1: \documentclass[12pt]{article}

2: \usepackage{amsthm,graphicx,amssymb}

3: 

4: \newcounter{theorem}\theoremstyle{remark}

5: \newtheorem{remark}[theorem]{Remark}

6: 

7: \newcommand*{\rmd}{\mathrm{d}}

8: \newcommand*{\e}{\mathbf e}%

9: \newcommand*{\eqref}[1]{(\ref{#1})}

10: \newcommand*{\tfrac}[2]{\textstyle{\frac{#1}{#2}}}

11: 

12: \DeclareSymbolFont{iso}{U}{txmia}{m}{it}

13: \DeclareMathSymbol{\rea}{\mathalpha}{iso}{"92}

14: \DeclareMathSymbol{\okr}{\mathalpha}{iso}{"93}

15: \DeclareMathSymbol{\nab}{\mathalpha}{iso}{"6E}

16: \DeclareMathSymbol{\isom}{\mathalpha}{iso}{"24}

17: 

18: \date{}

19: \title{Falling into the Schwarzschild black hole. Important details.}

20: \author{S. Krasnikov\thanks{The Central Astronomical Observatory of RAS,

21: M-140, Pul\-ko\-vo, St.~Petersburg, Russia. \emph{Email}:

22: Gennady.Krasnikov@pobox.spbu.ru}}

23: \begin{document}

24: \maketitle

25: 

26: \begin{abstract}

27: The Schwarzschild space is one of the best studied spacetimes and its

28: exhaustive considerations are easily accessible. Nevertheless, for some

29: reasons it is still surrounded by a lot of misconceptions, myths, and

30: ``paradoxes". In this pedagogical paper an attempt is made to give a

31: simple (i. e., without cumbersome calculations) but rigorous

32: consideration of the relevant questions. I argue that 1) an observer

33: falling into a Schwarzschild black hole will

34: \emph{not} see ``the entire history of the universe" 2) he will

35: \emph{not} cross the

36: horizon at the speed of light 3) when inside the hole, he will \emph{not}

37: see the (future)  singularity, and 4) the latter is \emph{not}

38: ``central".

39: \end{abstract}

40: 

41: 

42: 

43: 

44: 

45: \section{Introduction}

46: 

47: The Schwarzschild spacetime (alias maximally extended Schwarzschild

48: spacetime, alias the Kruskal spacetime) is certainly one of the best

49: studied solutions of the Einstein equations. A rare  textbook in

50: relativity does not dwell on that space, which is no surprise taking into

51: account its importance and (relative) simplicity. So, one might think

52: that no mysteries are harboured there any longer, a careful reading of

53: \cite[\S\S 31,32]{MTW} being able to give the answer to almost any

54: ``silly" question. This, however, is not quite so. For a person who has

55: not yet got used to the basic concepts of general relativity (equivalence

56: of all coordinate systems, the impossibility of attaching a preferred

57: extended reference system to an observer, etc.) the Schwarzschild space

58: is fraught with pitfalls. Such a person encounters various ``paradoxes"

59: and ``miracles" (exactly as in studying special relativity or quantum

60: mechanics) and it takes some work to sort them out. Unfortunately, the

61: areal of those paradoxes and miracles is not restricted to student

62: internet forums and  popular literature. They have infiltrated the

63: semi-popular, research, and even pedagogical works. Thus one can find

64: there the assertion that, just before  crossing a black hole horizon, an

65: astronaut in a single moment of his proper time will see the whole

66: infinitely long evolution of the external universe \cite{Regge,Cher}. He

67: will see how our Sun swells becoming a red giant, how the Earth skimming

68: over the upper atmosphere of the dying Sun evaporates in its glare, and

69: how the Sun later transforms into a white dwarf \dots\cite{Cher}.

70: Elsewhere one reads that the astronaut  will traverse the horizon at the

71: speed of light

72: \cite{KisLogMes6} and after crossing the horizon he will see the

73: ``central singularity" \cite{Cher}. The authors of these excerptions are

74: all scholars of authority, so one can only pity a student reading all

75: that.

76: % ***********************************************************************

77: 

78: 

79: Thus, it seems there is a need for a paper where the most puzzling

80: properties of the Schwarzschild space would be illuminated in an as clear

81: (but rigorous) manner as possible. In the following sections I treat |

82: hopefully just in that manner | a few  most ``controversial" issues,

83: which are: Will an observer falling into the black hole see the entire

84: future of our universe? Will he cross the horizon at the speed of light?

85: Will he see the singularity? Is that singularity ``central"? (The answers

86: to all four questions are negative). The reader is supposed to be

87: familiar with only the basics of semi-Riemannian geometry. Units are used

88: in which $G=c=1$.

89: \section{The locale}

90: \subsection{The geometry of the Schwarzschild spacetime}

91: The simplest (i.~e., non-ro\-tat\-ing and uncharged) black hole is

92: described, as everybody knows, by the spacetime $\mathcal M$:

93: \begin{equation}\label{eq:metr}

94: %\begin{split}

95: \rmd s^2=

96: 4m^2\left\{-\frac{4}{xe^{x}}\rmd u\rmd v

97:  +x^2(\rmd\theta^2 + \cos^2\theta\,\rmd\phi)\right\},

98: \end{equation}

99: $$

100: u,v  \in \rea,\quad x>0,

101: $$

102:  where $m$ is a positive

103: parameter (called the mass, see below) and $x=x(u,v)$

104:  is the function defined (implicitly) by the equation

105: \begin{equation}\label{eq:def r}

106: uv=(1-x)e^x.

107: \end{equation}

108: The importance of  $\mathcal M$ | it is this spacetime that we shall call

109: Schwarzschild's | lies, of course, in the fact that it is a spherically

110: symmetric solution to the vacuum Einstein equations and, moreover, by

111: Birkhoff's theorem it is the only such solution in the class of

112: maximal\footnote{ $\mathcal M$ cannot be extended, say, to the region

113: $x(u,v)<0$ because the scalar $R_{abcd}R^{abcd}$ diverges at $x\to 0$.}

114: globally hyperbolic spacetimes. It is often convenient to choose $x$ (or

115: $r$, which is almost the same) as a new coordinate. That cannot be done

116: in the entire $\mathcal M$ (as follows, for example, from the fact that

117: $\nabla x(0,0)=0$) and we shall restrict ourselves to the region

118: \[

119: \mathcal M_*\colon\qquad u<0,\quad v>0.

120: \]

121: (in Fig.~\ref{fig:Sch}a it is shown by dark gray).

122: \begin{figure}[tb]

123: \includegraphics[width=\textwidth]{sch_en.eps}\\

124: \hspace{0.15\textwidth}(a)

125: \hfill (b)\hspace*{0.15\textwidth}

126: \caption{The sections $\phi=const$, $\theta=const$ of the

127: Schwarzschild spacetime. (a). The dark gray region $\mathcal M_*$

128:  is asymptotically flat, it is the region

129: ``outside of the Schwarzschild black hole''. By the light gray the

130: regions $\mathcal M_-$ and $\mathcal M_+$  are shown, which are,

131: respectively, expanding  and contracting ``universes"; (b). The gray

132: region is the causal past of $\gamma$, i.~e., the union of the causal

133: pasts of all its points.\label{fig:Sch} This region includes \emph{all}

134: events that have ever been observed by $\gamma$. }

135: \end{figure}

136: There the transition to the coordinates

137: \begin{equation} \label{ScKoordI}

138: r\equiv 2mx,\qquad t\equiv 2m\ln(-v/u)

139: \end{equation}

140: brings the metric~(\ref{eq:metr}) to a more customary form:

141: \begin{equation}

142: \label{eq:UsSch}

143:  \rmd s^2=-(1-\tfrac{2m}{r})\rmd t^2 + (1-\tfrac{2m}{r})^{-1}\rmd

144: r^2

145:  +r^2(\rmd\theta^2 + \cos^2\theta\,\rmd\phi)

146: \end{equation}

147: $$

148: t\in \rea,\quad r>2m.

149: $$

150: To relate it to the everyday  consider the region $\mathcal

151: M_{r_0}\subset \mathcal M_*$ defined by the inequality $r>r_0>2m$.  The

152: region is spherically symmetric and asymptotically flat, and the metric

153: there solves the source-free Einstein equations. So, $\mathcal M_{r_0}$

154: (and | again by Birkhoff's theorem | only $\mathcal M_{r_0}$) describes

155: the universe outside a ball of radius $r_0$ and mass $m$. (The equation

156: for the $r$-coordinate of a radial geodesic parametrized by the proper

157: time $\tau$ is

158: \footnote{It is easily found by varying the ``geodesic Lagrangian"

159: \cite{MTW} $L= -(1-

160: \frac{2m}{r})\dot t^2+ (1-

161: \frac{2m}{r})^{-1}\dot r^2$ with respect to $r$.}

162: \begin{equation}\label{eq:radg}

163:   \ddot r = -\frac m{r^2},

164: \end{equation}

165: where the dot stands for the derivative by $\tau$. The comparison of this

166: equation with the Newtonian one justifies our interpretation of $m$ as

167: the mass).

168: 

169: 

170: The surfaces $u=0$ and $v=0$ (alias $x=1 $, alias $r=2m$) bounding

171: $\mathcal M_*$ are called \emph{horizons}. It should be emphasized that

172: the points of horizons  have no ``magic'' properties; each of them has a

173: small neighbourhood with exactly the same (in a qualitative sense)

174: properties as a neighbourhood of any other point of any spacetime: the

175: tidal forces here are finite,  massive bodies move on timelike curves,

176: the world lines of photons are null geodesics, etc. Now, what is there

177: \emph{beyond}  the horizon? One might naively expect that since the horizon

178: is a sphere (at each moment of time; we are discussing the section of the

179: spacetime by some simultaneity surface $\mathcal S$), then what it bounds

180: is a ball. Or rather a punctured ball, with a singularity at the center.

181: That would perfectly fit the idea that the Schwarzschild solution

182: ``describes the field of a point mass (located at the center, the

183: singular point of the metric)" \cite{LL_II}. The said idea goes back to

184: classics of the pre-Kruskal epoch \cite{konc} and is amazingly widespread

185: even today. It should be stressed therefore that the just drawn picture

186: though not \emph{wrong} ($\mathcal S$ can be chosen so as to justify it)

187: is, nevertheless, grossly misleading. We shall see, in particular, that

188: the term ``central" is applicable to Schwarzschild's singularity no more

189: than, say, to Friedmann's.

190: 

191: 

192: To perceive the real  geometry of the region

193: \[

194: \mathcal M_+\colon\qquad u>0,\quad 0<r<2m,

195: \]

196: (shadowed in light gray in Fig.~\ref{fig:Sch}a) it is instructive to

197: introduce there the coordinates

198: \begin{equation} \label{ScKoordII}

199: \eta\equiv 2mx,\qquad l\equiv 2m\ln(v/u).

200: \end{equation}

201: The metric then takes the form

202: $$

203: \rmd s^2=-(\tfrac{2m}{\eta}-1)^{-1}\rmd\eta^2 +

204: (\tfrac{2m}{\eta}-1)\rmd l^2

205:  +\eta^2(\rmd\theta^2 + \cos^2\theta\,\rmd\phi)

206: $$

207: $$

208: l\in \rea,\quad \eta\in (0,2m).

209: $$

210: 

211: \begin{remark}

212: The transformation \eqref{ScKoordI} is singular at $u=0$ and $v=0$.

213: Therefore it \emph{cannot} be extended to $\mathcal M_+$. In other words,

214: $(t,r)$ and $(\eta, l)$ are

215: \emph{different} coordinates. Unfortunately, this fact is overlooked

216: sometimes, which leads to much confusion and the talk about  ``space and

217: time swapping their roles'' inside the black hole.

218: \end{remark}

219: Thus, an observer after crossing the horizon finds himself in a

220: ``universe'' with not quite usual properties\footnote{This homogeneous,

221: anisotropic universe is a special case of the Kantowski-Sachs

222:  spacetime \cite{KS}.}. The ``space'' of that universe (i.~e., the surface

223: $\mathcal S$ given in this case by the equation $\eta={\rm const} $) is a

224: homogeneous cylinder $\rea^1\times\okr^2$. It is spherically symmetric,

225: but not isotropic, the distinguished direction being that along the

226: $l$-axis. At the same time, even though the surfaces $\eta={\rm const}$,

227: $l={\rm const}$ are spheres, one should not call the $l$-coordinate

228: ``radial", because the space is invariant w.~r.~t. translations in that

229: direction. Note that the space has \emph{neither} a singularity,

230: \emph{nor} a centre.

231: 

232: With time the geometry of $\mathcal M_+$ changes. This fact is not

233: surprising --- the Schwarz\-schild space as a whole is \emph{non-static},

234: even though it has a static  [as is seen from

235: \eqref{eq:UsSch}] region\footnote{The isometries

236: $\isom^*_A\colon$

237: $

238:  t\mapsto t + A,

239: $

240: that act on $\mathcal M_*$ \emph{can} be extended to the isometries

241: $\isom_A\colon$ $u\mapsto e^{-\frac{A}{4m}}u$, $v\mapsto

242: e^{\frac{A}{4m}}v$ acting on the entire $\mathcal M$, but in  $\mathcal

243: M_+$  the orbits of the group $\isom_A$ are spacelike.}

244: %

245: %

246:  $\mathcal M_*$. The radius of the cylinders $\mathcal S$ falls and

247:  it is its

248: vanishing at $\eta=0$ that is referred to as the Schwarzschild

249: singularity\footnote{The case in point is the ``upper singularity" in

250: Fig.~\ref{fig:Sch}; the other one is, of course, in the past.}.

251: Evidently,

252:  for any observer in $\mathcal M_+$ the singularity is \emph{in the

253:  future} and, in

254:  particular, \emph{nobody}

255:  (on whichever side of the horizon) can ever observe it.

256: 

257: 

258: \begin{remark} The surfaces of simultaneity could be chosen

259: differently, of course. For example, they could be defined by the

260: equation $u+v={\rm const}$ instead of $\eta={\rm const}$. In such a case

261: $\mathcal M$ would appear as an evolving wormhole \cite{MTW}. The throat

262: of the wormhole lies in $\mathcal M_\pm$ and connects two asymptotically

263: flat isometric ``universes" --- ours and $\mathcal M_*'$. As can be seen

264: from Fig.~\ref{fig:Sch}, the two universes are causally disconnected, but

265: a traveller from one of them is allowed to see some events in the other

266: (though not before traversing the horizon).

267: \end{remark}

268: 

269: 

270: 

271: Often it is $\mathcal M_*$ that is called  Schwarzschild's space and $t$,

272: $r$ --- Schwarzschild's coordinates, while $\mathcal M$ and $u,v$ are

273: called Kruskal--Szekeres'. The   coordinates  $u,v$  cover the entire

274: manifold. And, in studying the \emph{radial} motion, when only the

275: sections $\phi=const$, $\theta=const$ matter,  their additional advantage

276: is that the metric of those sections takes the form

277: \[

278: \rmd s^2=

279: -F(r)\rmd u\rmd v,\qquad

280:  F= 16m^2x^{-1}e^{-x},

281: \]

282: which simplifies significantly the analysis of their causal structure. A

283: curve in the $(u,v)$-plane is causal (i.~e., can be the world line of a

284: particle) if and only if in all its points the angle between its tangent

285: and the vertical  (i.~e., the line $u-v=const$) is $\leq 45^\circ$. Thus

286: the set of all points from which signals can come to a point $p$ | this

287: set is called the \emph{causal past} of $p$ | is the down-directed angle

288: with the vertex in $p$ and the sides parallel to the $u$- and $v$-axes.

289: And the  \emph{causal future} of $p$, i.~e., the set of all points at

290: which $p$ can be seen, is the angle vertical to that.

291: 

292: 

293: We shall consider only the  $(u,v)$-plane taking into account that the

294: non-radial motion complicates the analysis without adding anything

295: qualitatively new. So, by a ``signal" or ``motion", etc., from now on we

296: understand a ``radially propagating signal" or ``radial motion", etc.

297: \subsection{Schwarzschild  and free-falling observers}

298: 

299: To analyze the fall  into the black hole let us consider two observers

300: separating into a point $s$, see Fig.~\ref{fig:Sch}b. One of them, let us

301: label him $\alpha$, after the parting moves with constant $r$, $\phi$ and

302: $\theta$. Such observers | we shall call them Schwarzschild | are at rest

303: in the Schwarzschild coordinates, in which the metric does not depend on

304: time. It would be quite untrue, however, to regard Schwarzschild

305: observers as ``immobile'' or, at least, inertial. Their world lines are

306: \emph{not} geodesics; so the observers experience an acceleration $a$, the fact

307: well known (empirically) to the reader as all of us are, to high

308: accuracy, Schwarzschild observers with $r=R_{\oplus} $ in the metric with

309: $m=M_{\oplus}$ and each of us moves | in the instantaneously comoving

310: system | with the acceleration $a\approx 9.8\,$m/s$^2$.

311: 

312: 

313: The second observer, $\gamma $, falls freely, i.~e., his world line is a

314: radial geodesic $\gamma(\tau)$ with $\dot r(0)\leq 0$. The most important

315: fact about $\gamma $ is that at some moment $\tau_h<\infty$ of its proper

316: time it

317: \emph{unavoidably} meets the horizon.

318: \begin{proof}

319: As follows from \eqref{eq:radg} the function $x(\tau)$ is convex. At the

320: same time

321: \[

322: x(0)>1,\qquad \dot x(0)\leq 0.

323: \]

324: Hence, there \emph{is} $\tau_h>0$ such that $x(\tau_h)=1$. So, we only

325: have to prove that $\tau$ takes

326: \emph{all} values in $[0,\tau_h]$.

327: In other words, the observer $\gamma$ \emph{must} reach the horizon, if

328: he lives long enough, and our task is to prove that he does not cease to

329: exist before his clock shows $\tau_h$. Note that this follows neither

330: from \eqref{eq:radg}, nor from any general considerations: one could

331: imagine, for example, that $\gamma$ approaches the horizon like $\mu$ in

332: Fig.~\ref{fig:Sch}a and leaves $\mathcal M_*$ as $\tau\to a\in

333: (0,\tau_h]$. To exclude such a possibility notice that as long as $\gamma

334: $ stays in $\mathcal M_*$ the coordinate $v$ on it obeys the following

335: assessment

336: \begin{equation}\label{eq:ocen}

337: \ddot v=-(\ln F),_v{\dot

338: v}^2=(1-x^{-2})v^{-1}{\dot v}^2 <v^{-1}{\dot v}^2,

339: \end{equation}

340: where the first equality is the $v$-component of the geodesic equation

341: and the second follows from the simple chain

342: \[

343: (\ln F),_v=(\ln F)'(vu),_v/(vu)'=v^{-1}\ln'F /\ln'|vu|=v^{-1}(1-x^2)/x^2,

344: \]

345: in which we have made  use of \eqref{eq:def r}. Both $v$ and $\dot v$ are

346: positive in $\mathcal M_*$, and from

347: \eqref{eq:ocen} it follows immediately that

348: \[

349: v(\tau) \leq c_1e^{c_2e^{\tau/c_2}},

350: \]

351: where $c_{1,2}$ are some constants. Consequently, until $\gamma $ leaves

352:  $\mathcal M_*$, $v(\tau)$ is bounded on any interval.

353: \end{proof}

354: Once $\gamma $ enters $\mathcal M_+$ it cannot cross the horizon back and

355: inevitably terminates at the singularity ($r\to 0$ as

356: $\tau\to\tau_0<\infty$). Note that the same is true for \emph{any} causal

357: curve | geodesic or not | just because it has to stay within the right

358: angle with the vertex in its (arbitrary) point and the sides parallel to

359: the $u$- and $v$-axes.

360: 

361: All the abovesaid looks quite elementary. However, for the reasons

362: discussed in the Introduction we should discuss in more detail two

363: aspects of $\gamma$'s history.

364: 

365: %

366: %

367: \section{The velocity at the horizon}

368: %

369: 

370: It is common knowledge that an object \emph{similar to} the black hole

371: exists in   Newtonian physics too. If a ball has a sufficiently large

372: mass and small radius, the escape velocity $V_e$ may equal the speed of

373: light. But a body falling on such a ball  | with the zero initial speed |

374: from infinity would land just with $V_e$. Perhaps, it is such reasoning

375: that gave rise to a popular belief that a body crosses the horizon at the

376: speed of light. Is it true?

377: 

378: At first glance | yes. Indeed, consider a family of observers

379:  $\mathcal N $: each member $\nab_\tau$ meets $\gamma$

380:  in the corresponding point | in $\gamma(\tau) $ | and measures

381:  $\gamma$'s velocity in his, member's, proper reference system.

382: By the proper reference system we here understand a perfectly local and

383: well-defined entity | an orthonormal tetrad in $\gamma(\tau) $

384: \[

385: \{\e_{(i)}(\tau)\},\qquad  i=0,\dots,3,

386: \]

387: with the vector $\e_{(0)}$ tangent to the world line of $\nab_\tau$

388: (thus, instead of a family of observers we could speak about a tetrad

389: field along $\gamma $). Denote now by $\mathbf v(\tau)$ the 3-velocity of

390: $\gamma $ as measured by $\nab_\tau$, i.~e.,  found in the basis

391: $\{\e_{(i)}(\tau)\}$. If $\mathcal N $ is chosen (at $\tau<\tau_h$, of

392: course) to be the set of   Schwarzschild observers, then for a radial

393: $\gamma $ it can be shown, see \cite[(102.7)]{LL_II}, that

394: \[

395: |\mathbf v|=\sqrt{1-\xi (x-1)/x},

396: \]

397: where $\xi$ is a positive constant which depends on the choice of $\gamma

398: $. Thus

399:     \begin{equation}\label{eq:lim} |\mathbf

400: v|\rightarrow 1 \quad \mathrm{as}\quad \tau\to\tau_h-0.

401: \end{equation}

402: It is this fact that is interpreted sometimes as attainment of the speed

403: of light by a  falling body and thereby as self-inconsistency of general

404: relativity, see, e.~g., \cite{KisLogMes6}.

405: 

406: The falseness in that interpretation is that $|\mathbf v|$ is assumed to

407: be continuous in $\tau$. In fact, however, the properties of $|\mathbf

408: v|$ depend heavily on the choice of $\mathcal N $ (in this sense $\mathbf

409: v(\tau)$ characterizes $\mathcal N $ rather than $\gamma $). And in the

410: case under consideration, when $\mathcal N $ can\emph{not} be

411: complemented in a continuous way by an observer  meeting $\gamma $ in $h$

412: (such an observer would have to move with the speed of light), the vector

413: \[

414: \e_{(0)}=(v\partial_v-u\partial_u)/|v\partial_v-u\partial_u|

415: =

416: (v\partial_v-u\partial_u)/\sqrt{32m^2(1-1/x)}

417: \]

418: has obviously no limit as $\tau\to\tau_h-0$. So, $\mathbf v(\tau)$ could

419: have been continuous in $\tau_h$ only by a miracle. In other words,

420: \eqref{eq:lim} \emph{does not imply} $|\mathbf v(\tau_h)|=1$.

421: 

422: Actually the vector in $h$ tangent to $\gamma $ is \emph{timelike}. This

423: has nothing to do with relativity, or even with the  metric under

424: discussion, but follows from a fundamental geometric fact: a geodesic

425: timelike in a point ($s$ in this case) is timelike in \emph{all} points.

426: Thus in any orthonormal basis (i.~e., in a proper reference system of any

427: observer located in $h$) $\gamma$  crosses the horizon moving slower than

428: light.

429: 

430: 

431: 

432: 

433: \section{What will the falling observer see?}

434: Another widely met statement is ``From the point of view [or `in the

435: reference system', or `as measured by the clock'] of a remote observer it

436: takes infinite time for a body to reach the horizon". In this section I

437: argue that contrary to the first impression it is \emph{possible} to give

438: a meaning to that statement and even in three different ways (and,

439: indeed, in the literature all three meanings can be met). The statement

440: deserves a detailed analysis, because one of the three interpretations is

441: simply

442: \emph{wrong}.

443: 

444: The problem,  in essence, is that  an observer's clock measures the

445: observer's \emph{proper} time $\tau$, i.~e., for an observer with the

446: world line $\alpha=x^i(\xi)$ it measures the quantity

447: \[\tau(\xi)=\int_0^\xi\sqrt{g_{nk}\dot{x}^n\dot{x}^k}\,\rmd\xi'

448: \]

449: (the dot here is a derivative with respect to $\xi'$) and no reasonable

450: way is seen to make the clock measure a time interval between events

451: lying

452: \emph{off} $\alpha$ (in our case those are the events  $s$ and $h$).

453: Normally this causes no problems, because we can pick \emph{any}

454: coordinate system and measure the time by using it. In doing so we need

455: not bother to interpret the thus defined time as ``true", or ``measured

456: by the clock of this or that  observer". Hence, the first way to

457: interpret the above-mentioned statement is to reformulate it: ``It takes

458: infinite Schwarzschild time for a body to reach the horizon", or, more

459: strictly (since the Schwarzschild coordinates do not cover $h$):

460: ``Between $s$ and $h$ there are events on $\gamma$ with arbitrarily large

461: $t$". The latter statement is trivially true, see Fig.~\ref{fig:Sch}.

462: 

463: \begin{remark} Replacement of the words ``in the Schwarzschild

464: coordinates" with ``from the point of view of a remote observer" is quite

465: common. The point is that light signals sent by a Schwarzschild observer

466: at   regular intervals $\Delta\tau$ of his proper time are received by

467: another Schwarzschild observer also at regular intervals\footnote{Because

468: $\mathcal M_*$  is static.} $\Delta\tau'$ (whatever are the corresponding

469: intervals $\Delta t$ are $\Delta t'$ of the coordinate time). Which

470: enables one to ``synchronize the clocks" throughout the entire $\mathcal

471: M_*$, i.~e., to introduce the time coordinate by requiring that $\Delta

472: t=\Delta t'$ (and that is how the Schwarzschild time can, indeed, be

473: defined). The similarity of this procedure to that used for building a

474: reference frame in special relativity (a purely illusive similarity, of

475: course, since $\Delta\tau\neq\Delta\tau'$ even though the observers are

476: ``at rest" w.~r.~t.\ each other) can mislead one into the idea that the

477: Schwarzschild coordinates are ``more physical" than the others and, in

478: particular, an interval of $t$ is exactly ``the time by the clock of a

479: remote observer". It is this deeply non-relativistic idea that makes |

480:  actually simple | properties of $\mathcal M$ look paradoxical. One such

481: paradox has been already considered, another is considered below, and two

482: more are presented in Fig.~\ref{fig:touch}.

483: \end{remark}

484: 

485: 

486: 

487: 

488: \begin{figure}[tb]

489: \includegraphics[width=\textwidth]{exp1.eps}\\[\medskipamount]

490: \hspace*{0.3\textwidth}(a)

491: \hfill (b)\hspace*{0.15\textwidth}

492: \caption{(a). The waist of a free falling victim must cross the

493: horizon at some moment. But his head at that moment still makes a regular

494: observer (it moves slower than light w.~r.~t.\ a Schwarzschild observer).

495: So, does it mean that in ``the head's reference system" the feet have not

496: yet reached the horizon? (b). ``From the point of view" of a

497: Schwarzschild observer with the coordinate $r_0$ his distance to the

498: horizon is $\Delta=\int_{2m}^{r_0}(1-\frac{2m}{r})^{-1/2}\rmd r$.

499: Evidently $\Delta\rightarrow 0$ as $r_0\to 2m$ and hence at some $r_0$ he

500: will be only, say, 10$\,$cm far from it. What can prevent the observer

501: from simply stretching a hand and touching the horizon?

502: \label{fig:touch}}

503: \end{figure}

504: 

505: 

506: The fact by itself that two events are separated by an infinite

507: coordinate interval is vapid: it is true for \emph{any} pair of causally

508: related events if the time coordinate is chosen appropriately. A more

509: meaningful --- since geometric --- statement can be made if we turn from

510: a relation between two events ($h$ and $s$) to a relation between an

511: event and an observer ($h$ and $\alpha $, respectively), or two observers

512: ($\alpha $ and $\gamma$). Indeed, notice that the

513: \emph{whole}

514:  $\alpha$ lies in the causal future of the \emph{segment} of $\gamma$

515:  bounded by $s$ and $h$. Physically this means that all his | infinite | life

516:  $\alpha$ will be able to receive signals sent by his falling comrade

517:  \emph{before} the latter reached the horizon\footnote{The signals though

518:  must be sent more and more frequently and with more and more blue

519:  photons.}. $\alpha$ may interpret this fact in two ways depending on

520:  which spacelike surfaces he chooses as the surfaces of simultaneity

521:  (recall that in general relativity there is \emph{no} preferred choice):

522: \begin{enumerate}

523:   \item If the surfaces are more or less horizontal in Fig.~\ref{fig:Sch}

524:   (for instance,   events are regarded simultaneous if they have the

525:   same value of $u+v$), then $\alpha$ will find nothing unusual in

526:   receiving the messages from $\gamma$. Exactly as we speak of the

527:   light of a distant star  coming to us for years after the star died,

528:   $\alpha$ could say that due to a huge | and growing | delay he keeps

529:   receiving

530:   signals from $\gamma$ centuries after the latter \emph{actually}

531:   traversed the horizon.

532:   \item One can choose, however, the simultaneity surfaces to be more and

533:   more tilted (cf.~the surfaces $t=const$ in Fig.~\ref{fig:Sch}a). This |

534:   rather exotic | choice would mean that $\alpha$ considers the

535:   information received with every signal as more and more fresh. And he

536:   would be quite consistent claiming that the fall  is infinitely long.

537: \end{enumerate}

538: 

539: There is, however, another | opposite, in a sense | approach to what

540: should be called infinitely long by a remote observer's clock. Imagine a

541: point $p$ which contains  the \emph{entire} $\alpha $ in its causal past.

542: An observer $\omega $, if his world line passes through $p$, would see

543: the entire (infinite) history of $\alpha $: he will see $\alpha $ aging,

544: his sun swelling and reddening, its protons decaying, etc. In his turn,

545: $\alpha $ would be able at any moment to send a message to $\omega $

546: (though maybe not to receive a response). All in all it would be quite

547: legitimate to say that $\omega $ reaches $p$ in infinite time by

548: $\alpha$'s clock.

549: 

550: In this just formulated sense the statement that $\gamma$'s falling time

551: is infinite is \emph{wrong}. Indeed, as is seen from Fig.~\ref{fig:Sch}b,

552: $\alpha $ | like any other Schwarzschild observer | leaves the causal

553: past of $h$ and of the entire $\gamma$, too. So, the falling observer

554: will

555: \emph{not} see the entire future of the Universe. Moreover, he will see

556: nothing \emph{at all} beyond the shadowed region in Fig.~\ref{fig:Sch}b.

557: In particular, the last event in $\alpha$'s life observed  by $\gamma$

558: before the latter submerges into $\mathcal M_*$, is $d$.

559: 

560: 

561: \begin{remark}

562: In  Reissner-Nordstr\"{o}m and Kerr black holes under their event

563: horizons (which are quite similar to Schwarzschild's) there is another

564: remarkable surface --- the Cauchy horizon. And that horizon

565: \emph{does} have the property in discussion: an astronaut falling into the

566: black hole reaches the Cauchy horizon in a finite proper time and crosses

567: it in a point $p$ that contains in its causal past the whole ``external

568: universe". Such an astronaut, indeed, \emph{will be able} to see the

569: death of stars and galaxies, see, e.~g.,~\cite{Chandrasekhar}.

570: \end{remark}

571: 

572: \section*{Acknowledgements}

573: The author was supported in part by RNP Grant No.~2.1.1.6826.

574: \begin{thebibliography}{A}

575: \bibitem{MTW}C. W. Misner, K. S. Thorne, and J. A. Wheeler,

576: \emph{Gravitation} (Freeman, San Francisco, 1973).

577: \bibitem{Regge}T. Regge, \emph{Cronache dell'Universo}

578: (Boringhieri, Torino, 1981).

579: \bibitem{Cher}A. Cherepaschuk, in  \emph{Astronomiya: Vek XXI} ed.~by V. Surdin,

580: (Fryazino, 2007).

581: \bibitem{KisLogMes6}A. A. Logunov, M. A. Mestverishvili, and V. V.

582: Kiselev,  Phys.\ Part.\ Nucl. \textbf{37}, 317 (2006).

583: \bibitem{LL_II}L. D. Landau and E. M. Lifshitz, \emph{The Classical Theory

584: of Fields} (But\-terworth-Heinemann, 1980).

585: \bibitem{konc}A. S. Eddington, \emph{The mathematical Theory of Relativity}

586: (2-nd Edition. Cambridge, University Press, 1924); \\ V. Fock,

587: \emph{The Theory of Space, Time, and Gravitation} (Pergamon, New York, 1959).

588: \bibitem{KS}R. Kantowski and R. Sachs, J. Math. Phys. \textbf{7},

589: 443 (1966).

590: \bibitem{Chandrasekhar}S. Chandrasekhar,   \emph{The Mathematical Theory

591: of Black Holes } (Oxford University Press, New York, 1983); \\ B. Carter,

592: in

593: \emph{Black Holes - Les Houches 1972}, ed. by C. DeWitt and B.~S.~DeWitt,

594: (Gordon and Breach Science Publishers, New York, 1973).

595: \end{thebibliography}

596: 

597: 

598: \end{document}

599: