bijection, and clearly ${\exp_p}$ and its inverse (if exists) are smooth, thus ${\exp _p}$ is actually a diffeomorphism.

We will show that then ${g}$ is pulled back to ${T_pM\cong \mathbb{R}^n}$ under this map and using the usual spherical coordinates (i.e. normal coordinates around ${p}$), the pulled back metric is ${ds^2 +\sinh ^2 s \,d\theta^2}$, and thus ${(M,g)}$ is isometric to the hyperbolic space. For this purpose we make use of the Jacobi field.

First of all, fixed any ${v\in T_pM\cong \mathbb{R}^n}$ with length ${1}$. Let ${\gamma (s)=\exp_p(sv)}$. Fix ${s_0}$ and let ${E\in T_{\gamma(s_0)}M}$ such that ${\gamma'(s_0)\perp E}$ is orthonormal. Parallel transport ${E}$ along ${\gamma }$, which we denote by ${E(s)}$. We then claim that

$\displaystyle J(s)=\sinh s \,E(s) \ \ \ \ \ (3)$

is an Jacobi field with ${J(0)=0}$.

As

$\displaystyle \nabla _{\gamma'}\nabla _{\gamma'}J= \sinh s \,E(s)=J(s).$

By the Jacobi field equation

$\displaystyle \nabla _{\gamma'}\nabla _{\gamma'}J+ R(J, \gamma')\gamma'=0,$

we only have to prove that

$\displaystyle R(J, \gamma')\gamma'=-J. \ \ \ \ \ (4)$

To see this, note that (1) implies

$\displaystyle \nabla _X (\nabla \phi)=\phi X.$

Also (2) implies ${\gamma'(s)= \frac{\nabla \phi}{|\nabla \phi|}}$, so

$\displaystyle |\nabla \phi|^2R(J,\gamma')\gamma'= \nabla _J \nabla_{\nabla \phi}\nabla \phi-\nabla _{\nabla \phi} \nabla_J\nabla \phi- \nabla _{[J,\nabla \phi]}\nabla \phi. \ \ \ \ \ (5)$

By ${\nabla _X (\nabla \phi)=\phi X}$,

$\displaystyle \nabla _J\nabla _{\nabla \phi}(\nabla \phi)= \nabla _J(\phi\nabla \phi)= J(\phi)\nabla \phi+\nabla _J(\nabla \phi)=J(\phi)\nabla \phi+\phi^2 J=\phi^2 J \ \ \ \ \ (6)$

noting that ${J(\phi)= g(\nabla \phi, J)=0}$. We also have

$\displaystyle \nabla _{\nabla \phi}\nabla _J(\nabla \phi)= \nabla_{\nabla\phi} (\phi J)= |\nabla \phi|^2 J+ \phi \nabla _{\nabla \phi}J \ \ \ \ \ (7)$

and

$\displaystyle \nabla _{[J,\nabla \phi]}( \nabla \phi)= \phi[J, \nabla \phi]= \phi(\nabla _J \nabla \phi- \nabla _{\nabla \phi}J)= \phi^2J- \phi \nabla _{\nabla \phi}J. \ \ \ \ \ (8)$

(6)(7)(8) then gives the RHS of (5):

$\displaystyle |\nabla \phi|^2R(J,\gamma')\gamma'= -|\nabla \phi|^2J, \quad\text{i.e. }R(J,\gamma')\gamma'= -J.$

This proves (4). To finish the proof, for any vector ${X\in T_{\gamma(s_0)}M}$, ${X= X^\perp+ X^T}$, where ${X^T}$ is tangential to the sphere ${S_p(s_0)=\{s=s_0\}}$ and ${X^\perp}$ is a multiple of ${\frac{\partial}{\partial s}}$. By Gauss lemma, ${X^\perp\perp_g X^T}$ and so ${|X|_g^2= |X^\perp|^2+|X^T|^2}$. If we denote the Euclidean metric on ${T_pM}$ by ${\overline g}$ and push forward it to ${M}$ by ${\exp}$, which we still denote by ${\overline g}$, then by Gauss lemma, ${|X^\perp|_{\overline g}=|X^\perp|_{g}}$. Let ${x^i}$ be the normal coordinates around ${p}$ and so in this coordinates, ${X= \sum_i X^i \frac{\partial}{\partial x^i}}$. Take a Jacobi field ${J(s)}$ on ${\gamma(s)}$ such that ${J(s_0)=X^T}$ and ${J(0)=0}$ (it exists because there is no pair of conjugate points on $M$ (why?)). Then on one hand, it is well known that

$\displaystyle J(s)= \frac{s}{s_0} \sum_iX^i \frac{\partial}{\partial x^i} \ \ \ \ \ (9)$

On the other hand, we know from (3) that ${J(s)= \sinh s E(s)}$ where ${E(s)}$ is parallel along ${\gamma}$, this implies ${\nabla _{\gamma'(0)}J=E(0)}$, thus

$\displaystyle |X^T|^2= |J(s_0)|^2 = \sinh ^2 s_0\, |E(s)|^2= \sinh ^2 s_0 |E(0)|^2 =\sinh ^2 s_0\,|\nabla _{\gamma'(0)}J|^2.$

Using (9) and the fact that ${\nabla \frac{\partial }{\partial x^i}|_p=0}$, we know that ${\nabla_{\gamma'(0)}J= \frac{1}{s_0}\sum _i X^i \frac{\partial}{\partial x^i}|_p}$. So the above becomes

$\displaystyle |X^T|_g^2= \frac{\sinh ^2 s_0}{s_0^2}\left|\sum_i X^i \left.\frac{\partial }{\partial x^i}\right|_p\right|_g^2=\frac{\sinh ^2 s_0}{s_0^2}|X|_{\overline g}^2$

since ${g}$ agrees with ${\overline g}$ at ${p}$. We conclude that

$\displaystyle |X|_g ^2 =|X^\perp|^2 + \frac{\sinh^2 s}{s^2}|X^T|_{\overline g}^2$

for ${X\in T_{\gamma(s)}M}$.

Finally, we know that the Euclidean metric is given by

$\displaystyle \overline g=ds^2 +s^2 \;d\theta^2$

where ${d\theta ^2}$ is the standard metric on the unit sphere. So we can transform the above to

$\displaystyle g=ds^2 +\sinh ^2 s \;d\theta^2,$

which is the hyperbolic metric. $\Box$

 Remark 1 We will show by an example that the condition ${\phi> |\nabla \phi|}$ cannot be dropped. Consider the Riemannian manifold ${(M,g)=(\mathbb{R}\times N, ds^2 + e^{2s} h)}$ where ${(N, h)}$ is a complete Riemannian manifold. Let ${\phi=e^s}$. We first show that this function satisfies $\displaystyle \nabla ^2 \phi= \phi g.$ Choose local coordinates ${x^i, i=1,\cdots, n-1}$ on ${N}$. Denote ${g}$ by ${\langle\cdot,\cdot \rangle}$. Then $\displaystyle \begin{array}{rcl} \nabla ^2\phi(\partial_i,\partial_j) = \partial_i\partial_j \phi- \nabla _i \partial _j (\phi) = -\langle \nabla _i \partial _j, \partial_s\rangle \phi' &=& \langle \partial _j , \nabla_s\partial _i\rangle \phi'\\ &=& \frac{1}{2}\partial_s\langle \partial _j , \partial _i\rangle \phi'\\ &=& e^{3s} h_{ij} \\ &=& \phi g(\partial_i,\partial_j), \end{array}$ $\displaystyle \nabla ^2\phi(\partial_s,\partial_s)=\phi''=e^ s g(\partial_s,\partial_s)=\phi g(\partial_s,\partial_s),$ and $\displaystyle \begin{array}{rcl} \nabla ^2\phi(\partial_s,\partial_i) =\partial_s\partial_i \phi- \nabla _s \partial _i (\phi) &=& - \langle \nabla _s \partial _i,\partial_s\rangle \phi'\\ &=& - \frac{1}{2}\partial_i\langle \partial_s,\partial_s\rangle \phi'\\ &=&0\\ &=&\phi g(\partial _s , \partial_i). \end{array}$ We now show that ${(M,g)}$ is complete. We will denote ${e^{2s}h}$ by ${g_s}$ and the distance (resp. diameter) function corresponding to ${g_s}$ by ${d_{g_s}}$ (resp. ${diam_{g_s}}$), etc. Let ${K}$ be a closed and bounded subset in ${(M,g)}$, it suffices to show that it is compact. Let ${p_n=(x_n,y_n)}$ be a sequence in ${C}$. It is easy to see that ${\pi_1(C)}$ is bounded and thus by passing into subsequence, we can w.l.o.g assume that ${x_n \rightarrow x\in \mathbb{R}}$. Here ${\pi_1: \mathbb{R}\times N\rightarrow \mathbb{R}}$ is the projection. Also this shows that ${\pi_1(C)\subset [a,b]}$. Obviously we have ${e^{2a}h\leq g_s\leq e^{2b}h}$ for ${s\in \pi_1(C)}$. We now claim that ${y_n}$ is bounded in ${(N,h)}$. Indeed, for ${C_s= C\cap(\{s\}\times N)}$, ${e^a diam_h (C_s)=diam_{g_a}(C_s)\leq diam_{g_s}(C_s)\leq diam _{g}(C)}$. So ${d_h(y_m,y_n)\leq e^{-a} diam_g(C)}$. As ${(N,h)}$ is complete, we can pick a subsequence ${y_{n_k}\rightarrow y\in (N,h)}$ as ${k\rightarrow\infty}$. It remains to show that ${(x_{n_k}, y_{n_k})\rightarrow (x,y)}$ in ${(M,g)}$ and ${(x,y)\in C}$. But the second assertion follows from the first because ${C}$ is closed in ${(M,g)}$. To show the first claim, note that $\displaystyle \begin{array}{rcl} d_g((x_{n_k}, y_{n_k}), (x,y)) &\leq& |x_{n_k}-x|+ \max_s d_{g_s}(y_{n_k},y)\\ &\leq& |x_{n_k}-x|+ e^{2b} d_{h}(y_{n_k}, y)\\ &\rightarrow & 0 \end{array}$ as ${k\rightarrow \infty}$.
 Remark 2 As remarked by Pong Ting Kei, the condition ${\phi >|\nabla \phi|}$ can also be replaced by the condition $\displaystyle \phi\geq c>0$ for some constant ${c}$. Indeed, this condition together with ${\nabla ^2\phi=\phi g}$ implies that ${\nabla^2 \phi \geq cg}$, i.e. ${\phi}$ is strongly convex. By some convex analysis, this will show that ${\phi}$ must have a (necessarily unique) minimum, as in the proof of Theorem 1, and the proof can proceed as before. We now give some details of the convex analysis. First we can w.l.o.g. assume that ${c=1}$. Thus we have ${\nabla^2\phi\geq g}$. Fix any ${p,q\in M}$, let ${l}$ be the distance between them and ${\gamma:[0.l]\rightarrow M}$ be a unit speed geodesic joining ${p}$ to ${q}$. We let ${f(s)=\phi (\gamma(s))}$. From ${\nabla^2\phi\geq g}$, we immediately have ${f''\geq1}$ on ${[0,l]}$. By fundamental theorem of calculus, for ${b\geq a}$, $f'(b)\geq f'(a)+b-a$. So $\displaystyle \begin{array}{rcl} f(l)-f(\frac{l}{2})=\int_{\frac{l}{2}}^l f'(t)dt &\geq& \int_{\frac{l}{2}}^l( f'(\frac{l}{2})+ (t-\frac{l}{2}))dt\\ &=&f'(\frac{l}{2})\frac{l}{2}+ \int_0^\frac{l}{2}tdt\\ &=&f'(\frac{l}{2})\frac{l}{2}+ \frac{l^2}{8}. \end{array}$ Similarly $\displaystyle \begin{array}{rcl} f(\frac{l}{2})-f(0)=\int_0^{\frac{l}{2}} f'(t)dt &\leq& \int_0^{\frac{l}{2}}( f'(\frac{l}{2})- (\frac{l}{2}-t))dt\\ &=&f'(\frac{l}{2})\frac{l}{2}- \int_0^\frac{l}{2}sds\\ &=&f'(\frac{l}{2})\frac{l}{2}- \frac{l^2}{8}. \end{array}$ Subtract the second inequality from the first one, we have $\displaystyle \frac{1}{2}(f(0)+f(l))-f(\frac{l}{2})\geq\frac{l^2}{8}.$ We have showed that for any minimal geodesic ${\gamma}$ joining ${p,q}$, $\displaystyle \frac{1}{2}(\phi(p)+\phi(q))-\frac{d(p,q)^2}{8}\geq\phi(\gamma(\frac{d(p,q)}{2})).$ We now prove the each sublevel set ${\{x\in M: \phi(x)\leq c\}}$ is bounded, with the diameter depending on ${c}$. Indeed, if ${f(p), f(q)\leq c}$, join them by a geodesic ${\gamma}$ of length ${l=d(p,q)}$, then $\displaystyle 0<\phi(\gamma(\frac{l}{2}))\leq \frac{1}{2}(\phi(p)+\phi(q))-\frac{d(p,q)^2}{8}\leq c-\frac{d(p,q)^2}{8}.$ Thus ${d(p,q)<\sqrt{8c}}$. Therefore ${diam\{\phi\leq c\}\leq \sqrt{8c}}$. The sublevel sets are all closed and bounded, and so is compact, therefore the minimum of ${\{\phi\leq a\}\neq \emptyset}$ must be the minimum of ${\phi}$ in ${M}$.

Therefore we have

 Theorem 2 (Pong Ting Kei’s theorem, 2011) Let ${(M^n,g)}$ be a complete connected Riemannian manifold. Then it is isometric to the hyperbolic space if and only if there is a smooth function ${\phi}$ on ${M}$ such that $\displaystyle \nabla ^2 \phi = \phi g$ and $\displaystyle \phi\geq c>0.$

This entry was posted in Geometry. Bookmark the permalink.

### 2 Responses to A condition for a Riemannian manifold to be a hyperbolic space

1. gilbertweinstein says:

Hi, right below equation 2, should that read $\phi^{-1}(cosh(s))$ is the sphere $S_p(s)$? [Oh, yes. Corrected, thanks! -KKK]

2. 林 立偉 says:

Ultra-high-technology intelligence reincarnation resurrection.

People

When arrogant body language Morohito Buddha saying: “Thou shalt good listen. Annoyance from various industries in the past, was now living, and to Morone; complex now made from various industry today, so karma, was next to the body and Morone. I today who say music analogy to Cummings justice, such as millet, all Synergy, it was raw sprouts, then this seed is really not raw sprouts, seeds Miegu buds will grow. Son Miegu infrequently, blastomycosis it constantly. Buddha uke abuzz case. Although the complex without me, karma yet. ‘”

(Added: upper seat does not admit nor deny the soul of the soul, but another point of view because of the upper seat think karma reincarnation generate another person A new person B after death, birth and death is birth and death on this issue under constant continuous. Obviously not a general commentators say person A soulless off after death did not, it is not generally have the soul of those who think there is a separate eternal individual (soul).)

In abstract algebra, power series focus of the study is its nature as a semi-algebraic ring. Power series coefficients domains are real or complex numbers, or other domain is no longer important, convergence and divergence is also not discussed. Such pulled out of the algebraic concept is known in the form of power series. In a combination of formal power series algebra has important uses, such as generating functions and identities used in many combinations of derivation.

Part of the immaterial. Many religions believe, and the role played dominate living soul within the body of man or other substances

Quantum mechanics is a description of microscopic matter physics branch, together with the theory of relativity is considered to be the two basic pillars of modern physics,

Hourglass is a timing equipment, there are sand clock, sand pots and other nicknames. Hourglass usually consists of two linked quicksand pool combination, and the time will depend on the calculation of the inside seal fitted with some quicksand. Japanese places quite aptly called when sand dollars.

Quantum gravity, the theory of quantum description of gravity field, which belongs to square one of the theories.

In about 300 BC, the ancient Greek mathematician Euclid established contact angle and the distance between space law, is now called Euclidean geometry. Euclidean plane first developed a two-dimensional object on a deal with a “geometry” He then analyzes the “solid geometry” three-dimensional objects, all of Euclid’s axioms are arranged to Euclid.

In Euclidean space, n-dimensional sphere refers to the space in the n-1 dimensional sphere with the radius contained. It is a collection that contains all the distance to less than a certain value to a fixed point of the point.

There is a downside clepsydra, water jug ​​in the cold weather will be cold even cause icing, so that the flow rate is very slow or does not flow, but too expensive to use mercury, which use sand instead of water, and the simplification, called “Hourglass.”

Shapes are different, the most traditional and common is funnel-shaped (about half a sphere-shaped and is similar to, but slightly sharp), the central ball to open a hole, connect another quicksand pool, the other side is covered by wood. The rest are varied, such as a triangle, square, pentagon, irregularly shaped cylindrical …… Jie, and two quicksand pools may have different shapes. The most common material is glass.

[Simplified strokes]: 12 [Traditional strokes]: 12 [good and bad moral]: fierce [Pinyin]: gui [Wubi input]: jthk [stroke]: vertical, horizontal, horizontal, vertical downwards Na Na [detailed explanation of horizontal, vertical fold] : sundial ashadowcastbythesun; time; sundial

Compass, also known as compass, Luo Geng, compass disk, etc., is an important tool in feng shui feng shui is not no activity.

Starting from the first letter of the Bible, looking for a possible jump sequence from 1,2,3 letters in sequence to skip the thousands of letters, to see what can spell the word, and then starting from the first two letters, again and again. Until the last letter of the Bible.

For example Rips ExplAineD thaT eacH codE is a Case Of adDing Every fourth or twelth or fiftieth to form a word derived implicit message is READ THE CODE.

Tower of Babel (Hebrew: מגדל בבל Migdal Bavel; also translated Babel, Tower of Babel, or paraphrase of Babel), Babel has the “accident” means in Hebrew.

Marxist doctrine millennium year or millennial (English: Millennialism or chiliasm) concept from the “Millennium”, which means the length of time for a thousand cycles.

Angular momentum is the origin of the object to the associated physical displacement and momentum in physics, classical mechanics is expressed as to the origin of the cross product of the displacement and momentum, usually written as L. Angular momentum is a vector.

Shadow (also known as shadow, back), light is produced opaque dark object blocking range, and the opposite direction of the light source. Shadow is a two-dimensional cross-section profile, blocking reverse projection light objects. Movies size, shape, angle of incidence changes with the light.

The two black holes intertwined, and then they separated, a wormhole link can be made between them (Basically, a shortcut). Similarly, from string theory to view, entangled two quarks will have the same effect.

Quark (English: quark, also translated as “sub-layer” or “loss of child”) is an elementary particle, also constitute the basic unit of matter. Quarks combine with each other to form a composite particle, called hadrons, the most stable hadrons are protons and neutrons, which are units of nuclei [1].

Rigid steric shape determines its centroid position and its orientation (up to six degrees of freedom).

1 rigid body’s “position”: Pick rigid internal point G to represent the entire rigid body, usually set the object centroid or centroid for this. Observations from space reference system S, G is the point where the entire rigid body position in space. The concept can be applied to the position vector to represent: the starting point for the origin of the reference frame vector S, the end of the point G. Setting the position of the rigid body requires three coordinates, for example, by Cartesian coordinates, the coordinates of the three coordinates x-, y- coordinates, z- coordinates. This used up three degrees of freedom.

In classical mechanics, Newton’s second law describes the force and acceleration is proportional to this law, also known as “acceleration law.” Suppose a net force is applied to the object is zero, the acceleration is zero, the speed is constant, since momentum is the product of mass and velocity, so momentum is conserved. In electrodynamics, the acceleration of the movement of charged particles was emits electromagnetic radiation.

Watch the 600 km distance ten times the solar mass black hole (mimic diagram), the background for the Milky Way

General relativity is a theory of gravity with a geometric language to describe Albert Einstein published in 1916, which represents the highest level of modern physics theory gravity. Classical general relativity Newton’s law of universal gravitation is included in the framework of the special theory of relativity, and the equivalence principle and build applications on this basis. In general relativity, gravity is described as a geometric property of space-time (curvature); And this space-time curvature of space-time and energy in the material and radiation – momentum tensor directly linked, their contact information that is Einstein’s gravitational field equation (a second-order nonlinear partial differential equations group).

Spirit of the sphere, the geodesic is a great circle. Three spherical geodesic triangle composed of non-negative soul. In differential geometry, the Riemann curvature tensor, or Riemann tensor is a standard way to express curvature of Riemannian manifold manifolds and, more generally, it can be said that there is an affine connection manifold curvature, including non-twist rate or torsion. Non-“high”, “burn”, “expensive.” Part of the substance. Along the geodesic torsion, spiritual twist (or tilt) [Properties] panel [conversion] Soul

(Scientific name: Homo sapiens, Homo sapiens, meaning “wise man”)

V above, is probably not correct, there are a lot less, be enough?

A lot of information and intelligence to find books or historical geography.