1:
2: \documentstyle[epsf]{aipproc}
3:
4: \begin{document}
5: \title{Testing cosmological variability of fundamental constants}
6:
7: \author{D.~A.~Varshalovich, A.~Y.~Potekhin, and A.~V.~Ivanchik}
8: \address{Ioffe Physico-Technical Institute,
9: 194021 St.\,Petersburg, Russia}
10:
11: \maketitle
12:
13: \begin{abstract}
14: One of the topical problems of contemporary physics is a possible
15: variability of the fundamental constants.
16: Here we consider possible variability of two dimensionless constants
17: which are most important for calculation of atomic and molecular spectra
18: (in particular, the X-ray ones):
19: the fine-structure constant $\alpha=e^2/\hbar c$ and
20: the proton-to-electron mass ratio $\mu=m_p/m_e$.
21: Values of the physical constants in the early epochs are estimated
22: directly from observations of quasars --
23: the most powerful sources of radiation,
24: whose spectra were formed when the Universe was
25: several times younger than now.
26: A critical analysis of the available results
27: leads to the conclusion that present-day data do not reveal
28: any statistically significant evidence for
29: variations of the fundamental constants under study.
30: The most reliable upper limits to possible variation rates
31: at the 95\% confidence level,
32: obtained in our work, read:
33: $$
34: |\dot\alpha/\alpha| < 1.4\times 10^{-14}{\rm~yr}^{-1},
35: \quad
36: |\dot\mu/\mu| < 1.5\times10^{-14}{\rm~yr}^{-1}
37: $$
38: on the average over the last $10^{10}$ yr.
39: \end{abstract}
40:
41: \section*{Introduction}
42: Contemporary theories (SUSY GUT, superstring and others)
43: not only predict the dependence of fundamental physical
44: constants on energy\footnote{
45: The prediction of the theory that
46: the fundamental constants depend on the energy of interaction
47: has been confirmed in experiment.
48: In this paper, we consider only the space-time variability
49: of their low-energy limits.}, but also
50: have cosmological solutions in which low-energy values of
51: these constants vary with the cosmological time.
52: The predicted variation at the present epoch is small but non-zero,
53: and it depends on theoretical model.
54: In particular,
55: Damour and Polyakov \cite{Damour} have developed
56: a modern version of the string theory, whose parameters
57: could be determined from
58: cosmological variations of the coupling constants and
59: hadron-to-electron mass ratios.
60: Clearly, a discovery of these variations would be
61: a great step in our understanding of Nature.
62: Even a reliable upper bound on a possible variation rate
63: of a fundamental constant presents a valuable tool
64: for selecting viable theoretical models.
65:
66: Historically, a hypothesis that the fundamental constants
67: may depend on the {\it cosmological time\/} $t$
68: (that is the age of the Universe) was first discussed
69: by Milne \cite{Milne} and Dirac \cite{Dirac}.
70: The latter author proposed his famous ``large-number hypothesis''
71: and suggested that the gravitational constant was directly
72: proportional to $t$.
73: Later the variability of fundamental constants
74: was analyzed, using different arguments, by
75: Gamow \cite{Gamow}, Dyson \cite{Dyson}, and others.
76: The interest in the problem has been revived due to recent
77: major achievements in GUT and Superstring models (e.g.,~\cite{Damour}).
78:
79: Presently, the fundamental constants are being measured with a relative
80: error of $\sim 10^{-8}$.
81: These measurements obviously rule out considerable variations of the constants
82: on a short time scale,
83: but do not exclude their changes over the lifetime
84: of the Universe, $ \sim 1.5\times10^{10}$ years.
85: Moreover, one cannot rule out the possibility that the constants differ
86: in widely separated regions of the Universe; this could
87: be disproved only by astrophysical observations
88: and different kinds of experiments.
89:
90: Laboratory experiments cannot trace
91: possible variation of a fundamental constant
92: during the entire history of the Universe.
93: Fortunately, Nature has provided us with a tool
94: for direct measuring the physical constants in the early epochs.
95: This tool is based on observations of quasars,
96: the most powerful sources of radiation.
97: Many quasars belong to most distant objects we can observe.
98: Light from the distant quasars travels to us about $10^{10}$~years.
99: This means that the quasar
100: spectra registered now were formed $\sim 10^{10}$ years ago.
101: The wavelengths of the lines observed in these spectra
102: ($\lambda_{\rm obs}$) increase compared to their
103: laboratory values ($\lambda_{\rm lab}$) in proportion
104: $\lambda_{\rm obs}$ = $\lambda_{\rm lab} (1 + z)$, where
105: the {\it cosmological redshift\/} $z$ can be used to determine the age
106: of the Universe at the line-formation epoch.
107: In some cases, the redshift is as high as $z\sim 3{-}5$,
108: so that the intrinsically far-ultraviolet lines
109: are registered in the visible range.
110: The examples are demonstrated in Fig.~\ref{fig1}.
111: Analysing these spectra we may study
112: the epoch when the Universe was several times younger than now.
113:
114: \begin{figure}[!ht]
115: \begin{center}
116: \epsfbox[50 470 550 700]{figsakh.ps}
117: \end{center}
118: \caption{Portions of quasar spectra which show
119: the absorption lines of H$_2$ and Si{\sc\,iv} with large redshifts,
120: $z=(\lambda_{\rm obs}-\lambda_{\rm lab})/\lambda_{\rm lab}\approx 2.8$.
121: The lower horizontal axis gives the wavelengths
122: in the observer's frame ($\lambda_{\rm obs}$) and the upper
123: axis gives the wavelengths
124: in the quasar's frame $\lambda_{\rm lab}$ (in {\AA}).
125: (a)~The spectrum (thin line) of the quasar PKS 0528--250,
126: obtained with the 4-meter CTIO telescope (Chile), containing
127: H$_2$ lines which belong to the L 4--0 branch of the spectrum;
128: thick line plots the spectral fit.
129: (b)~The spectrum of the quasar HS 1946+76,
130: obtained with the 6-meter SAO telescope (Russia), containing
131: Si{\sc\,iv} doublet lines which correspond to the
132: $^2$S$_{1/2}\to ^2$P$_{3/2}$ and $^2$S$_{1/2}\to ^2$P$_{1/2}$
133: transitions.
134: \label{fig1}
135: }
136: \end{figure}
137:
138: Here we review briefly the studies of the space-time
139: variability of the fine-structure constant $\alpha$
140: and the proton-to-electron mass ratio $\mu$.
141:
142: \section*{Fine-structure constant}
143: Various tests of the fundamental constant variability differ in
144: space-time regions of the Universe which they cover.
145: {\em Local tests\/} relate to the values of constants
146: on the Earth and in the Solar system.
147: In particular, {\em laboratory tests\/}
148: infer the possible variation of certain combinations
149: of constants ``here and now'' from comparison of different
150: frequency standards. {\em Geophysical tests\/}
151: impose constraints
152: on combinations of fundamental constants over the past history
153: of the Solar system, although most of these constraints are very indirect.
154: In contrast, {\em astrophysical tests\/} allows one to ``measure''
155: the values of fundamental constants in distant areas of the early
156: Universe.
157:
158: \subsection*{Local tests}
159:
160: \subsubsection*{Laboratory experiments}
161:
162: There were a number of laboratory experimets aimed at
163: detection of trends of the fundamental constants
164: with time by comparison of frequency standards which have
165: different dependences on the constants.
166: We mention only two of the published experiments.
167:
168: Comparison of H-masers with Cs-clocks during 427 days
169: revealed a relative (H--Cs) frequency drift with a rate
170: $1.5\times10^{-16}$ per day, while the rates of (H-H) and (Cs--Cs)
171: drifts (i.e., the drifts between identical standards,
172: used to control their stability) were less than $1\times10^{-16}$ per day
173: \cite{Demidov}.
174: A similar result was found in comparison of a Hg$^+$-clock with
175: a H-maser during 140 days \cite{Prestage}:
176: the rate of the relative frequency drift was less than $(2\pm1)\times10^{-16}$
177: per day.
178:
179: Such a drift is treated as a consequence of a difference
180: in the long-term stability of different atomic clocks.
181: In principle, however, it may be caused by variation of $\alpha$.
182: That is why it gives an upper limit to the $\alpha$ variation \cite{Prestage}:
183: $|\dot\alpha/\alpha| \leq 3.7\times10^{-14}{\rm~yr}^{-1}$.
184:
185: \subsubsection*{Geophysical tests}
186:
187: The strongest bound on the possible time-variation rate of $\alpha$
188: was derived in 1976 by Shlyakhter \cite{Shlyakhter},
189: and recently, from a more detailed analysis,
190: by Damour and Dyson \cite{DD}, who obtained
191: $|\dot\alpha/\alpha| < 0.7\times 10^{-16}{\rm~yr}^{-1}$,
192: The analysis was based on measurements of isotope ratios
193: in the Oklo site in Gabon, where a unique natural uranium
194: nuclear fission reactor
195: had operated 1.8 billion years ago.
196: The isotope ratios of samarium
197: produced in this reactor by the neutron capture reaction
198: $^{149}$Sm$+n\to^{150}$Sm$+\gamma$
199: would be completely different, if the energy of the nuclear
200: resonance responsible for this capture were shifted at least by 0.1 eV.
201:
202: Another strong bound,
203: $|\dot\alpha/\alpha| < 5\times10^{-15}{\rm~yr}^{-1}$,
204: was obtained by Dyson \cite{Dyson}
205: from an isotopic analysis of natural radioactive decay products
206: in meteorites.
207:
208: A weak point of these tests is their dependence
209: on the model of the phenomenon,
210: fairly complex, involving many physical effects.
211: For instance, Damour and Dyson \cite{DD} estimated possible shift
212: of the above-mentioned resonance due to the $\alpha$ variation,
213: assuming that the Coulomb energy of the
214: excited state of $^{150}$Sm$^*$,
215: responsible for the resonance,
216: is not less than the Coulomb energy of the
217: {\em ground\/} state of $^{150}$Sm.
218: In absence of experimental data on the nuclear state in question,
219: this assumption is not justified,
220: since heavy excited
221: nuclei often have Coulomb energies smaller than those for
222: their ground states \cite{Kalvius}.
223: Furthermore,
224: a correlation between the constants of strong and electroweak
225: interactions (which is likely in the frame of modern theory)
226: might lead to further softening
227: of the mentioned bounds by 100-fold, to
228: $|\dot\alpha/\alpha| < 5\times10^{-15}{\rm~yr}^{-1}$,
229: as noted by Sisterna and Vucetich
230: \cite{Sisterna}.
231:
232: In addition, the local tests
233: cannot be extended to distant space regions and to the early
234: Universe, since the law of possible space-time variation of
235: $\alpha$ is unknown {\it a priory}.
236: It is the extragalactic astronomy that allows us to study these
237: remote regions of spacetime,
238: in particular the regions which were causally disconnected
239: at the epoch of formation of the observed absorption spectra.
240:
241: \subsection*{Astrophysical tests}
242:
243: To find out whether $\alpha$ changed
244: over the cosmological time, we have studied
245: the fine splitting of the doublet lines of Si{\sc\,iv}, C{\sc\,iv},
246: Mg{\sc\,ii} and other ions,
247: observed in the spectra of distant quasars.
248: According to quantum electrodynamics, the relative splitting
249: of these lines $\delta\lambda/\lambda$
250: is proportional to $\alpha^2$ (neglecting very small corrections).
251: Consequently, if $\alpha$
252: changed with time, then $\delta\lambda/\lambda$ would depend on
253: the cosmological redshift $z$.
254: This method of measuring $\alpha$ in distant regions
255: of the Universe had been first suggested by Savedoff \cite{Savedoff}
256: and was used later by other authors.
257: For instance, Wolfe {\it et al.\/} \cite{Wolfe} derived an estimate
258: $|\dot{\alpha}/\alpha| < 4\times 10^{-12}\ {\rm yr}^{-1}$
259: from an observation of the Mg{\sc\,ii} absorption doublet at $z = 0.524$.
260:
261: An approximate formula which relates a deviation
262: of $\alpha$ at redshift $z$ from its current value, $\Delta\alpha_z$,
263: with measured $\delta\lambda/\lambda$ in the extragalactic spectra
264: and in laboratory reads
265: \begin{equation}
266: \Delta\alpha_z \approx {c_r\over2}
267: \left[ {(\delta\lambda/\lambda)_z \over (\delta\lambda/\lambda)_0}
268: -1 \right],
269: \label{Dalpha}
270: \end{equation}
271: where $c_r\sim 1$ takes into account radiation corrections \cite{Dzuba}:
272: for instance, for Si{\sc\,iv} $c_r\approx0.9$.
273:
274: Many high-quality quasar spectra measured in the last decade
275: have enabled us to significantly increase the accuracy
276: of determination of $\delta\lambda/\lambda$ at large $z$.
277: An example of the spectra observed is shown in Fig.~\ref{fig1}.
278: For the present report, we have
279: selected
280: the results of high-resolution observations \cite{Petitjean,VPI,Outram},
281: most suitable for an analysis of $\alpha$ variation.
282: The values of $\Delta\alpha/\alpha$ calculated from these data
283: according to Eq.~(\ref{Dalpha}) are given in Table~\ref{tab1}.
284:
285: \begin{table}
286: \caption{Variation of $\alpha$ value estimated from redshifted
287: Si{\sc\,iv} fine-splitting doublets.}
288: \label{tab1}
289: \begin{tabular}{lddc}
290: {\bf Quasar} & $z$ &$\Delta\alpha/\alpha$ & {\bf Ref.} \\
291: \tableline
292: HS 1946+76 & 3.050079 & 1.58 & \cite{VPI} \\
293: HS 1946+76 & 3.049312 & 0.34 & \cite{VPI} \\
294: HS 1946+76 & 2.843357 & 0.59 & \cite{VPI} \\
295: S4 0636+76 & 2.904528 & 1.37 & \cite{VPI} \\
296: S5 0014+81 & 2.801356 & -1.80 & \cite{VPI} \\
297: S5 0014+81 & 2.800840 & -1.70 & \cite{VPI} \\
298: S5 0014+81 & 2.800030 & 1.11 & \cite{VPI} \\
299: & & & \\
300: PKS 0424$-$13 & 2.100027 & -4.51 & \cite{Petitjean} \\
301: Q 0450$-$13 & 2.230199 & -1.48 & \cite{Petitjean} \\
302: Q 0450$-$13 & 2.104986 & 0.02 & \cite{Petitjean} \\
303: Q 0450$-$13 & 2.066646 & 1.03 & \cite{Petitjean} \\
304: & & & \\
305: J 2233$-$60 & 1.867484 & -1.92 & \cite{Outram} \\
306: J 2233$-$60 & 1.869756 & -2.21 & \cite{Outram} \\
307: J 2233$-$60 & 1.871074 & -1.41 & \cite{Outram} \\
308: J 2233$-$60 & 1.925971 & 1.11 & \cite{Outram} \\
309: J 2233$-$60 & 1.941979 & 0.48 & \cite{Outram} \\
310: \end{tabular}
311: \end{table}
312:
313: As a result, we obtain a new estimate of the possible
314: deviation of the fine-structure constant at $z = 2$--4
315: from its present ($z = 0$) value:
316: \begin{equation}
317: \Delta\alpha/\alpha =
318: (-4.6\pm4.3\,[{\rm stat}]\pm1.4\,[{\rm syst}])\times10^{-5},
319: \end{equation}
320: where the statistical error is obtained from the scatter of
321: astronomical data (at large $z$) and the systematic one
322: is estimated
323: from the uncertainty of the fine splitting measurement
324: in the laboratory
325: \cite{Morton,Kelly}
326: (at $z=0$, which serves as the reference
327: point for the estimation of $\Delta\alpha$).
328: The corresponding upper limit
329: of the $\alpha$ variation rate averaged over $\sim10^{10}$~yr
330: is
331: \begin{equation}
332: |\dot{\alpha}/\alpha| < 1.4\times10^{-14}{\rm~yr}^{-1}
333: \label{alpha-bound}
334: \end{equation}
335: (at the 95\% confidence level).
336: This constraint is much more stringent than those
337: obtained from all but one previous astronomical observations.
338: The notable exception is presented
339: by Webb et al.\ \cite{Webb}, who have analysed
340: spectroscopic data of similar quality, but
341: estimated $\alpha$ from comparison of Fe{\sc\,ii}
342: and Mg{\sc\,ii} fine-splitted walelengths
343: in extragalactic spectra and in the laboratory.
344: Their result indicates
345: a tentative time-variation of $\alpha$:
346: $\Delta\alpha/\alpha=(-1.9\pm0.5)\times10^{-5}$ at $z=1.0$--1.6.
347: Note, however,
348: two important sources of a possible systematic error
349: which could mimic the effect:
350: (a) Fe{\sc,ii} and Mg{\sc\,ii} lines used
351: are situated in different orders of the echelle-spectra,
352: so relative shifts in calibration of the different orders
353: can simulate the effect of $\alpha$-variation,
354: and (b) were the relative abundances of Mg isotopes
355: changing during the cosmological evolution,
356: the Mg{\sc\,ii} lines would be subjected to an additional
357: $z$-dependent shift relative to the Fe{\sc\,ii} lines,
358: quite sufficient to simulate the variation of $\alpha$
359: (this shift can be easily estimated from recent
360: laboratory measurements \cite{Pickering}).
361: In contrast, the method based on the fine splitting
362: of a line of the same ion species (Si{\sc\,iv} in the above example)
363: is not affected by these two uncertainty sources.
364: Thus we believe that the restriction (\ref{alpha-bound})
365: is the most reliable at present for the long-term history of the Universe.
366:
367: According to our analysis,
368: some theoretical
369: models are inconsistent with observations.
370: For example, power laws $\alpha\propto t^n$
371: with $n = 1$, $-1/4$, and $-4/3$,
372: published by various authors in 1980s, are excluded.
373: Moreover, the Teller--Dyson's hypothesis on the
374: logarithmic dependence of $\alpha$ on $t$
375: \cite{Teller,Dyson} has also been shown to be inconsistent with observations.
376:
377: Many regions of formation of the spectral lines,
378: observed at large redshifts in different
379: directions in the sky, had been causally disconnected at the
380: epochs of line formation.
381: Thus, no information could have been exchanged
382: between these regions of the Universe
383: and, in principle, the fundamental constants could
384: be different there.
385: However, a separate analysis \cite{SpSciRev}
386: has shown that $\alpha$ value is the same in different directions
387: in the sky
388: within the $3\sigma$ relative error
389: $|\Delta\alpha/\alpha| < 3\times10^{-4}$.
390:
391: \section*{Proton-to-electron mass ratio}
392: %
393: The dimensionless constant $\mu = m_{\rm p}/m_{\rm e}$ approximately equals
394: the ratio of the constants of strong interaction $g^2/(\hbar c)\sim 14$
395: and electromagnetic interaction $\alpha\approx 1/137.036$, where
396: $g$ is the effective coupling constant
397: calculated from the amplitude of nucleon--$\pi$-meson scattering
398: at low energy.
399:
400: In order to check the cosmological variability of $\mu$
401: we have used high-redshift absorption lines of molecular hydrogen H$_2$
402: in the spectrum of the quasar PKS 0528--250.
403: This is the first (and, in a sense, unique)
404: high-redshift system of H$_2$ absorption lines
405: discovered in 1985 \cite{LV85}.
406: A study of these objects yields information of paramount importance
407: on the physical conditions $\sim10^{10}$ years ago.
408:
409: A possibility
410: of distinguishing between
411: the cosmological redshift of spectral wavelengths
412: and shifts due to a variation of $\mu$ arises from the
413: fact that the electronic, vibrational, and rotational energies of
414: H$_2$ each undergo a different dependence on the reduced mass of
415: the molecule. Hence
416: comparing ratios of wavelengths $\lambda_i$ of various H$_2$
417: electron-vibration-rotational lines in a quasar spectrum at some
418: redshift $z$ and in laboratory (at $z=0$), we can trace
419: variation of $\mu$.
420: The method had been used previously by Foltz {\it et al.} \cite{FCB},
421: whose analysis was corrected later in our papers \cite{VL93,SpSciRev,VP96}.
422: In the latter papers,
423: we calculated the sensitivity coefficients $K_i$
424: of the wavelengths $\lambda_i$ with respect
425: to possible variation of $\mu$ and applied a linear regression analysis to
426: the measured redshifts of individual lines $z_i$ as function of $K_i$.
427: An illustration of the wavelength dependences on the mass of
428: the nucleus is given in Table \ref{tab2},
429: where a few resonance wavelengths of hydrogen, deuterium, and
430: tritium molecules
431: are listed. One can see that, as the nuclear mass increases,
432: different wavelengths shift in different directions.
433: More complete tables, as well as two algorithms
434: of $K_i$ calculation, are given in Refs.\ \cite{SpSciRev,Lanzetta}.
435:
436: \begin{table}
437: \caption{Comparison of wave\-lengths of electron-vibro-rotational lines
438: for H$_2$, D$_2$, and $T_2$.}
439: \label{tab2}
440: \begin{tabular}{cdddc}
441: $i$ & $\lambda_i({\rm H}_2)$ & $\lambda_i({\rm D}_2)$ &
442: $\lambda_i({\rm T}_2)$ & $K_i$ \\
443: \tableline
444: L 0--0 R(1) & 1108.633 & 1103.351 & 1101.021 & $-8.18\times10^{-3}$ \\
445: L 0--2 R(1) & 1077.697 & 1081.153 & 1082.760 & $+5.35\times10^{-3}$ \\
446: L 0--9 R(1) & 992.013 & 1015.610 & 1027.218 & $+3.80\times10^{-2}$ \\
447: \end{tabular}
448: \end{table}
449:
450: Thus, if the proton mass in the epoch of line formation
451: were different from the present value, the
452: measured $z_i$ and $K_i$ values would correlate:
453: \begin{equation}
454: {z_i\over z_k} =
455: {(\lambda_i/\lambda_k)_z \over (\lambda_i/\lambda_k)_0}
456: \simeq
457: 1+(K_i-K_k)\left({\Delta\mu\over\mu}\right).
458: \label{1}
459: \end{equation}
460:
461: \begin{figure}[!ht]
462: \begin{center}
463: \epsfbox[90 260 610 580]{x99h2.ps}
464: \end{center}
465: \caption{Redshift values
466: inferred from an analysis of separate spectral features
467: in an H$_2$ absorption system in the spectrum of
468: the quasar PKS 0528$-$250,
469: plotted vs.\ $\lambda_i(\mu)$-sensitivity coefficients $K_i$.
470: The slanted solid line shows the most probable regression
471: and the dashed ones corespond to $\pm1\sigma$
472: deviations of the slope.}
473: \label{fig2}
474: \end{figure}
475:
476: We have performed a $z$-to-$K$ regression analysis using
477: a modern high-resolution spectrum
478: of PKS 0528$-$250 \cite{Lanzetta}.
479: Several tens of the H$_2$ lines have been identified;
480: a portion of the spectrum which reveales some of the lines
481: is shown in Fig.~\ref{fig1}.
482: The redshift estimates for individual absorption lines
483: with their individual errorbars are plotted in Fig.~\ref{fig2}
484: against their sensitivity coefficients.
485: The resulting parameter estimate and $1\sigma$ uncertainty is
486: \begin{equation}
487: \Delta \mu/\mu = (-11.5\pm7.6{\rm\,[stat]}\pm1.9{\rm\,[syst]})
488: \times 10^{-5}.
489: \label{dmu}
490: \end{equation}
491: The $2 \sigma$ confidence interval to $\Delta \mu / \mu$ is
492: %
493: \begin{equation}
494: |\Delta \mu / \mu| < 2.0\times 10^{-4}.
495: \label{limits}
496: \end{equation}
497: %
498: Assuming that the age of the Universe is $\sim 15$ Gyr
499: the redshift of the H$_2$ absorption system
500: $z=2.81080$ corresponds to the elapsed time of 13 Gyr
501: (in the standard cosmological model).
502: Therefore we arrive at the restriction
503: \begin{equation}
504: |\dot{\mu}/\mu|< 1.5\times 10^{-14}{\rm ~yr}^{-1}
505: \label{constraint}
506: \end{equation}
507: on the variation rate of $\mu$, averaged over 90\% of
508: the lifetime of the Universe.
509:
510: \section*{Conclusions}
511: Despite the theoretical prediction of the time-dependences
512: of fundamental constants,
513: a statistically significant variation of any of the constants
514: have not been reliably detected up to date,
515: according to our point of view substantiated above.
516: The upper limits obtained indicate that the constants
517: of electroweak and strong interactions
518: did not significantly change over the last 90\% of the history
519: of the Universe.
520: This shows that more precise measurements and observations
521: and their accurate statistical analyses are required
522: in order to detect the expected variations of the
523: fundamental constants.
524:
525: {\bf Acknowledgements.}
526: This work was performed in frames of the Project 1.25
527: of the Russian State Program ``Fundamental Metrology''
528: and supported by the grant RFBR 99-02-18232.
529:
530: \newcommand{\article}[5]{#1, {\it #3} {\bf #4}, #5 (#2).}
531:
532: \begin{references}
533: \bibitem{Damour}
534: \article{Damour, T., and Polyakov, A.M.}{1994}{Nucl.\ Phys.}{B 423}{532}
535: \bibitem{Milne}
536: \article{Milne, E.}{1937}{Proc.\ R.\ Soc.}{A 158}{324}
537: \bibitem{Dirac}
538: \article{Dirac, P.A.M.}{1937}{Nature}{139}{323}
539: \bibitem{Gamow}
540: \article{Gamow, G.}{1967}{Phys.\ Rev.\ Lett.}{19}{759}
541: \bibitem{Dyson}
542: \article{Dyson, F.J.}{Cambridge Univ., Cambridge, 1972}{{\rm in}
543: Aspects of Quantum Theory\/,
544: {\rm edited by A.~Salam and E.~P. Wigner,}}{}{p.~213}
545: \bibitem{Demidov}
546: \article{Demidov, N.A., Ezhov, E.M., Sakharov, B.A., et al.}{1992}{%
547: {\rm in} Proc.\ of 6th European Frequency and Time Forum}{}{p.~409}
548: \bibitem{Prestage}
549: \article{Prestage, J.D., Tjoelker, R.L., and Maleki, L.}{1995}{%
550: Phys.\ Rev.\ Lett.}{14}{3511}
551: \bibitem{Shlyakhter}
552: \article{Shlyakhter, A.I.}{1976}{Nature}{25}{340}
553: \bibitem{DD}
554: \article{Damour, T., and Dyson, F.J.}{1996}{Nucl.\ Phys.}{B 480}{37}
555: \bibitem{Kalvius}
556: \article{Kalvius, G.M., and Shenoy, G.K.}{1974}{%
557: Atomic and Nuclear Data Tables}{14}{639}
558: \bibitem{Sisterna}
559: \article{Sisterna, P.D., and Vucetich, H.}{1990}{%
560: Phys. Rev. D}{41}{1034}
561: \bibitem{Savedoff}
562: \article{Savedoff, M.P.}{1956}{Nature}{264}{340}
563: \bibitem{Wolfe}
564: \article{Wolfe, A.M., Brown, R.L., and Roberts, M.S.}{1976}{%
565: Phys.\ Rev.\ Lett}{37}{179}
566: \bibitem{Dzuba}
567: \article{Dzuba, V.A., Flambaum, V.V., and Webb, J.K.}{e-print:
568: physics/9808021}{%
569: Phys.\ Rev.\ A}{}{submitted}
570: \bibitem{Petitjean}
571: \article{Petitjean, P., Rauch, M., and Carswell, R.F.}{1994}{%
572: Astron.\ Astrohys.}{91}{29}
573: \bibitem{VPI}
574: \article{Varshalovich, D.A., Panchuk, V.E., and Ivanchik, A.V.}{1996}{%
575: Pis'ma Astron.\ Zh. (Engl.\ transl.: Astron.\ Lett.)}{22}{8}
576: \bibitem{Outram}
577: \article{Outram, P.J., Boyle, B.J., Carswell, R.F., Hewett, P.C.,
578: and Williams, R.E.}{1999}{%
579: Mon.\ Not.\ Roy.\ Astron.\ Soc.}{305}{685}
580: \bibitem{Morton}
581: \article{Morton, D.C.}{1991}{Astrophys.\ J.\ Suppl.}{77}{119};
582: (E) {\bf 81}, 883 (1992)
583: \bibitem{Kelly}
584: \article{Kelly, R.L.}{1987}{J.\ Phys.\ Chem.\ Ref.\ Data NBS}{16}{Suppl.\,1}
585: \bibitem{Webb}
586: \article{Webb, J.K., Flambaum, V.V., Churchill, C.W.,
587: Drinkwater, M.J., and Barrow, J.D.}{1999}{%
588: Phys.\ Rev.\ Lett.}{82}{884}
589: \bibitem{Pickering}
590: \article{Pickering, J.C., Thorne, A.P., and Webb, J.K}{1998}{%
591: Monthly Not.\ R.\ Astron.\ Soc.}{300}{131}
592: \bibitem{Teller}
593: \article{Teller, E.}{1948}{Phys.\ Rev.}{73}{801}
594: \bibitem{SpSciRev}
595: \article{Varshalovich, D.A. and Potekhin, A.Y.}{1995}{%
596: Space Sci.\ Rev.}{74}{259}
597: \bibitem{LV85}
598: \article{Levshakov, S.A., and Varshalovich, D.A.}{1985}{%
599: Monthly Not.\ R.\ Astron.\ Soc.}{212}{517}
600: \bibitem{FCB}
601: \article{Foltz, C.B., Chaffee, F.H., and Black, J.H.}{1988}
602: {Astrophys.\ J.}{324}{267}
603: \bibitem{VL93}
604: \article{Varshalovich, D.A. and Levshakov, S.A.}{1993}{JETP Lett.}
605: {58}{237}
606: \bibitem{VP96}
607: \article{Varshalovich, D.A., and Potekhin, A.Y.}{1996}{%
608: Pis'ma Astron.\ Zh. (Engl.\ transl.: Astron.\ Lett.)}{22}{3}
609: \bibitem{Lanzetta}
610: \article{Potekhin, A.Y., Ivanchik, A.V., Varshalovich, D.A.,
611: Lanzetta, K.M., et al.}{1998}{%
612: Astrophys.\ J.}{505}{523}
613: \end{references}
614:
615: \end{document}
616: \bibitem{PV94}
617: \article{Potekhin, A.Y., and Varshalovich, D.A.}{1994}{%
618: Astron.\ Astrophys.\ Suppl.\ Ser.}{104}{89}
619: \bibitem{IPV}
620: \article{Ivanchik, A.V., Potekhin, A.Y., and Varshalovich, D.A.}{1999}{%
621: Astron.\ Astrophys.}{343}{439}
622: