1: \documentclass{llncs} 

2: \usepackage[colorlinks=true,

3: linkcolor=webgreen,

4: filecolor=webbrown,

5: citecolor=webgreen]{hyperref}

6: 

7: \definecolor{webgreen}{rgb}{0,.5,0}

8: \definecolor{webbrown}{rgb}{.6,0,0}

9: 

10: 

11: \usepackage{psfig,amssymb}

12: \pagestyle{plain}

13: 

14: \def\binom#1#2{{#1}\choose{#2}}

15: \newcommand{\eqn}[1]{(\ref{#1})}

16: \newcommand{\bsq}{\vrule height .9ex width .8ex depth -.1ex}

17: 

18: \begin{document}

19: \title{My Favorite Integer Sequences}

20: \author{N. J. A. Sloane}

21: \institute{Information Sciences Research \\

22:            AT\&T Shannon Lab \\

23:            Florham Park, NJ 07932-0971 USA \\

24:            Email: \href{mailto:njas@research.att.com}{njas@research.att.com}

25: }

26: \date{}

27: \maketitle 

28: 

29: \begin{abstract}

30: This paper gives a brief description of the author's database of integer sequences,

31: now over 35 years old, together with a selection of a few of the most interesting

32: sequences in the table.

33: Many unsolved problems are mentioned.

34: 

35: \vspace{.5\baselineskip}

36: This paper was published (in a somewhat different form) in

37: {\em Sequences and their Applications (Proceedings of SETA '98)}, C. Ding, T. Helleseth and H. Niederreiter (editors), Springer-Verlag, London, 1999, pp. 103-130. 

38: Enhanced pdf version prepared Apr 28, 2000.

39: 

40: \vspace{.5\baselineskip}

41: The paragraph on ``Sorting by prefix reversal'' on page \pageref{sorting}

42: was revised Jan. 17, 2001.

43: \end{abstract}

44: 

45: \section{How it all began}

46: I started collecting integer sequences in December 1963, when I was a graduate student at

47: Cornell University, working on perceptrons (or what are now called neural networks).

48: Many graph-theoretic questions had arisen, one of the simplest

49: of which was the following.

50: 

51: Choose one of the $n^{n-1}$ rooted labeled trees with $n$ nodes at random, and pick a random node:

52: what is its expected height above the root?

53: To get an integer sequence, let $a_n$ be the sum of the heights of all nodes in all

54: trees, and let $W_n = a_n/n$.

55: The first few values $W_1, W_2 , \ldots$ are

56: \begin{center}

57: 0, 1, 8, 78, 944, 13800, 237432, $\ldots$,

58: \end{center}

59: a \htmladdnormallink{sequence}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=000435} engraved on my memory.

60: I was able to calculate about ten terms, but I needed to know how

61: $W_n$ grew in comparison with $n^n$, and it was impossible to guess this from 

62: so few terms.

63: So instead I tried to guess a formula for the $n$th term.

64: Again I was unsuccessful.

65: Nor could I find this sequence in Riordan's book \cite{R1}, although

66: there were many sequences that somewhat resembled it.

67: 

68: So I started collecting all the sequences I could find, entering them on punched cards,

69: thinking that if any of these sequences came up in another problem,

70: at least I would know what {\em they} were.

71: 

72: I never did find that sequence in the literature, but I learned P\'{o}lya's theory of counting

73: and (with John Riordan's help) obtained the answer, which appears in \cite{RS1}.

74: There {\em is} a simple formula, although maybe not simple enough to be guessed:

75: $$W_n = (n-1)! \sum_{k=0}^{n-2} \frac{n^k}{k!} ~.$$

76: An old formula of Ramanujan \cite{Wat29} then implies that

77: $$\frac{W_n}{n^n} \sim \sqrt{\frac{2\pi}{n}} , \quad\mbox{as}\quad n \to \infty~,

78: $$

79: which was what I needed.

80: That sequence became number \htmladdnormallink{A435}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=000435} in the collection.

81: 

82: The idea of a ``dictionary'' of integer sequences was received with enthusiasm by many

83: people, and in 1973 I published \cite{HIS}, containing about 2400 sequences,

84: arranged lexicographically.

85: One correspondent commented on the book by saying

86: ``There's the Old Testament, the New Testament and the Handbook of Integer Sequences''.

87: 

88: Over the next twenty years an enormous amount of new material arrived,

89: preprints, reprints, postcards, typewritten letters, handwritten letters, etc., and it was

90: not until 1995 that --- with Simon Plouffe's help ---

91: a sequel \cite{EIS} appeared.

92: This contained 5500 sequences.

93: 

94: Around the same time I set up two services that can be used to consult the database

95: via electronic mail.

96: The first has the address

97: \href{mailto:sequences@research.att.com}{sequences@research.att.com},

98: and simply looks up a sequence in the table.

99: The second

100: email address, which is

101: \href{mailto:superseeker@research.att.com}{superseeker@research.att.com},

102: tries hard to find an explanation even for a sequence not in the table.

103: 

104: A large number of new sequences started arriving as soon as \cite{EIS}

105: appeared, and when in 1996 the total number reached 16,000, three times the number in the book,

106: I decided to set up a web site for the database \cite{OEIS}.

107: New sequences still continue to pour in, at about 10000 per year.

108: At present, in April 1999, the database contains about 50000 sequences, and the web site

109: receives over 2500 hits per day.

110: 

111: The main reason for this rapid expansion is that in the two books I only included sequences

112: that had been or were about to be published.

113: For the on-line version, where storage space is no longer a limitation, any well-defined

114: and sufficiently interesting sequence is eligible for inclusion.

115: 

116: There is now also an electronic {\em Journal of Integer Sequences}

117: \cite{JIS} and a mailing list for sequence fans \cite{Fan}.

118: 

119: The following sections describe how the database is used (Section 2) and the kinds of sequences

120: it contains (Section 3).

121: Section 4 discusses a few ``hard'' sequences and Section 5 some recursive examples.

122: Then Sections 6--8 describe sequences associated with meandering rivers

123: and stamp-folding, extremal codes and lattices, and Levine's sequence.

124: 

125: Besides integer sequences, the table also contains many examples of arrays of numbers,

126: Pascal's triangle, for example --- see Section 9.

127: The final three sections discuss sequences associated with the Wythoff array, the boustrophedon transformation of sequences,

128: and Tchoukaillon solitaire.

129: 

130: \section{How the database is used}

131: The main applications of the database are in identifying sequences or in finding out the

132: current status of a known sequence.

133: 

134: The database has been called a mathematical analogue of a ``fingerprint file''

135: \cite{Cipra}.

136: You encounter a number sequence, and wish to know if anyone has ever come across it before.

137: If your sequence is in the database, the reply will provide a description, the first 50

138: or so terms (usually enough to fill three lines on the screen), and,

139: when available, formulae, recurrences,

140: generating functions, references,

141: computer code for producing the sequence, links to relevant web sites, etc.

142: 

143: Let me illustrate how the database is used with a typical story.

144: Last summer the following question arose at AT\&T Labs in connection with a quantization

145: problem \cite{SVS}.

146: Given an $n$-dimensional lattice $\Lambda$, for which integers $N$ does $\Lambda$ have a sublattice of index $N$ that is geometrically similar to $N$?

147: 

148: For the two-dimensional root lattice $A_2$, for example, it is easy to see that a necessary

149: and sufficient condition is that $N$ be of the form $a^2 + ab + b^2$.

150: 

151: However, for the four-dimensional lattice $A_4$ the situation is more complicated.

152: The first thing we did was to run a computer search, which showed that $A_4$ has a similar

153: sublattice of index $N= c^2$ if and only if $c$ is one of the numbers

154: $$1,4,5,9,11,16,19,20,25,29,31,36, \ldots$$

155: This turned out to be sequence \htmladdnormallink{A31363}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=031363} in the database, with a reference to Baake \cite{Baa1},

156: where it appears

157: in an apparently different context

158: as the indices of coincidence site sublattices in a certain

159: three-dimensional quasicrystal.

160: \cite{Baa1} identifies these numbers as those positive integers in which all primes congruent to 2 or 3 (modulo 5) appear to an even power.

161: This was a very useful hint in getting started on our problem.

162: We were able to show that this is the correct condition for the $A_4$ lattice, and to find analogous results for a number of

163: other lattices \cite{CRS}.

164: (However, we have not yet found a direct connection between $A_4$ and the quasicrystal problem.

165: Nevertheless, the occurrence of the same numbers in the two problems cannot be entirely

166: coincidental.)

167: 

168: My files contain many similar stories.

169: Some other examples can be found in Chapter 3 of \cite{EIS}.

170: 

171: Most of the applications however are less dramatic.

172: One encounters a sequence in the middle of a calculation,

173: \htmladdnormallink{perhaps}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=001405}

174: \begin{center}

175: 1,1,2,3,6,10,20,35,70,126, 252, $\ldots$,

176: \end{center}

177: and one wants to know quickly what it is --- preferably a formula (in this case it is ${\binom{n}{[n/2]}}$)

178: or generating function.

179: 

180: It is worth emphasizing a special case of this: the

181: simplification of binomial coefficient sums.

182: Powerful methods are available for simplifying such sums by

183: computer \cite{Nem}, \cite{AB}.

184: But if one is in a hurry, one can first try evaluating the initial terms and looking up the sequence

185: in the table.

186: E.g. the sum

187: $$a(n) = \sum_{k=0}^n {\binom{2n-2k}{n-k}}^2 {\binom{2k}{k}}$$

188: produces

189: $$1, 8,88,1088, 14296, 195008 , \ldots$$

190: (\htmladdnormallink{A36917}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=036917}), and the entry in the table supplies a recurrence

191: $$n^3 a(n) = 16 \left( n - \frac{1}{2} \right)

192: (2n^2 -2n+1) a(n-1) - 256 (n-1)^3 a(n-2)$$

193: and a reference (\cite{AB}, p. ix, although the sum is misstated in the first printing).

194: I have begun entering

195: into the database

196: all the sequences corresponding to the singly-indexed identities in Gould's table \cite{Gould}

197: of binomial coefficient identities.

198: 

199: A related application is in identifying arithmetic inequalities.

200: Suppose you suspect that

201: \begin{equation}\label{EqIa}

202: \sigma (n) \ge d(n) + \phi (n) , \quad\mbox{for}\quad

203: n \ge 2 ~,

204: \end{equation}

205: where $\sigma, d$ and $\phi$ are respectively the sum of divisors, number of divisors, and Euler totient functions.

206: You could evaluate the sequence formed by the difference of the two sides, which for $n \ge 1$ is

207: $$-1, 0,0,2,0,6,0,7,4,10,0,18,0,14,12, \ldots$$

208: (\htmladdnormallink{A46520}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=046520}).

209: The table then points you to a reference (\cite{HNT}, \S I.3.1) where this is stated as a theorem.

210: (Again I would like to get more examples of such sequences.)

211: 

212: Another important use for the database is in finding out the present state of knowledge about

213: some problem.

214: A second story will illustrate this.

215: The number of Latin squares of order $n$ $(Q_n )$ is given by sequence \htmladdnormallink{A315}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=000315}

216: (see Fig.~\ref{FLS1}).

217: Computing $Q_n$ is one of the famous hard problems in combinatorics.

218: \begin{figure}[htb]

219: $$

220: \begin{array}{r@{~~~~}r}

221: n & \multicolumn{1}{c}{Q_n} \\ \hline

222: 1 & 1 \\

223: 2 & 1 \\

224: 3 & 1 \\

225: 4 & 4 \\

226: 5 & 56 \\

227: 6 & 9408 \\

228: 7 & 16942080 \\

229: 8 & 535281401856 \\

230: 9 & 377597570964258816 \\

231: 10 & 7580721483160132811489280

232: \end{array}

233: $$

234: 

235: \caption{Sequence \protect\htmladdnormallink{A315}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=000315}, the number of Latin squares of order $n$ (McKay and Rogoyski \cite{MR95}).}

236: \label{FLS1}

237: \end{figure}

238: 

239: In 1991 Brendan McKay, at the Australian National University in Camberra,

240: computed $Q_{10}$ (which as can be seen is a rather large number).

241: When he checked the database he found that the same value had recently been obtained by

242: Eric Rogoyski of Cadence Design Systems in San Jose, California.

243: As it turned out the two methods were similar but not identical, and he and Rogoyski ended up writing a joint paper

244: \cite{MR95}, acknowledging the database for bringing them together.

245: 

246: 

247: \section{Types of Sequences}

248: The database contains sequences from all branches of science, including

249: \begin{itemize}

250: \item

251: enumeration problems (combinatorics, graph theory, lattices, etc.)

252: \item

253: number theory (number of solutions to $x^2 + y^2 + z^2 =n$, etc.)

254: \item

255: game theory (winning positions, etc.)

256: \item

257: physics (paths on lattices, etc.)

258: \item

259: chemistry (sizes of clusters of atoms, etc.)

260: \item

261: computer science (number of steps to sort $n$ things, etc.)

262: \item

263: communications ($m$-sequences, weight distributions of codes, etc.)

264: \item

265: puzzles

266: \item

267: etc.

268: \end{itemize}

269: 

270: To be accepted into the database, a sequence must be well-defined and interesting.

271: However, when in doubt, my tendency is to accept rather than to reject.

272: The amazing coincidences of the Monstrous Moonshine investigations \cite{MM}

273: make it difficult to say that a particular sequence, no matter how obscure,

274: will never be of interest.

275: 

276: Readers are urged to send me any sequences they come across that are not at present in the

277: database.

278: There is a convenient electronic form for this purpose in \cite{OEIS}.

279: Some of the reasons for sending in your sequence are as follows:

280: \begin{itemize}

281: \item

282: this stakes your claim to it

283: \item

284: your name is immortalized

285: \item

286: the next person who comes across it will be grateful and, not least,

287: \item

288: you may benefit from this yourself, when you come across the same sequence some weeks from now.

289: \end{itemize}

290: 

291: Often one finds that a particular project may involve dozens of sequences, all variants of a few

292: basic ones.

293: Ideally you should send them all to the database!

294: 

295: 

296: \section{Hard sequences}

297: One of the keywords used in the database is ``hard'',

298: which indicates that the term following those given is not known.

299: Besides the Latin squares problem mentioned above, some other classic hard sequences are the following: \\

300: 

301: \noindent{\bf Projective planes.}

302: The number of projective planes of orders $n=2,3, \ldots, 10$ (\htmladdnormallink{A1231}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=001231}):

303: $$1,1,1,1,0,1,1,4,0 ,$$

304: where the last term refers to the result of Clement Lam et~al. \cite{LamTS89}

305: (completing work begun in \cite{MacST73})

306: that there is no projective plane of order 10. \\

307: 

308: \noindent{\bf The Poincar\'{e} conjecture.}

309: The number of differential structures on the $n$-sphere, for

310: $n=1,2, \ldots, 16$ (\htmladdnormallink{A1676}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=001676}):

311: $$1,1,1?, 1,1,1, 28, 2, 8, 6, 992, 1,3,2, 16256,2,$$

312: as given by Kervaire and Milnor \cite{KM63}.

313: The {\em Poincar\'{e} conjecture} is that the third term is 1. \\

314: 

315: \noindent{\bf Dedekind's problem.}

316: The number of monotone Boolean functions of $n$ variables, for $n=1,2,3, \ldots, 8$

317: (\htmladdnormallink{A7153}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=007153}):

318: $$

319: \begin{array}{l}

320: 1,4,18,166, 7579, 7828352, 2414682040996, \\

321: 56130437228687557907786$$

322: \end{array}

323: $$

324: where the last term was computed by Wiedemann \cite{Wie}.

325: (As is the case for many of these examples, there are several other versions of this sequence in the database.) \\

326: 

327: \noindent{\bf The Hadamard maximal determinant problem.}

328: What is the maximal determinant of an $n \times n$ $\{0,1\}$-matrix?

329: The values for $n=1,2, \ldots, 13$ are (\htmladdnormallink{A3432}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=003432}):

330: $$1,1,2,3,5,9, 32, 56, 144, 320, 1458, 3645, 9477,$$

331: where the last two terms are due to Ehlich, and Ehlich and Zeller \cite{Ehl1},

332: \cite{EhZ}.

333: For $n \equiv -1~(\bmod~4)$, Hadamard showed that the $n$th term is equal to 

334: $$(n+1)^{(n+1)/2} /2^n~,$$

335: provided that what is now called a Hadamard matrix of order $n$

336: exists.

337: In some cases conference matrices give the answer when

338: $n \equiv 1~(\bmod~4)$, but

339: the problem of finding the other terms in the sequence seems 

340: to have been untouched for over 35 years.

341: It would be nice to have confirmation of the above values as well as

342: some more terms! \\

343: 

344: \noindent{\bf Enumerating Hadamard matrices.}

345: The number of Hadamard matrices

346: of orders $n=4,8,12, \ldots, 28$ (\htmladdnormallink{A7299}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=007299}) is

347: $$1,1,1,5,3,60, 487,$$

348: where the last entry is the work of Kimura \cite{Kimur86},

349: \cite{Kimur94a},

350: \cite{Kimur94b},

351: \cite{KimOh86}.

352: The {\em Hadamard conjecture} is that such a matrix always exists if $n$ is

353: a multiple of 4.

354: Judging by how rapidly these numbers are growing, this should not be hard to prove,

355: yet it has remained an open question for a century.

356: Of course, as the example in Section 7 shows, such numerical evidence can be

357: misleading. \\

358: 

359: \noindent{\bf The kissing number problem.}

360: How many spheres can touch another sphere of the same size?

361: For arrangements that occur as part of a lattice packing,

362: the answers are known for $n=1,2, \ldots, 9$ (\htmladdnormallink{A1116}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=001116}):

363: $$2,6,12,24,40, 72, 126, 240 , 272,$$

364: the last term being due to Watson (see \cite{SPLAG}).

365: For nonlattice packings, all we know is

366: $$2,6,12,?,?,?,?,240, \ge 306~.$$

367: The best bounds known in dimensions 4, 5, 6 and 7 are respectively

368: $$\mbox{24--25, 40--46, 72--82} \quad\mbox{and}\quad

369: \mbox{126--140}$$

370: --- see \cite{SPLAG} for further information. \\

371: 

372: \paragraph{\bf Sorting by prefix reversal.}\label{sorting}

373: If you can only reverse segments that include the initial term of the current permutation, how many reversals are needed to

374: transform an arbitrary permutation of $n$ letters to the identity permutation?

375: To state this another way \cite{AMM75}:

376: \begin{quote}

377: The chef in our place is sloppy, and when he prepares a stack of pancakes they come out all different sizes.

378: Therefore, when I deliver them to a customer, on the way to the table I

379: rearrange them (so that the smallest winds up on top, and so on, down to the largest at the bottom)

380: by grabbing several from the top and flipping them over,

381: repeating this (varying the number I flip) as many times as necessary.

382: If there are $n$ pancakes, what is the maximum number of flips (as a function

383: $f(n)$ of $n$) that I will ever have to use to rearrange them?

384: \end{quote}

385: The only exact values known are $f(1),\ldots, f(9)$:

386: $$0,1,3,4,5,7,8,9,10,$$

387: due to Garey, Johnson and Lin, and Robbins (see \cite{AMM75}, \cite{GaP79}).

388: It is also known that $f(n) \ge n+1$ for $n \ge 6$, $f(n) 

389: \ge 17n/16$ if $n$ is a multiple of 16

390: (so $f(32) \ge 34$), and $f(n) \le (5n+5)/3$, the last two bounds\footnote{(Added May 5, 2000.)

391: The bound $f(n) \le (5n+5) /3$ was independently obtained by E. Gy\"{o}ri and G. Tur\'{a}n, Stack of pancakes,

392: {\em Studia Sci. Math. Hungar.}, {\bf 13} (1978), 133--137.

393: I thank L\'{a}szl\'{o} Lov\'{a}sz for pointing out this reference.}

394: being due to

395: Gates and Papadimitriou \cite{GaP79}.

396: Again it would be nice to have more terms.

397: 

398: Note added Jan. 17, 2001.

399: John J. Chew III (Department of Mathematics, University of Toronto)

400: has found that $f(10)$ through $f(13)$ are $11, 13, 14, 15$, respectively,

401: so the sequence begins:

402: $$0,1,3,4,5,7,8,9,10,11,13,14,15$$

403: This has now been added to the database as sequence

404: \htmladdnormallink{A058986}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=058986}.

405: 

406: \section{Recursive sequences}

407: Whereas the sequences in the previous section enumerated some class of objects, the following are self-generated. \\

408: 

409: \noindent{\bf Differences $=$ complement} (\htmladdnormallink{A5228}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=005228}):

410: $$1,3,7,12,18,26,35,45,56,69,83,98,114, \ldots$$

411: The differences (\htmladdnormallink{A30124}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=030124})

412: $$2,4,5,6,8,9,10,11,13,14,15,16,17,19, \ldots$$

413: are the terms not in the sequence!

414: This is one of many fine self-generating sequences from Hofstadter \cite{GEB}. \\

415: 

416: \noindent{\bf Golomb's sequence.}

417: The $n$th term is the number of times $n$ appears (\htmladdnormallink{A1462}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=001462}):

418: $$1,2,2,3,3,4,4,4,5,5,5,6,6,6,6,7,7,7,7,8, \ldots$$

419: The $n$th term is the nearest integer to (and converges to)

420: $$\tau^{2- \tau} n^{\tau -1} ~,$$

421: where $\tau = (1+ \sqrt{5})/2$ \cite{Gol66},

422: \cite[Section E25]{UPNT}. \\

423: 

424: \noindent{\bf Wilson's primeth recurrence.}

425: $a_{n+1}$ is the $a_n$-th prime (\htmladdnormallink{A7097}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=007097}), shown in Fig.~\ref{FRP1}.

426: The sequence was sent in by R. G. Wilson V \cite{RGW}, and the last few terms were computed

427: by P. Zimmermann \cite{Zim1} and M. Del\'{e}glise \cite{Del1}.

428: Their algorithm is a slightly speeded up version of an algorithm for computing

429: $\pi (x)$, the number of primes not exceeding $x$, due to

430: J. C. Lagarias, V. S. Miller and A. M. Odlyzko (see \cite{LO82}).

431: It is quite remarkable that it is possible to compute so many terms of this

432: sequence.

433: 

434: \begin{figure}[htb]

435: $$

436: \begin{array}{r}

437: 1 \\

438: 2 \\

439: 3 \\

440: 5 \\

441: 11 \\

442: 31 \\

443: 127 \\

444: 709 \\

445: 5381 \\

446: 52711 \\

447: 648391 \\

448: 9737333 \\

449: 174440041 \\

450: 3657500101 \\

451: 88362852307 \\

452: 2428095424619 \\

453: 75063692618249 \\

454: 2586559730396077 \\

455: \end{array}

456: $$

457: 

458: \caption{$a_{n+1}$ is the $a_n$-th prime.}

459: \label{FRP1}

460: \end{figure}

461: 

462: \noindent{\bf Recam\'{a}n's sequences.}

463: (i) $a_n = a_{n-1} -n$ if $a_{n-1} -n > 0$ and $a_{n-1} -n$

464: has not already occurred in the sequence, otherwise $a_n = a_{n-1} +n$ (\htmladdnormallink{A5132}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=005132}):

465: $$1,3,6,2,7,13,20,12,21,11,22,10,23,9, \ldots$$

466: (ii) $a_{n+1} = a_n/n$ if $n$ divides $a_n$, otherwise $a_{n+1} = na_n$ (\htmladdnormallink{A8336}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=008336}):

467: $$1,1,2,6,24,120,20,140, 1120, 10080, \ldots$$

468: These were sent in by B. Recam\'{a}n \cite{Rec1}.

469: How fast do they grow? \\

470: 

471: \noindent{\bf The \$10,000 sequence.}

472: In a colloquium talk at AT\&T Bell Labs \cite{Bell}, John Conway

473: discussed the sequence (\htmladdnormallink{A4001}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=004001})

474: $$1,1,2,2,3,4,4,4,5,6,7,7,8,8,8,8,9, \ldots$$

475: defined by (for $n \ge 3$)

476: $$a(n+1) = a(a(n)) + a(n+1 - a(n)) ~.$$

477: (In words, $a(n+1)$ is the $a(n)$th term in

478: from the left plus the $a(n)$th term in from the right.)

479: This sequence seems to have been introduced by either David Newman or Douglas Hofstadter around 1986.

480: In his talk,

481: Conway said that he could prove that $\frac{a(n)}{n} \to \frac{1}{2}$,

482: and offered \$10,000 for finding the exact $n$ at which

483: $\left| \frac{a(n)}{n} - \frac{1}{2} \right|$ last exceeds $\frac{1}{20}$.

484: My colleague Colin Mallows did not take long to analyze the sequence, and came up with an answer of

485: 6083008742 \cite{CCS1}.

486: 

487: Colin tells me that in fact the problem is actually much easier than either he or John Conway had believed, and the true answer is 1489.

488: A recent paper by Kubo and Vakil \cite{KV1} also studies this sequence and its generalizations.

489: See also \cite[Section E31]{UPNT}.

490: 

491: Many variants of this sequence have not yet been analyzed.

492: Even the rate of growth of this one is not known:

493: $a(1) = a(2) =1$, $a(n) = a(a(n-2)) + a(n-a(n-2))$, \htmladdnormallink{A5229}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=005229} \cite{CCS1},

494: which begins

495: $$1,1,2,3,3,4,5,6,6,7,7,8,9,10, 10,10, 11,12,12, \ldots$$ \\

496: 

497: \noindent{\bf The Prague clock sequence.}

498: This is not really recursive

499: in the same sense as the preceding sequences, but it seems to fit in here.

500: It is, in any case, delightful.

501: This is \htmladdnormallink{A28354}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=028354}:

502: 1, 2, 3, 4, 32, 123, 43, 2123, 432, 1234, 32123, 43212, 34321, 23432, 123432, 

503: 1234321, 2343212, 3432123, 4321234, 32123432, 123432123, 43212343, 

504: 2123432123, 432123432, 1, 2, 3, 4, 32, $\ldots$

505: According to \cite{Hor}, the sequence indicates how the astronomical clock in Prague strikes the hours.

506: There is a single bell which

507: (i) makes from 1 to 4 strokes at a time,

508: (ii)~the number of strokes follows the sequence

509: $$\ldots 321 2343 212343 \ldots ~,$$

510: (iii)~at the $n$th hour, for $n=1,2, \ldots, 24$, the strokes add to $n$, and

511: (iv)~at the 25th hour there is a single stroke (so the sequence has period 24).

512: As the reader will see by studying the sequence, its existence depends on two coincidences! \\

513: 

514: \noindent{\bf Cayley's mistake.}

515: Since the sequences in the database are numbered A1, A2, A3, $\ldots$, several people humorously proposed the ``diagonal'' sequence

516: (now \htmladdnormallink{A31135}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=031135}) in which the $n$th term is equal to the $n$th term of $An$:

517: $$1,2,1,0,2,3,0,6,8,4,63,1,316,42,16, \ldots ,$$

518: and the even less well-defined sequence (now \htmladdnormallink{A37181}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=037181})

519: with $n$th term equal to \htmladdnormallink{A31135}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=031135}$(n)+1$.

520: I resisted adding these sequences for a long time, partly out of a desire to maintain

521: the dignity of the database, and partly because \htmladdnormallink{A22}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=000022}

522: was only known to 11 terms!

523: 

524: Sequence \htmladdnormallink{A22}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=000022} gives the number of ``centered hydrocarbons with $n$ atoms'',

525: and is based on an 1875 paper of Cayley \cite{Cay75}.

526: The paper is extremely hard to follow, and gives incorrect values for $n=12$ and 13.

527: The errors it contains were reproduced in \cite{BuS65}, \cite{HIS} and \cite{EIS},

528: even though Herrmann \cite{Her80} had pointed out these errors in 1880.

529: As far as we can tell, a correct verion of this sequence was never published until

530: Eric Rains and I did so in 1999 \cite{RS99}.

531: Although Henze and Blair, P\'{o}lya and many others have written articles enumerating related families of chemical compounds (see \cite{RS99} for a brief survey),

532: this sequence seems to have been forgotten for over 100 years.

533: 

534: Once we determined what it was that Cayley was trying to enumerate,

535: P\'{o}lya's counting theory quickly gave the answer,

536: and the correct version of \htmladdnormallink{A22}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=000022} is now in the database (as

537: are the two diagonal sequences mentioned above -- mostly because users 

538: of the database kept proposing them).

539: 

540: \section{Meanders and stamp-folding}

541: The meandric and stamp-folding numbers are similar to the better-known Catalan numbers (sequence \htmladdnormallink{A108}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=000108})

542: in that they are fundamental, easily described combinatorial quantities that arise in

543: many different parts of mathematics, but differ from them in that there is no

544: known formula,

545: and in fact seem to be quite hard to compute.

546: 

547: The stamp-folding problem has a history going back to at least Lucas \cite{Luc}

548: in the nineteenth century, while the meandric problem

549: seems to have been first mentioned by Poincar\'{e} \cite{Poin}.

550: The meandric problem asks:

551: in how many different ways can a river (starting in the South-West and flowing East)

552: cross a road $n$ times?

553: For $n=5$ crossings there are $M_5 =8$ possibilities, shown in Fig. \ref{FMe1}.

554: The sequence $M_1$, $M_2$, $M_3$, $\ldots$ (\htmladdnormallink{A5316}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=005316}) begins

555: $$1,1,2,3,8,14,42,81,262,538,1828, \ldots$$

556: These are called {\em meandric} numbers, since the river {\em meanders} across the road.

557: \begin{figure}[htb]

558: \centerline{\psfig{file=arrows.ps,width=3in}}

559: 

560: \caption{The eight ways a river (going from South-West to North-East) can cross a road five times.}

561: \label{FMe1}

562: \end{figure}

563: The even-numbered terms $M_2$, $M_4$, $M_6 , \ldots$ give the number of different ways an oriented line can cross a Jordan curve

564: (\htmladdnormallink{A5315}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=005315}).

565: There are several other interpretations, one of which is the number of

566: ``simple alternating transit mazes'' \cite{Ph88}, \cite{Ph89}.

567: 

568: The stamp-folding problem asks the same question, but now the line is only semi-infinite.

569: Equivalently, how many ways are there to fold a strip of $n$ stamps?

570: Pictures illustrating the first few terms

571: of the stamp-folding sequence (\htmladdnormallink{A1011}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=001011})

572: can be found on the front cover of \cite{HIS}

573: and in Figure M4587 of \cite{EIS}.

574: The sequence begins

575: $$1,1,2,5,14,38,120, 353,1148, 3527, \ldots$$

576: 

577: No polynomial-time algorithm is known for computing either sequence (on the other hand

578: it is not known that such algorithms do not exist).

579: The best algorithms known require on the order of $nC_n$ steps, where $C_n = \frac{1}{n+1} {\binom{2n}{n}}$ is the $n$th

580: Catalan number (\htmladdnormallink{A108}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=000108}):

581: these algorithms are due to Koehler \cite{Koe} for the stamp-folding problem and to

582: Knuth and Pratt \cite{KP89} and Reeds \cite{RSM} for the meandric problem.

583: They are too complicated to describe here.

584: 

585: Using these algorithms, Stephane Legendre \cite{Leg} has extended the stamp-folding sequence to 26 terms and the meandric sequence to 25 terms.

586: Knuth and Pratt \cite{KP89} have computed 17 terms of the $M_2$, $M_4$, $M_6 , \ldots$ subsequence.

587: Lando and Zvonkin \cite{LZ92}, \cite{LZ93} and Di Francesco, Golinelli and Guitter \cite{FGG96},

588: \cite{FGG95}, \cite{FGG97} have also studied these sequences.

589: 

590: The exact rate of growth of these sequences is not known.

591: The best bounds for $M_{2n}$ presently known

592: appear to be due to Reeds, Shepp and McIlroy \cite{RSM}.

593: It is easy to see that $M_{2n}$ is submultiplicative, and that $C_n \le M_{2n} \le C_n^2$, which implies that

594: $$\mu = \lim_{n \to \infty} M_{2n}^{1/n}$$

595: exists and satisfies $4 \le \mu \le 16$.

596: In \cite{RSM} it is shown that

597: $$8.8 \le \mu \le 13.01 ~.$$

598: 

599: Besides the papers already mentioned, the meandric and stamp-folding problems have recently

600: been discussed by

601: Arnold \cite{Arn88}, Di~Francesco \cite{DiF96},

602: Harris \cite{Har98},

603: Lando and Zvonkin \cite{LZ92}, \cite{LZ93}, Lunnon \cite{Lun68},

604: Sade \cite{Sade},

605: Smith \cite{WDS} and

606: Touchard \cite{Touch50}.

607: 

608: \section{Extremal codes and lattices}

609: Let $C$ be a binary linear self-dual code of length $n$ in which the weight of every codeword

610: is a multiple of 4, and let

611: $$W_C (x,y) = \sum_{c \in C}

612: x^{n- wt(c)} y^{wt(c)}$$

613: be its weight enumerator.

614: Examples are the Hamming code of length 8, with weight enumerator

615: $$f = x^8 + 14 x^4 y^4 + y^8 ~,$$

616: and the Golay code of length 24, with weight enumerator

617: $$g = x^{24} + 759 x^{16} y^8 + 2576 x^{12} y^{12} + 759 x^8 y^{16} + y^{24} ~.$$

618: 

619: A remarkable theorem of Gleason says that the weight enumerator of any 

620: such code $C$ is a polynomial in $f$ and $g$.

621: (References for this section are

622: \cite{SPLAG}, \cite{MS77},

623: \cite{RS98}.)

624: E.g. if $C$ has length $n=72$, its weight enumerator can be written as

625: $$W_C = a_0 f^9 + a_1 f^6 g + a_2 f^3 g^2 + a_3 g^3 ~,$$

626: for rational numbers $a_0, \ldots, a_3$.

627: If we choose these numbers so that the minimal distance of

628: this (hypothetical) code $C$ is as large as possible, we find that

629: $$W_C =1+ 0x^4 + 0x^8 +0x^{12} +

630: 249849 x^{16} + 18106704 x^{20} + 462962955 x^{24} + \cdots ~,

631: $$

632: so that $C$ would have minimal distance 16 (the coefficients form sequence

633: \htmladdnormallink{A18236}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=018236}).

634: The coefficients in this ``extremal'' weight enumerator are all nonnegative integers, but whether a code exists with this weight enumerator is an important unsolved question.

635: 

636: One can perform this calculation for any length that is a multiple of 8, and in \cite{MaS} it was shown that

637: when as many initial terms as possible are set to zero,

638: the next term in the extremal weight enumerator is always

639: positive (so the minimal distance of the hypothetical extremal code is precisely $4[n/24] + 4$).

640: 

641: Lengths that are multiples of 24 are especially interesting.

642: Figure \ref{FEC1} shows the leading term in the extremal weight enumerator

643: for a binary self-dual code of length $n=24m$

644: (this is sequence \htmladdnormallink{A34414}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=034414}).

645: \begin{figure}[htb]

646: $$

647: \begin{array}{r@{~~~~}r}

648: n & \multicolumn{1}{c}{\mbox{Coefficient}} \\ \hline

649: 0 & 1 \\

650: 24 & 759 \\

651: 48 & 17296 \\

652: 72 & 249849 \\

653: 96 & 3217056 \\

654: 120 & 39703755 \\

655: 144 & 481008528 \\

656: 168 & 5776211364 \\

657: 192 & 69065734464

658: \end{array}

659: $$

660: 

661: \caption{Number of codewords of minimal weight $n/6 +4$ in extremal weight enumerator of length $n$.}

662: \label{FEC1}

663: \end{figure}

664: It is ``obvious'' that these numbers are growing rapidly, and in fact it is shown in \cite{MaS} that if $n=24m$ then the leading

665: coefficient is

666: $${\binom{24m}{5}} {\binom{5m-2}{m-1}} / {\binom{4m+4}{5}} ~.$$

667: The next term in the extremal weight enumerator (the number of codewords

668: of weight $4[n/24] +8$) also grows rapidly --- see Fig \ref{FEC2} (sequence \htmladdnormallink{A34415}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=034415}).

669: \begin{figure}[htb]

670: $$

671: \begin{array}{r@{~~~~}r}

672: n & \multicolumn{1}{c}{\mbox{Coefficient}} \\ \hline

673: 0 & 1 \\

674: 24 & 2576 \\

675: 48 & 535095 \\

676: 72 & 18106704 \\

677: 96 & 369844880 \\

678: 120 & 6101289120 \\

679: 144 & 90184804281 \\

680: 168 & 1251098739072 \\

681: 192 & 16681003659936

682: \end{array}

683: $$

684: 

685: \caption{Number of codewords of next-to-minimal weight $n/6+8$ in extremal weight enumerator of

686: length $n$.}

687: \label{FEC2}

688: \end{figure}

689: 

690: Again it is ``obvious'' that these numbers also grow rapidly, especially if one examines a more extensive

691: table, where the number of  digits continues to increase with 

692: each term.

693: Yet --- this shows the danger of drawing conclusions just from numerical data ---

694: it is proved in \cite{MOS} that this sequence goes negative at $n= 3696$ and stays

695: negative forever.

696: (See Theorem 29 of \cite{RS98} for more precise information.)

697: 

698: So codes corresponding to these extremal weight enumerators

699: certainly do not exist for $n \ge 3696$ (since the

700: weight enumerator of a genuine code must have nonnegative

701: coefficients).

702: They are known to exist for $n=24$ (the Golay code)

703: and 48 (a quadratic residue code),

704: but every case from 72 to 3672 is open.

705: For further details, and in particular for a description

706: of the analogous situation when $n$ is a multiple of 8 but not of 24, see \cite{RS98}.

707: 

708: 

709: The situation for lattices is similar.

710: Consider an even unimodular lattice $\Lambda$ of dimension $n$, with theta series

711: $$\Theta_\Lambda (q) = \sum_{u \in \Lambda} q^{u \cdot u} ~.$$

712: Examples are the eight-dimensional root lattice $E_8$, with theta series

713: \begin{eqnarray*}

714: f & = & 1 + 240 q^2 + 2160q^4 + 6720 q^6 + \cdots \\

715: & = & 1+ 240 \sum_{m=1}^\infty \sigma_3 (m) q^{2m} ~,

716: \end{eqnarray*}

717: where $\sigma_3 (m)$ is the sum of the cubes of the divisors of $m$,

718: and the

719: 24-dimensional Leech lattice, with theta series

720: \begin{eqnarray*}

721: g & = & 1+ 196560 q^4 + 16773120q^6 +

722: 398034000 q^8 + \cdots \\

723: & = & f^3 - 720 q^2 \prod_{m=1}^\infty (1-q^{2m} )^{24} ~.

724: \end{eqnarray*}

725: The Taylor series expansion of the last product defines the famous Ramanujan numbers.

726: (The coefficients of $f$ and $g$ give sequences \htmladdnormallink{A4009}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=004009} and \htmladdnormallink{A8408}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=008408}, respectively;

727: the Ramanujan numbers form sequence \htmladdnormallink{A594}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=000594}.)

728: 

729: A theorem due essentially to Hecke says that the theta series of any such lattice

730: $\Lambda$ is a polynomial in $f$ and $g$.

731: We may then define extremal theta series just as we defined extremal weight enumerators.

732: It is known \cite{MOS} that the leading term

733: in the extremal theta series is positive, but that

734: again the next coefficient eventually becomes negative (even

735: though the sequence begins with about 1700 exponentially

736: growing terms).

737: The dimension where the first negative coefficient

738: appears is around 41000.

739: The leading terms are shown in Fig.~\ref{FEC3} (sequence \htmladdnormallink{A34597}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=034597}).

740: \begin{figure}[htb]

741: $$

742: \begin{array}{r@{~~~~}r}

743: n & \multicolumn{1}{c}{\mbox{Coefficient}} \\ \hline

744: 0 & 1 \\

745: 24 & 196560 \\

746: 48 & 52416000 \\

747: 72 & 6218175600 \\

748: 96 & 565866362880 \\

749: 120 & 45792819072000

750: \end{array}

751: $$

752: 

753: \caption{Number of shortest vectors $u$ (with $u \cdot u = n/12 +2$) in extremal theta series at dimension $n$.}

754: \label{FEC3}

755: \end{figure}

756: 

757: Lattices corresponding to the extremal theta series exist

758: for $n=24$ (the Leech lattice) and 48 (at least three

759: inequivalent lattices exist), but every

760: case from 72 to about 41000 is open.

761: The 24- and 48-dimensional extremal lattices

762: provide record kissing numbers in those dimensions;

763: it would be nice to have a 72-dimensional example.

764: 

765: \section{Levine's sequence}

766: In the summer of 1997 Lionel Levine \cite{Lev97}

767: submitted a new sequence to the table,

768: a sequence of such beauty that it immediately caught

769: the attention of several of my colleagues.

770: It is constructed via the array in Fig.~\ref{FL}.

771: \begin{figure}[htb]

772: $$

773: \begin{array}{|ccccccccccccccccccccccc|} \hline

774: &&&&&&&&&&&&&&&&&&&&&1 & \underline{1} \\ \hline

775: &&&&&&&&&&&&&&&&&&&&&1 & \underline{2} \\ \hline

776: &&&&&&&&&&&&&&&&&&&&1 & 1 & \underline{2} \\ \hline

777: &&&&&&&&&&&&&&&&&&&1 & 1 & 2 & \underline{3} \\ \hline

778: &&&&&&&&&&&&&&&&1 & 1 & 1 & 2 & 2 & 3 & \underline{4} \\ \hline

779: &&&&&&&&&1 & 1 & 1 & 1 & 2 & 2 & 2 & 3 & 3 & 4 & 4 & 5 & 6 & \underline{7} \\ \hline

780: 1& 1 & 1 & 1 & 1 & 1 & 1 & 2 & 2 & 2 & 2 & 2 & 2 & 3 & 3 & 3 & 3 & 3 & 4 & 4 & ~ & ~ & ~ \\

781: && 4 & 4 & 5 & 5 & 5 & 5 & 6 & 6 & 6 & 7 & 7 & 7 & 8 & 8 & 9 & 9 & 10 & 11 & 12 & 13 & \underline{14} \\  \hline

782: \end{array}

783: $$

784: \caption{The array that produces Levine's sequence

785: $1,2,2,3,4,7,14, \ldots$.}

786: \label{FL}

787: \end{figure}

788: If a row of the array contains the numbers

789: $$a_1 ~~a_2 ~~ a_3 ~ \ldots ~ a_{k-1} ~~ a_k$$

790: then the next row contains

791: $$

792: \mbox{$a_k$  1's,}

793: \quad

794: \mbox{$a_{k-1}$  2's,} \quad

795: \mbox{$a_{k-2}$  3's, $\ldots$}

796: $$

797: Levine's sequence (\htmladdnormallink{A11784}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=011784}) is obtained by taking the last term in each row:

798: \begin{equation}\label{EqLe1}

799: \begin{array}{l}

800: 1, 2, 2, 3, 4, 7, 14, 42, 213, 2837, 175450, 

801: 139759600, 6837625106787, \\

802: 266437144916648607844, 

803: 508009471379488821444261986503540, ...

804: \end{array}

805: \end{equation}

806: 

807: The terms grow unexpectedly rapidly!

808: The $n$th term $L_n$ is

809: \begin{itemize}

810: \item[(i)]

811: the sum of the elements in row $n-2$

812: \item[(ii)]

813: the number of elements in row $n-1$

814: \item[(iii)]

815: the last element in row $n$

816: \item[(iv)]

817: the number of 1's in row $n+1$

818: 

819: $\ldots$

820: \end{itemize}

821: Furthermore, if $s(n,i)$ denotes the sum of the first $i$ elements in

822: row $n$, then we have

823: \begin{itemize}

824: \item[(v)]

825: $L_{n+2} = s (n, L_{n+1} )$

826: \item[(vi)]

827: $L_{n+3} = \displaystyle\sum_{i=1}^{L_{n+1}} s(n,i)$

828: \item[(vii)]

829: $L_{n+4} = \displaystyle\sum_{i=1}^{L_{n+1}} {\displaystyle\binom{s(n,i)+1}{2}}$.

830: \end{itemize}

831: The latter identity was found by Allan Wilks

832: \cite{Wilks}, who also found a more complicated formula for

833: $L_{n+5}$, and used it to compute the last two terms

834: shown in \eqn{EqLe1}.

835: No other terms are known!

836: 

837: As to the rate of growth, we have only a crude estimate.

838: Bjorn Poonen and Eric Rains \cite{PoRa97} showed that

839: \begin{equation}\label{EqL3}

840: \log L_n \sim c \tau^n ~,

841: \end{equation}

842: where $\tau = (1+ \sqrt{5})/2$.

843: {\em Sketch of Proof.}

844: (a)~$L_{n+2} \le L_{n+1} L_n$, and so $\log L_n$ is bounded above by a Fibonacci-like sequence.

845: (b)~The sum of the $(n+1)$st row is at most

846: $$

847: {\binom{\left[ \frac{L_{n+2}}{L_n} \right] +1}{2}} L_n ~,

848: $$

849: which implies

850: $$\frac{L_{n+3}}{2L_{n+2}} \ge \frac{L_{n+2}}{2L_{n+1}} \frac{L_{n+1}}{2L_n}$$

851: and so $\log (L_{n+1} / 2L_n )$ is bounded below by a Fibonacci-like sequence.

852: 

853: Colin Mallows \cite{clm} has determined numerically that a reasonably

854: good approximation to $L_n$ is given by

855: $$\frac{1}{c_1} e^{c_2 \tau^n}$$

856: where $c_1 \approx 0.277$, $c_2 \approx 0.05427$.

857: It would be nice to have better estimates for $L_n$, and one or more

858: additional terms.

859: 

860: \section{Arrays of Numbers}

861: Besides number {\em sequences}, the database also contains {\em arrays}

862: of numbers that have been converted to sequences.

863: Triangular arrays are read by rows, in the obvious way.

864: E.g. Pascal's triangle of binomial coefficients

865: $$

866: \begin{array}{ccccccccc}

867: ~ & ~ & ~ & ~ & 1 \\

868: ~ & ~ & ~ & 1 & ~ & 1 \\

869: ~ & ~ & 1 & ~ & 2 & ~ & 1 \\

870: ~ & 1 & ~ & 3 & ~ & 3 & ~ & 1 \\

871: 1 & ~ & 4 & ~ & 6 & ~ & 4 & ~ & 1 \\

872: \end{array}

873: $$

874: becomes sequence \htmladdnormallink{A7318}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=007318}:

875: $$1,1,1,1,2,1,1,3,3,1,1,4,6,4,1, \ldots$$

876: 

877: Square arrays are read by antidiagonals, usually in this order:

878: $$

879: \begin{array}{lllll}

880: a_0 & a_2 & a_5 & a_9 & \ldots \\

881: a_1 & a_4 & a_8 & \multicolumn{2}{l}{\ldots} \\

882: a_3 & a_7 & \multicolumn{3}{l}{\ldots} \\

883: a_6 & \multicolumn{4}{l}{\ldots} \\

884: \multicolumn{5}{l}{\ldots}

885: \end{array}

886: $$

887: E.g. the Nim-addition table \cite{ONAG}

888: $$\begin{array}{cccccc}

889: 0 & 1 & 2 & 3 & 4 & \ldots \\

890: 1 & 0 & 3 & 2 & 5 & \ldots \\

891: 2 & 3 & 0 & 1 & 6 & \ldots \\

892: 3 & 2 & 1 & 0 & 7 & \ldots \\

893: \cdot & \cdot & \cdot

894: \end{array}

895: $$

896: becomes sequence \htmladdnormallink{A3987}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=003987}:

897: $$0,1,1,2,0,2,3,3,3,3,4,2,0,2,4, \ldots$$

898: 

899: Other classical arrays are the Stirling numbers of both kinds, Eulerian numbers,

900: etc.

901: 

902: A less well-known array arises from {\bf Gilbreath's conjecture}.

903: This conjecture states that if one writes down the primes in a row, and underneath the absolute values of the differences, as in Fig. \ref{FG},

904: then the leading terms (shown $\underline{\mbox{underlined}}$)

905: of all rows except the first are equal to 1 (\cite{UPNT}, \S A10).

906: The corresponding sequence (\htmladdnormallink{A36262}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=036262}) is

907: $$2,1,3,1,2,5,1,0,2,7,1,2,2,4,11,1, \ldots$$

908: Odlyzko \cite{Od93} has verified the conjecture out to $3 \times 10^{11}$.

909: \begin{figure}[htb]

910: $$

911: \begin{array}{ccccccccccccccccccc}

912: 2~~~ & ~~~~ & 3~~~  & ~~~~ & 5~~~ & ~ & 7~~~ & ~~~~ & 11~~ & ~ & 13~~ & ~ & 17~~ & ~ & 19~~ & ~ & 23~~ & ~~~~ & \ldots \\

913: ~ & \underline{1} & ~ & 2 & ~ & 2 & ~ & 4 & ~ & 2 & ~ & 4 & ~ & 2 & ~ & 4 & ~ & \ldots \\

914: ~ & ~ & \underline{1} & ~ & 0 & ~ & 2 & ~ & 2 & ~ & 2 & ~ & 2 & ~ & 2 & ~ & \ldots \\

915: ~ & ~ & ~ & \underline{1} & ~ & 2 & ~ & 0 & ~ & 0 & ~ & 0 & ~ &0 & ~ & \ldots \\

916: ~ & ~ & ~ & ~ & \underline{1} & ~ & 2 & ~ & 0 & ~ & 0 & ~ & 0 & ~ & \ldots \\

917: ~ & ~ & ~ & ~ & ~ &  \underline{1} & ~ & 2 & ~ & 0 & ~ & 0 & ~ & \ldots \\

918: \multicolumn{16}{c}{\ldots}

919: \end{array}

920: $$

921: \caption{Gilbraith's conjecture is that the leading

922: terms of all rows in this array

923: except the first are always 1 (the top row contains the primes, subsequent rows are the absolute

924: values of the differences of the previous row).}

925: \label{FG}

926: \end{figure}

927: 

928: \section{The Wythoff array}

929: This array shown in Fig. \ref{FW} has many wonderful properties, some of which are mentioned here.

930: I learned about most of these properties from John Conway

931: \cite{JHC}, but this array has a long history

932: --- see Fraenkel and Kimberling \cite{FrKi94},

933: Kimberling \cite{Kim93},

934: \cite{Kim93a}, \cite{Kim94},

935: \cite{Kim95}, \cite{Kim95a},

936: \cite{Kim95b},

937: \cite{Kim97}, Morrison \cite{Mor80} and Stolarsky \cite{Sto76},

938: \cite{Sto77}.

939: It is related to a large number of sequences in the database (the main

940: entry is \htmladdnormallink{A35513}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=035513}).

941: \begin{figure}[htb]

942: $$

943: \begin{array}{rr|rrrrrrrrr}

944: 0 & 1 & 1 & 2 & 3 & 5 & 8 & 13 & 21 & 34 & 55 \\

945: 1 & 3 & 4 & 7 & 11 & 18 & 29 & 47 & 76 & .. \\

946: 2 & 4 & 6 & 10 & 16 & 26 & 42 & 68 & .. \\

947: 3 & 6 & 9 & 15 & 24 & 39 & 63 & .. \\

948: 4 & 8 & 12 & 20 & 32 & 52 & .. \\

949: 5 & 9 & 14 & 23 & 37 & 60 & .. \\

950: 6 & 11 & 17 & 28 & 45 & 73 & .. \\

951: 7 & 12 & 19 & 31 & 50 & 81 & .. \\

952: . & .. & .. & .. & .. & .. & ..

953: \end{array}

954: $$

955: \caption{The Wythoff array.}

956: \label{FW}

957: \end{figure}

958: 

959: \paragraph{Construction (1).}

960: The two columns to the left of the vertical

961: line consist respectively of the nonnegative

962: integers $n$, and the

963: {\em lower Wythoff sequence}

964: (\htmladdnormallink{A201}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=000201}), whose $n$th term is $[(n+1) \tau ]$.

965: The rows are then filled in by the Fibonacci rule that each term is the sum of the two previous terms.

966: The entry $n$ in the first column is the {\em index} of that row.

967: 

968: \paragraph{Definition.}

969: The {\em Zeckendorf expansion} of a number $n$ is obtained by repeatedly subtracting the

970: largest possible Fibonacci number until nothing

971: remains.

972: E.g. $100 = 89+8+3= F_{11} + F_6 + F_4$.

973: The {\em Fibonacci successor} to $n$, $Sn$, say, is found by replacing

974: each $F_i$ in the Zeckendorf expansion by $F_{i+1}$.

975: E.g. the Fibonacci successor to 100 is $S100 = F_{12} + F_7 + F_5 = 144 + 13 + 5 = 162$.

976: 

977: \paragraph{Construction (2).}

978: The two columns to the left of the vertical line in Fig. \ref{FW} read $n$, $1+Sn$;

979: then after the vertical line the row continues

980: $$

981: m \quad Sm \quad SSm \quad SSSm \quad SSSSm \quad \ldots ~,$$

982: where $m=n+1 +Sn$.

983: \paragraph{Construction (3).}

984: Let $\{S1, S2, S3, \ldots \} = \{2,3,5,7,8,10,11, \ldots \}$ be the sequence of

985: Fibonacci successors (\htmladdnormallink{A22342}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=022342}).

986: The first column to the right of the line consists of the numbers not in that sequence: 1, 4, 6, 9, 12, $\ldots$ (\htmladdnormallink{A7067}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=007067}).

987: The rest of each row is filled in by repeatedly applying $S$.

988: \paragraph{Construction (4).}

989: The entry in row $n$ and column $k$ is

990: $$[(n+1) \tau F_{k+2} ] + F_{k+1} n$$

991: (where $k=0$ indicates the first column to the right of the vertical line).

992: 

993: Some properties of the array to the right of the line

994: are the following:

995: \begin{itemize}

996: \item[(i)]

997: The first row consists of the

998: Fibonacci sequence $1,2,3,5,8, \ldots$

999: \item[(ii)]

1000: Every row satisfies the Fibonacci recurrence.

1001: \item[(iii)]

1002: The leading term in each row is the smallest number not found

1003: in any earlier row.

1004: \item[(iv)]

1005: Every positive integer appears exactly once.

1006: \item[(v)]

1007: The terms in any row or column are monotonically increasing.

1008: \item[(vi)]

1009: Every positive Fibonacci-type sequence (i.e. satisfying

1010: $a(n)=a(n-1)+a(n-2)$ and eventually positive) appears

1011: as some row of the array.

1012: \item[(vii)]

1013: The terms in any two adjacent rows alternate.

1014: \end{itemize}

1015: There are infinitely many arrays with Properties 1--7, see \cite{Kim95a}.

1016: 

1017: The $n$th term of the {\em vertical para-Fibonacci sequence}

1018: $$

1019: 0, 0, 0, 1, 0, 2, 1, 0, 3, 2, 1, 4, 0, 5, 3, 2, 6, 1, 7, 4, 0, 8, 5,  \ldots

1020: $$

1021: (\htmladdnormallink{A19586}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=019586})

1022: gives the index (or parameter) of the row of the Wythoff array that contains $n$.

1023: This sequence also has some nice fractal-like properties:

1024: 

1025: (a)~If you delete the first occurrence of each number, the

1026: sequence is unchanged.

1027: Thus if we delete the underlined numbers from

1028: $$

1029: \underline{0}, 0, 0, \underline{1}, 0, \underline{2}, 1, 0,

1030: \underline{3}, 2, 1, \underline{4}, 0, \underline{5},

1031: 3, 2, \underline{6}, 1, \underline{7}, 4, 0, \underline{8}, 5, \ldots

1032: $$

1033: we get

1034: $$

1035: 0, 0, 0, 1, 0, 2, 1, 0, 3, 2, 1, 4, 0, 5, 3, 2, 6, 1, 7, 4, 0, 8, 5, \ldots

1036: $$

1037: again!

1038: 

1039: (b)~Between any two consecutive  0's

1040: we see a permutation of the first few positive integers, and these

1041: nest, so the sequence can be rewritten as (read across the rows):

1042: $$

1043: \begin{array}{ccccccccccccc}

1044: 0 \\

1045: 0 \\

1046: 0& ~ & ~ & ~ & ~ & ~ & ~ && 1 \\

1047: 0& ~ & ~ & ~ & ~ & 2 & ~ & ~ & 1 \\

1048: 0& ~ & ~ & 3 & ~ & 2 & ~ & ~ & 1 & ~ & ~ & 4 \\

1049: 0& ~ & 5 & 3 & ~ & 2 & ~ & 6 & 1 & ~ & 7 & 4 \\

1050: 0 & 8 & 5 & 3 & 9 & 2 & 10 & 6 & 1 & 11 & 7 & 4 & 12

1051: \end{array}$$

1052: 

1053: 

1054: The $n$th term of the {\em horizontal para-Fibonacci sequence}

1055: $$

1056: 1, 2, 3, 1, 4, 1, 2, 5, 1, 2, 3, 1, 6, 1, 2, 3, 1, 4, 1, 2, 7, 1, 2,  \ldots

1057: $$

1058: (\htmladdnormallink{A35612}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=035612})

1059: gives the index (or parameter) of the column of the Wythoff

1060: array that contains $n$.

1061: This sequence also has some nice properties.

1062: 

1063: I hope I have said enough to convince you that the Wythoff array is

1064: well worth studying and full of surprises.

1065: 

1066: \section{The Boustrophedon transform}

1067: The Taylor series for $\sin x$ and $\cos x$ are easily remembered,

1068: but most people have trouble with

1069: \begin{eqnarray*}

1070: \tan x & = &

1071: 1x + 2 \frac{x^3}{3!} + 16 \frac{x^5}{5!} + 272 \frac{x^7}{7!} + \cdots ~, \\

1072: {\rm sec} ~x & = & 1 + 1 \frac{x^2}{2!} +

1073: 5 \frac{x^4}{4!} + 61 \frac{x^6}{6!} + \cdots ~.

1074: \end{eqnarray*}

1075: However, their coefficients can be calculated from the array in Fig. \ref{FB}.

1076: \begin{figure}[htb]

1077: $$

1078: \begin{array}{ccccccccccccccc}

1079: ~ & ~ & ~ & ~ & ~ & ~ & ~ & 1 \\

1080: ~ & ~ & ~ & ~ & ~ & ~ & 0 & ~ & 1 \\

1081: ~ & ~ & ~ & ~ & ~ & 1 & ~ & 1 & ~ & 0 \\

1082: ~ & ~ & ~ & ~ & 0 & ~ & 1 & ~ & 2 & ~ & 2 \\

1083: ~ & ~ & ~ & 5 & ~ & 5 & ~ & 4 & ~ & 2 & ~ & 0 \\

1084: ~ & ~ & 0 & ~ & 5 & ~ & 10 & ~ & 14 & ~ & 16 & ~ & 16 \\

1085: ~ & 61 & ~ & 61 & ~ & 56 & ~ & 46 & ~ & 32 & ~ & 16 & ~ & 0 \\

1086: 0 & ~ & 61 & ~ & 122 & ~ & 178 & ~ & 224 & ~ & 256 & ~ & 272 & ~ & 272 \\

1087: \multicolumn{15}{c}{\cdots}

1088: \end{array}

1089: $$

1090: \caption{The secant-tangent triangle.}

1091: \label{FB}

1092: \end{figure}

1093: The nonzero entries on the left are the {\em secant numbers} (\htmladdnormallink{A364}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=000364}):

1094: $$1,1,5,61,1385,50521, 2702765, \ldots ,$$

1095: and those on the right are the

1096: {\em tangent numbers} (\htmladdnormallink{A182}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=000182})

1097: $$1,2, 16, 272, 7936, 353792, 22368256, \ldots$$

1098: while the combination of the two sequences (\htmladdnormallink{A111}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=000111}):

1099: \begin{equation}\label{EqB1}

1100: 1,1,1,2,5,16,61,272,1385, 7936, \ldots

1101: \end{equation}

1102: are usually called the {\em Entringer numbers}.

1103: The latter count permutations of

1104: \linebreak

1105: $\{1,2, \ldots, n \}$ that

1106: alternately fall and rise.

1107: 

1108: This array is filled by a rule somewhat similar to that for Pascal's

1109: triangle:

1110: the rows are scanned alternately from right to left

1111: and left to right,

1112: the leading entry in each row is 0,

1113: and every subsequent entry is the sum of the previous

1114: entry in the same row and the entry above it in the previous row.

1115: (This is the {\em boustrophedon} or ``ox-plowing'' rule.)

1116: The earliest reference I have seen to this triangle is Arnold \cite{Arn91},

1117: who calls it the Euler-Bernoulli

1118: triangle.

1119: However, it may well be much older origin.

1120: \cite{bous} gives many other references.

1121: 

1122: Richard Guy \cite{Guy95} observed that if the entries at the beginnings of the rows

1123: are changed from $1,0,0, \ldots$ to

1124: say $1,1,1,1,1, \ldots$ or to $1,2,4,8,16, \ldots$ then the

1125: numbers that appear at the ends of the rows form interesting-looking

1126: sequences that were not to be found in \cite{EIS}, and asked if they had a combinatorial interpretation.

1127: Using $1,1,1, \ldots$ for example the triangle becomes

1128: $$

1129: \begin{array}{ccccccccccc}

1130: ~ & ~ & ~ & ~ & ~ & 1 \\

1131: ~ & ~ & ~ & ~ & 1 & ~ & 2 \\

1132: ~ & ~ & ~ & 4 & ~ & 3 & ~ & 1 \\

1133: ~ & ~ & 1 & ~ & 5 & ~ & 8 & ~ & 9 \\

1134: ~ & 24 & ~ & 23 & ~ & 18 & ~ & 10 & ~ & 1 \\

1135: 1 & ~ & 25 & ~ & 48 & ~ & 66 & ~ & 76 & ~ & 77 \\

1136: \multicolumn{11}{c}{\cdots}

1137: \end{array}

1138: $$

1139: yielding the sequence (\htmladdnormallink{A667}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=000667})

1140: \begin{equation}\label{EqB2}

1141: 1,2,4,9,24,77, 294, 1309, \ldots

1142: \end{equation}

1143: We may regard this process as carrying out a transformation (the

1144: {\em boustrophedon transform})

1145: of sequences: if

1146: the numbers at the beginnings of the rows are $a_0$, $a_1$, $a_2 , \ldots$

1147: (the input sequence) then the numbers at the ends of the rows,

1148: $b_0$, $b_1$, $b_2, \ldots$ (say) are the output

1149: sequence.

1150: In \cite{bous} we showed that there is a simple relationship

1151: between the input and

1152: output sequences:

1153: their exponential generating functions

1154: $${\cal A} (x) = \sum_{n=0}^\infty a_n \frac{x^n}{n!} , \quad

1155: {\cal B} (x) = \sum_{n=0}^\infty b_n \frac{x^n}{n!}

1156: $$

1157: are related by

1158: $${\cal B} (x) = ({\rm sec}~x + \tan x) {\cal A} (x) ~.$$

1159: We also give a combinatorial interpretation of the $\{b_n\}$.

1160: E.g. in \eqn{EqB2}, $b_n$ is the number of up-down subsequences

1161: of $\{1, \ldots, n\}$, so that $b_3 =9$ corresponds to

1162: $\emptyset ,1,2,3,12,13,23,132,231$.

1163: 

1164: The Entringer sequence \eqn{EqB1} then has the property that it shifts

1165: one place left under the boustrophedon transform.

1166: The lexicographically earliest sequence that shifts {\em two}

1167: places left under this transform (\htmladdnormallink{A661}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=000661}) is

1168: $$1,0,1,1,2,6,17, 62, 259, 1230, 6592, \ldots$$

1169: We do not know what this enumerates!

1170: 

1171: Many examples of similar ``eigen-sequences'' for other transformations

1172: of sequences can be found in Donaghey \cite{MB13},

1173: Cameron \cite{MB6}, and especially \cite{eigen}.

1174: E.g. the sequence giving the number of planted

1175: achiral trees \cite{MB19}, \cite{MB25} (\htmladdnormallink{A3238}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=003238}):

1176: $$1,1,2,3,5,6,10,11,16,19,26, \ldots$$

1177: has the property that it shifts left one place under the ``inverse M\"{o}bius transformation'' given by

1178: $$b_n = \sum_{d|n} a_d ~.$$

1179: 

1180: \section{Tchoukaillon solitaire (or Mancala, or Kalahari)}

1181: These are ancient board games, with hundreds of variants

1182: and many different names.

1183: The version to be described here is called Tchoukaillon solitaire.

1184: It has been studied by several authors (see for example

1185: Betten \cite{Bet88} and Broline and Loeb \cite{BL95}).

1186: It is played on a board with a row of holes numbered

1187: $0,1,2, \ldots$ (see Fig. \ref{FTc1}).

1188: \begin{figure}[htb]

1189: \centerline{\psfig{file=FTc1.ps,width=4in}}

1190: \caption{A move in Tchoukaillon solitaire.}

1191: \label{FTc1}

1192: \end{figure}

1193: 

1194: The game begins with $n$ stones placed anywhere except in hole 0.

1195: A move consists in picking up the stones in some hole and placing one in each lower-numbered hole.

1196: If the last stone falls in hole 0 then play continues, otherwise the game is lost.

1197: The objective is to get all the stones into hole 0.

1198: 

1199: The game is interesting because there is a unique winning position for any number of stones.

1200: These winning positions are shown in Fig. \ref{FTc2} (sequence \htmladdnormallink{A28932}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=028932}), and can be found by playing the game backwards.

1201: \begin{figure}[htb]

1202: $$

1203: \begin{array}{r@{~~~~}r}

1204: n & \multicolumn{1}{c}{\mbox{Position}} \\ \hline

1205: 0 & 0 \\

1206: 1 & 1 \\

1207: 2 & 20 \\

1208: 3 & 21 \\

1209: 4 & 310 \\

1210: 5 & 311 \\

1211: 6 & 4200 \\

1212: 7 &4201 \\

1213: 8 &4220 \\

1214: 9 &4221 \\

1215: 10 &53110 \\

1216: 11 &53111 \\

1217: 12 &642000 \\

1218: 13 & 642001 \\

1219: .. & ...~~~~

1220: \end{array}

1221: $$

1222: 

1223: \caption{The unique winning position for $n$ stones in Tchoukaillon solitaire.}

1224: \label{FTc2}

1225: \end{figure}

1226: 

1227: The array can be more explicitly constructed by the rule that if the first 0

1228: in a row (counting from the right)

1229: is in position $i$, then the next row is obtained by writing $i$ in position $i$ and subtracting 1 from all earlier positions.

1230: The sequence of successive values of $i$ (\htmladdnormallink{A28920}{http://www.research.att.com/cgi-bin/access.cgi/as/njas/sequences/eisA.cgi?Anum=028920}) is

1231: $$1,2,1,3,1,4,1,2,1,5,1,6,1,2, \ldots$$

1232: 

1233: Let $t(k)$ denote the position where $k$ occurs for the first time in this sequence.

1234: The values of $t(1)$, $t(2)$, $t(3), \ldots$ are (sequence A2491):

1235: $$1,2,4,6,10,12,18,22,30,34,42, \ldots$$

1236: This sequence has some very nice properties.

1237: It has been investigated by (in addition to the references mentioned above)

1238: David \cite{Dav57}, Erd\H{o}s and Jabotinsky \cite{ErJa58} and

1239: Smarandache \cite{smar}.

1240: 

1241: (i)

1242: $t(n)$ can be obtained by starting with $n$ and successively

1243: rounding up to the next multiple of $n-1$, $n-2, \ldots, 2,1$.

1244: E.g. if $n=10$, we obtain

1245: $$10 \to 18 \to 24 \to 28 \to 30 \to 30 \to 32 \to 33 \to 34 \to 34 ~,$$

1246: so $t(10) =34$.

1247: 

1248: (ii)

1249: The sequence can be obtained by a sieving process:

1250: write $1,2, \ldots$ in a column.

1251: To get the second column, cross off 1 and every second number.

1252: To get the third column, cross off the first and every third number.

1253: Then cross off the first and every fourth number, and so on (see Fig. \ref{FTc3}).

1254: The top number in column $n$ is $t(n)$.

1255: Comparison of Figures \ref{FTc2} and \ref{FTc3} shows that connection with the solitaire game.

1256: \begin{figure}[htb]

1257: \centerline{\psfig{file=FTc3.ps,width=4in}}

1258: \caption{A sieve to generate the sequence $t(1), t(2), \ldots = 1,2,4,6,10,12,18, \ldots$.

1259: At stage $n$, the first number and every $n$th are crossed off.}

1260: \label{FTc3}

1261: \end{figure}

1262: 

1263: (iii)

1264: Finally, Broline and Loeb \cite{BL95} (extending the work of

1265: the other authors mentioned) show that, for large $n$,

1266: $$t(n) = \frac{n^2}{\pi} + O(n) ~.$$

1267: It is a pleasant surprise to see $\pi$ emerge from such a simple game.

1268: 

1269: \begin{thebibliography}{15} 

1270: 

1271: \bibitem{Arn88}

1272: V. I. Arnold, A branched covering of ${\mathbb C} {\mathbb P}^2 \to S^4$,

1273: hyperbolicity and projective topology [Russian],

1274: {\em Sibir. Mat. Zhurn.} {\bf 29} (No. 2, 1988)

1275: 36--47 $=$ {\em Siberian Math. J.} {\bf 29} (1988), 717--725.

1276: 

1277: \bibitem{Arn91}

1278: V. I. Arnold, Bernoulli-Euler updown numbers associated with

1279: function singularities, their combinatorics and arithmetics,

1280: {\em Duke Math. J.} {\bf 63} (1991), 537--555.

1281: 

1282: \bibitem{Baa1}

1283: M. Baake, Solution of the coincidence problem

1284: in dimensions $d \le 4$, pp. 199--237 of

1285: R. V. Moody, ed., {\em The Mathematics of Long-Range

1286: Periodic Order},

1287: Kluwer, Dordrecht, 1997.

1288: 

1289: \bibitem{eigen}

1290: M. Bernstein and N. J. A. Sloane,

1291: \htmladdnormallink{Some canonical sequences of integers}{http://www.research.att.com/~njas/doc/eigen.pdf}, {\em Linear Alg. App.} {\bf 226--228}

1292: (1995), 57--72.

1293: 

1294: \bibitem{Bet88}

1295: D. Betten,

1296: Kalahari and the sequence ``Sloane No. 377'',

1297: {\em Ann. Discrete Math.} {\bf 37} (1988), 51--58.

1298: 

1299: \bibitem{BL95}

1300: D. M. Roline and D. E. Loeb,

1301: The combinatorics of Mancala-type games:

1302: Ayo, Tchoukaillon, and $1 / \pi$,

1303: {\em J. Undergrad. Math. Applic.} {\bf 16} (1995), 21--36.

1304: 

1305: \bibitem{BuS65}

1306: R. G. Busacker and T. L. Saaty,

1307: {\em Finite Graphs and Networks}, McGraw-Hill, NY, 1965, see p.~201.

1308: 

1309: 

1310: \bibitem{MB6}

1311: P. J. Cameron,

1312: Some sequences of integers, {\em Discrete Math.}

1313: {\bf 75} (1989), 89--102.

1314: 

1315: \bibitem{Cay75}

1316: A. Cayley,

1317: Ueber die analytischen Figuren,

1318: welche in der Mathematik B\"{a}ume genannt werden und ihre Anwendung auf die Theorie

1319: chemischer Verbindungen,

1320: {\em Ber. deutsch. chem. Ges.} {\bf 8} (1875), 1056--1059.

1321: 

1322: \bibitem{Cipra}

1323: B. Cipra,

1324: Mathematicians get an on-line fingerprint file,

1325: {\em Science}

1326: {\bf 265} (22 July 1994), 473.

1327: 

1328: \bibitem{ONAG}

1329: J. H. Conway,

1330: {\em On Numbers and Games},

1331: Academic Press, NY, 1976.

1332: 

1333: \bibitem{Bell}

1334: J. H. Conway,

1335: Some crazy sequences,

1336: videotaped talk at AT\&T Bell Labs,

1337: July 15, 1988.

1338: 

1339: \bibitem{JHC}

1340: J. H. Conway, personal communication, 1999.

1341: 

1342: \bibitem{MM}

1343: J. H. Conway and S. P. Norton,

1344: Monstous moonshine, {\em Bull. London Math. Soc.} {\bf 11} (1979), 308--339.

1345: 

1346: \bibitem{CRS}

1347: J. H. Conway,

1348: E. M. Rains,

1349: N. J. A. Sloane,

1350: \htmladdnormallink{On the existence of similar sublattices}{http://www.research.att.com/~njas/doc/sim.pdf},

1351: {\em Canad. J. Math.}, to appear.

1352: 

1353: \bibitem{SPLAG}

1354: J. H. Conway and N. J. A. Sloane,

1355: {\em \htmladdnormallink{Sphere Packing, Lattices and Groups}{http://www.research.att.com/~njas/doc/splag.html}},

1356: Springer-Verlag, NY, 3rd. ed., 1998.

1357: 

1358: \bibitem{Dav57}

1359: Y. David,

1360: On a sequence generated by a sieving process,

1361: {\em Riveon Lematematika} {\bf 11} (1957), 26--31.

1362: 

1363: \bibitem{Del1}

1364: M. Del\'{e}glise,

1365: Computation of large values of $\pi (x)$,

1366: preprint, 1996.

1367: 

1368: \bibitem{DiF96}

1369: P. Di Francesco, \htmladdnormallink{Meander determinants}{http://www.research.att.com/~njas/doc/qc2.pdf},

1370: {\em Commun. Math. Phys.} {\bf 191} (1998),

1371: 543--583;

1372: LANL e-print hep-th/9612026.

1373: 

1374: \bibitem{FGG95}

1375: P. Di Francesco, O. Golinelli and E. Guitter,

1376: \htmladdnormallink{Meanders: a direct enumeration approach}{http://www.research.att.com/~njas/doc/qc2.pdf},

1377: {\em Nucl. Phys.} {\bf B 482} (1996), 497--535;

1378: LANL e-print hep-th/9607039.

1379: 

1380: \bibitem{FGG96}

1381: P. Di Francesco, O. Golinelli and E. Guitter,

1382: \htmladdnormallink{Meander, folding and arch statistics}{http://www.research.att.com/~njas/doc/qc2.pdf},

1383: {\em Math. Comp. Modeling} {\bf 26} (1997), 97--147;

1384: LANL e-print hep-th/950607039.

1385: 

1386: \bibitem{FGG97}

1387: P. Di Francesco, O. Golinelli and E. Guitter,

1388: \htmladdnormallink{Meanders and the Temperley-Lieb algebra}{http://www.research.att.com/~njas/doc/qc2.pdf},

1389: {\em Commun. Math. Phys.} {\bf 186} (1997),

1390: 1--59,

1391: LANL e-print hep-th/9602025.

1392: 

1393: \bibitem{MB13}

1394: R. Donaghey,

1395: Binomial self-inverse sequences and tangent coefficients,

1396: {\em J. Combin. Theory} {\bf A 21} (1976), 155--163.

1397: 

1398: \bibitem{AMM75}

1399: H. Dweighter, Problem E2569,

1400: {\em Amer. Math. Monthly} {\bf 82} (1975), 1010.

1401: Comments by M. R. Garey, D. S. Johnson and

1402: S. Lin, {\em loc. cit.} {\bf 84} (1977), 296.

1403: \bibitem{Ehl1}

1404: H. Ehlich,

1405: Determinantenabschaetzungen f\"{u}r bin\"{a}re Matrizen,

1406: {\em Math. Z.} {\bf 83} (1964), 123--132.

1407: 

1408: \bibitem{EhZ}

1409: H. Ehlich and K. Zeller,

1410: Bin\"{a}re Matrizen, {\em Zeit. Angew. Math. Mech.} {\bf 42} (1962), 20--21.

1411: 

1412: \bibitem{ErJa58}

1413: P. Erd\H{o}s and E. Jabotinsky, On a sequence of integers

1414: generated by a sieving process,

1415: {\em Indag. Math.} {\bf 20} (1958), 115--128.

1416: 

1417: \bibitem{FrKi94}

1418: A. Fraenkel and C. Kimberling,

1419: Generalized Wythoff arrays, shuffles and interspersions,

1420: {\em Discrete Math.} {\bf 126} (1994), 137--149.

1421: 

1422: 

1423: \bibitem{GaP79}

1424: W. H. Gates and C. H. Padadimitriou,

1425: Bounds for sorting by prefix reversal,

1426: {\em Discrete Math.} {\bf 27} (1979), 47--57.

1427: 

1428: \bibitem{MB19}

1429: G. Gati, F. Harary and R. W. Robinson,

1430: Line colored trees with extendable automorphisms,

1431: {\em Acta Math. Scientia} {\bf 2} (1982), 105--110.

1432: 

1433: \bibitem{Gol66}

1434: S. W. Golomb,

1435: Problem 5407,

1436: {\em Amer. Math. Monthly} {\bf 73} (1966), 674;

1437: {\bf 74} (1967), 740--743.

1438: 

1439: \bibitem{Gould}

1440: H. W. Gould,

1441: {\em Combinatorial Identities},

1442: Morgantown, WV, 1972.

1443: 

1444: \bibitem{UPNT}

1445: R. K. Guy,

1446: {\em Unsolved Problems in Number Theory},

1447: Springer-Verlag, NY, 2nd ed., 1994.

1448: 

1449: \bibitem{Guy95}

1450: R. K. Guy,

1451: personal communication, 1995.

1452: 

1453: \bibitem{MB25}

1454: F. Harary and R. W. Robinson,

1455: The number of achiral trees,

1456: {\em J. Reine Angewandte Math.}

1457: {\bf 278} (1975), 322--335.

1458: 

1459: \bibitem{Har98}

1460: M. G. Harris,

1461: \htmladdnormallink{A diagrammatic approach to the meander problem}{http://www.research.att.com/~njas/doc/qc2.pdf}, LANL e-print hep-th/9807193.

1462: 

1463: \bibitem{Her80}

1464: F. Hermann, Ueber das Problem, die Anzahl der isomeren Paraffine der Formel

1465: $C_n H_{2n+2}$ zu bestimmen, {\em Ber. deutsch. chem. Ges.} {\bf 13} (1880), 792.

1466: 

1467: \bibitem{GEB}

1468: D. R. Hofstadter,

1469: {\em G\"{o}del, Fscher, Bach:

1470: An Eternal Golden Braid},

1471: Vintage Books, NY, 1980.

1472: 

1473: \bibitem{Hor}

1474: D. Horsk\'{y}, {\em Pra\v{z}sk\'{y} Orloj}, Panorama, Prague, 1988, pp. 76--78.

1475: 

1476: 

1477: \bibitem{JIS}

1478: {\em Journal of Integer Sequences},

1479: \htmladdnormallink{www.research.att.com/$\sim$njas/sequences/JIS/}{http://www.research.att.com/~njas/seqeunces/JIS/}.

1480: 

1481: \bibitem{KM63}

1482: M. A. Kervaire

1483: and J. W. Milnor,

1484: Groups of homotopy sphere I,

1485: {\em Ann. Math.} {\bf 77} (1963), 504--537.

1486: 

1487: \bibitem{Kim91}

1488: C. Kimberling, Problem 1615, {\em Crux Mathematicorum} {\bf 17} (No. 2, 1991),

1489: 44; {\bf 18} (1993), 82--83.

1490: 

1491: \bibitem{Kim93}

1492: C. Kimberling,

1493: Orderings of the set of all positive Fibonacci sequences,

1494: in G. E. Bergum et al., editors,

1495: {\em Applications of Fibonacci Numbers} {\bf 5} (1993), 405--416.

1496: 

1497: \bibitem{Kim93a}

1498: C. Kimberling, Interspersions and dispersions,

1499: {\em Proc. Amer. Math. Soc.} {\bf 117} (1993), 313--321.

1500: 

1501: \bibitem{Kim94}

1502: C. Kimberling, The first column of an interspersion, {\em Fib. Quart.}

1503: {\bf 32} (1994), 301--314.

1504: 

1505: \bibitem{Kim95}

1506: C. Kimberling,

1507: Numeration systems and fractal sequences, {\em Acta Arithmetica} {\bf 73} (1995), 103--117.

1508: 

1509: \bibitem{Kim95a}

1510: C. Kimberling,

1511: Stolarsky interspersions, {\em Ars Combinatoria} {\bf 39} (1995), 129--138.

1512: 

1513: \bibitem{Kim95b}

1514: C. Kimberling,

1515: The Zeckendorf array equals the Wythoff array, {\em Fib. Quart.}

1516: {\bf 33} (1995), 3--8.

1517: 

1518: \bibitem{Kim97}

1519: C. Kimberling,

1520: Fractal sequences and interspersions, {\em Ars Combinatoria} {\bf 45} (1997), 157.

1521: 

1522: \bibitem{Kimur86}

1523: H. Kimura,

1524: Hadamard matrices of order 28 with automorphism

1525: groups of order two,

1526: {\em J. Combin. Theory} {\bf A 43} (1986),

1527: 98--102

1528: 

1529: \bibitem{Kimur94a}

1530: H. Kimura,

1531: Classification of {H}adamard matrices of order 28 with Hall sets,

1532: {\em Discrete Math.}

1533: {\bf 128} (1994),

1534: 257--268.

1535: 

1536: \bibitem{Kimur94b}

1537: H. Kimura,

1538: Classification of {H}adamard matrices of order 28,

1539: {\em Discrete Math.}

1540: {\bf 133} (1994),

1541: 171--180.

1542: 

1543: \bibitem{KimOh86}

1544: H. Kimura

1545: and H. Ohmori,

1546: Construction of {H}adamard matrices of order 28,

1547: {\em Graphs Combin.} {\bf 2} (1986),

1548: 247--257.

1549: 

1550: \bibitem{KP89}

1551: D. E. Knuth and V. R. Pratt,

1552: personal communication, 1989.

1553: 

1554: \bibitem{Koe}

1555: J. E. Koehler, Folding a strip of stamps,

1556: {\em J. Combin. Theory} {\bf 5} (1968), 135--152.

1557: 

1558: \bibitem{KV1}

1559: T. Kubo and R. Vakil,

1560: On Conway's recursive sequence,

1561: {\em Discrete Math.} {\bf 152} (1996), 225--252.

1562: 

1563: \bibitem{LO82}

1564: J. C. Lagarias and A. M. Odlyzko,

1565: New algorithms for computing $\pi (x)$,

1566: pp. 176--193 of D. V. Chudnovsky et~al., eds.,

1567: {\em Number Theory (New York, 1982)},

1568: Lecture Notes Math. {\bf 1052}, Springer-Verlag, NY.

1569: 

1570: \bibitem{LamTS89}

1571: C. W. H. Lam, L. Thiel, and S. Swiercz,

1572: The non-existence of finite projective planes of order 10,

1573: {\em Canad. J. Math.} {\bf 41} (1989),

1574: 1117--1123.

1575: 

1576: \bibitem{LZ92}

1577: S. K. Lando and A. K. Zvonkin,

1578: Meanders, {\em Selecta Mathematica Sovietica}

1579: {\bf 11} (No. 2, 1992), 117--144.

1580: 

1581: \bibitem{LZ93}

1582: S. K. Lando and A. K. Zvonkin,

1583: Plane and projective meanders,

1584: {\em Theoretical Computer Science}

1585: {\bf 117} (1993), 227--241.

1586: 

1587: \bibitem{Leg}

1588: S. Legendre, personal communication.

1589: 

1590: \bibitem{Lev97}

1591: L. Levine, personal communication, 1997.

1592: 

1593: \bibitem{Luc}

1594: E. Lucas,

1595: {\em Th\'{e}orie des Nombres},

1596: Gauthier-Villars, Paris, 1891.

1597: 

1598: \bibitem{Lun68}

1599: W. F. Lunnon,

1600: A map-folding problem,

1601: {\em Math. Comp.} {\bf 22} (1968), 193--199.

1602: 

1603: \bibitem{MS77}

1604: F. J. MacWilliams

1605: and N. J. A. Sloane,

1606: {\em \htmladdnormallink{The Theory of Error-Correcting Codes}{http://www.research.att.com/~njas/doc/ms77.html}},

1607: North-Holland, Amsterdam, 1977.

1608: 

1609: \bibitem{MacST73}

1610: F. J. MacWilliams, N.~J.~A. Sloane, and J. G. Thompson,

1611: On the existence of a projective plane of order 10,

1612: {\em J. Combin. Theory} {\bf A~14} (1973),

1613: 66--78.

1614: 

1615: \bibitem{CCS1}

1616: C. L. Mallows, Conway's challenge sequence,

1617: {\em Amer. Math. Monthly} {\bf 98} (1991), 5--20; {\bf 99} (1992), 563--564.

1618: 

1619: \bibitem{clm}

1620: C. L. Mallows,

1621: personal communication, 1997.

1622: 

1623: \bibitem{MOS}

1624: C. L. Mallows, A. M. Odlyzko and N. J. A. Sloane,

1625: Upper bounds for modular forms, lattices and codes,

1626: {\em J. Algebra} {\bf 36} (1975), 68--76.

1627: 

1628: \bibitem{MaS}

1629: C. L. Mallows and N. J. A. Sloane,

1630: An upper bound for self-dual codes,

1631: {\em Information and Control}

1632: {\bf 22}

1633: (1973),

1634: 188--200.

1635: 

1636: \bibitem{MR95}

1637: B. D. McKay and E. Rogoyski,

1638: Latin squares of order 10,

1639: {\em Electron. J. Combin.} {\bf 2} (1995), \#N3.

1640: 

1641: \bibitem{bous}

1642: J. Millar, N. J. A. Sloane and N. E. Young,

1643: \htmladdnormallink{A new operation on sequences: the boustrophedon transform}{http://www.research.att.com/~njas/doc/bous.pdf},

1644: {\em J. Combin. Theory}

1645: {\bf A 76} (1996), 44--54.

1646: 

1647: \bibitem{HNT}

1648: D. S. Mitrinovi\'{c},

1649: J. S\'{a}ndor and B. Crstici,

1650: {\em Handbook of Number Theory},

1651: Kluwer, Dordrecht, 1996.

1652: 

1653: \bibitem{Mor80}

1654: D. R. Morrison,

1655: A Stolarsky array of Wythoff pairs, in {\em A Collection of Manuscripts

1656: Related to the Fibonacci Sequence}, Fibonacci Assoc., Santa Clara, CA, 1980, pp. 134--136.

1657: 

1658: \bibitem{Nem}

1659: I. Nemes,

1660: M. Petkov\v{s}ek, H. S. Wilf and D. Zeilberger,

1661: How to do Monthly problems with your computer,

1662: {\em Amer. Math. Monthly} {\bf 104} (1997), 505--519.

1663: 

1664: \bibitem{Od93}

1665: A. M. Odlyzko,

1666: Iterated absolute values of differences of successive

1667: primes,

1668: {\em Math. Comp.} {\bf 61} (1993), 373--380.

1669: 

1670: \bibitem{AB}

1671: M. Petkov\v{s}ek,

1672: H. S. Wilf and D. Zeilberger,

1673: {\em $A=B$},

1674: Peters, Wellesley, MA, 1996.

1675: 

1676: \bibitem{Ph88}

1677: A. Phillips, Simple alternating transit mazes,

1678: preprint, June 1988.

1679: A shorter version appears on the author's home page at

1680: www.math.sunysb.edu/$\sim$tony/mazes.

1681: 

1682: \bibitem{Ph89}

1683: A. Phillips, La topologia dei labirinti,

1684: pp. 57--67 of M. Emmer,

1685: ed.,

1686: {\em L'occhio di Horus:

1687: itinerari nell'imaginario matematico},

1688: Istit. Enciclopedia Ital., Rome, 1989.

1689: 

1690: \bibitem{Poin}

1691: H. Poincar\'{e}, Sur un

1692: th\'{e}or\`{e}me de g\'{e}om\'{e}trie,

1693: {\em Rend. Circ. Mat. Palermo} {\bf 33} (1912),

1694: 375--407 $=$ {\em Oeuvres}, VI, pp. 499--538.

1695: 

1696: \bibitem{PoRa97}

1697: B. Poonen and E. M. Rains,

1698: personal communication, 1997.

1699: 

1700: \bibitem{RS98}

1701: E. M. Rains and N. J. A. Sloane,

1702: \htmladdnormallink{Self-dual codes}{http://www.research.att.com/~njas/doc/self.pdf},

1703: pp. 177--294 of

1704: {\em Handbook of Coding Theory},

1705: ed. V. S. Pless and W. C. Huffman, Elsevier, 1998.

1706: 

1707: \bibitem{RS99}

1708: E. M. Rains and N. J. A. Sloane,

1709: \htmladdnormallink{On Cayley's enumeration of alkanes (or 4-valent trees)}{http://www.research.att.com/~njas/sequences/JIS/cayley.html},

1710: {\em J. Integer Sequences} {\bf 2} (1999), Article 99.1.1.

1711: 

1712: \bibitem{Rec1}

1713: B. Recam\'{a}n, personal communication.

1714: 

1715: \bibitem{RSM}

1716: J. A. Reeds, L. A. Shepp and M. D. McIlroy,

1717: Numerical bounds for the Arnold ``meander'' constant, manuscript, 1991.

1718: 

1719: \bibitem{R1}

1720: J. Riordan,

1721: {\em An Introduction to Combinatorial Analysis},

1722: Wiley, NY, 1958.

1723: 

1724: \bibitem{RS1}

1725: J. Riordan and N. J. A. Sloane,

1726: The enumeration of rooted trees by total height,

1727: {\em J. Austral. Math. Soc.} {\bf 10} (1969), 278--282.

1728: 

1729: \bibitem{Sade}

1730: A. Sade, {\em Sur les Chevauchements des Permutations},

1731: published by the author, Marseille, 1949.

1732: 

1733: \bibitem{Fan}

1734: {\em SeqFan Mailing List},

1735: home page \htmladdnormallink{www.ccr.jussieu.fr/gmpib/seqfan/seqfan.html}{http://www.ccr.jussieu.fr/gmpib/seqfan/seqfan.html}.

1736: 

1737: \bibitem{SVS}

1738: S. D. Servetto, V. A. Vaishampayan and N. J. A. Sloane,

1739: Multiple description lattice vector quantization,

1740: pp. 13--22 of

1741: {\em Proceedings DCC '99: Data Compression Conference (Snowbird, 1999)},

1742: ed. J. A. Storer and M. Cohn,

1743: IEEE Computer Society Press, Los Alamitos, CA, 1999.

1744: 

1745: \bibitem{HIS}

1746: N. J. A. Sloane, {\em A Handbook of Integer Sequences}, Academic Press, NY, 1973.

1747: 

1748: 

1749: \bibitem{OEIS}

1750: N. J. A. Sloane,

1751: {\em \htmladdnormallink{The On-Line Encyclopedia of Integer Sequences}{http://www.research.att.com/~njas/sequences}},

1752: published electronically at www.research.att.com/$\sim$njas/sequences.

1753: 

1754: 

1755: \bibitem{EIS}

1756: N. J. A. Sloane and S. Plouffe,

1757: {\em \htmladdnormallink{The Encyclopedia of Integer Sequences}{http://www.research.att.com/~njas/sequences/book.html}},

1758: Academic Press, 1995.

1759: 

1760: 

1761: 

1762: \bibitem{smar}

1763: F. Smarandache,

1764: {\em Only Problems, Not Solutions!},

1765: Xiquan. Publ.,

1766: Phoenix and Chicago, 1993.

1767: 

1768: \bibitem{WDS}

1769: W. D. Smith, {\em Studies in Computational Geometry Motivated

1770: by Mesh Generation},

1771: Ph.D. Dissertation,

1772: Dept. Appl. Math.,

1773: Princeton Univ., Sept. 1988.

1774: 

1775: \bibitem{Sto76}

1776: K. B. Stolarsky, Beatty sequences, continued fractions,

1777: and certain shift operators, {\em Canad. Math. Bull.} {\bf 19} (1976), 472--482.

1778: 

1779: \bibitem{Sto77}

1780: K. B. Stolarsky,

1781: A set of generalized Fibonacci sequences such

1782: that each natural number belongs to exactly one, {\em Fib. Quart.} {\bf 15} (1977), 224.

1783: 

1784: \bibitem{Touch50}

1785: J. Touchard,

1786: Contributions \`{a} l'\'{e}tude du probl\`{e}me des timbres postes,

1787: {\em Canad. J. Math.} {\bf 2} (1950), 385--398.

1788: 

1789: \bibitem{Wat29}

1790: G. N. Watson,

1791: Theorems stated by Ramanujan (V):

1792: approximations connected with $e^x$, {\em Proc. London

1793: Math. Soc.} {\bf 29} (1929), 293--308.

1794: 

1795: \bibitem{Wie}

1796: D. H. Wiedemann, personal communication.

1797: 

1798: \bibitem{Wilks}

1799: A. R. Wilks, personal communication, 1997.

1800: 

1801: \bibitem{RGW}

1802: R. G. Wilson, V, personal communication.

1803: 

1804: \bibitem{Zim1}

1805: P. Zimmermann, personal communication, 1988.

1806: 

1807: 

1808: \end{thebibliography} 

1809: \end{document} 

1810: