nlin0503069/iso.tex
1: \documentclass[12pt]{article}
2: 
3: \usepackage{graphicx}
4: \usepackage{amsmath, amssymb}
5: 
6: \def\calp{{\mathcal P}}
7: \def\fq{\mathbb{F}_{q}}
8: \def\fd{\mathbb{F}_{2}}
9: \def\nm{\mathbb{N}}
10: 
11: 
12: %%%%%%%%%%%%%%%%%%%%%%%%%% Format de la page %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
13: \begin{document}
14: 
15: \title{\bf Finite geometries and diffractive orbits in isospectral billiards}
16: \author{Olivier Giraud\\ 
17: Laboratoire de Physique Th\'eorique,\\
18:  UMR 5152 du CNRS, \\
19: Universit\'e Paul Sabatier,\\
20:  31062 Toulouse Cedex 4,\\
21: France\\}
22: \maketitle
23: 
24: 
25: \begin{abstract}
26: Several examples of pairs of isospectral planar domains have been produced
27: in the two-dimensional Euclidean space by various methods. 
28: We show that all these examples rely on the symmetry between 
29: points and blocks in finite projective spaces; from the properties 
30: of these spaces, one can derive a relation between Green functions 
31: as well as a relation between diffractive orbits in isospectral billiards.
32: \end{abstract}
33: 
34: 
35: 
36: 
37: 
38: %=================================================================
39: %====================Introduction=================================
40: %=================================================================
41: It is almost forty years now that Mark  Ka{\v c} addressed his 
42: famous question ``Can one hear the shape of a drum'' \cite{Kac66}. 
43: The problem was to know whether there exists
44:  isospectral domains, that is non-isometric bounded regions of 
45: space for which the sets $\{E_n, n\in\nm\}$ of solutions of the 
46: stationary Schr\"odinger equation $(\Delta+E)\Psi=0$, with some 
47: specified boundary conditions, would coincide.
48: Instances of isospectral pairs have finally been found for Riemannian manifolds
49:  \cite{Be89, Br88}, and more recently for two-dimensional Euclidean 
50: connected compact domains (we will call such planar domains ''billiards'') 
51: \cite{GorWebWol92}.
52: The proof of isospectrality was based on Sunada's theorem, which allows to 
53: construct isospectral pairs from groups related by a so-called 
54: ''almost conjugate'' property \cite{Sun85}. 
55: The two billiards given in \cite{GorWebWol92} to illustrate the existence of
56: isospectral billiards are represented in \ref{billards}a: they are made 
57: of 7 copies of a base tile (here a right isosceles triangle) unfolded with
58: respect to its edges in two different ways. 
59: It turns out that in this example, isospectrality can be 
60: proved directly by giving an explicit linear map between the two 
61: domains \cite{Cha95}. Eigenfunctions in one billiard can be expressed as a
62: linear superposition of eigenfunctions of the other billiard, in such
63: a way that boundary conditions are fulfilled; isospectrality is
64: ensured by linearity of Schr\"odinger equation. This 
65: ''transplantation'' method allowed to generalize the example of 
66: \cite{GorWebWol92} : any triangle can replace 
67: the base triangle of \ref{billards}a, what matters is the way 
68: the initial base tile is unfolded (see \ref{billards}b).
69: \begin{figure}[ht]
70: \begin{center}
71: \includegraphics[width=0.68\linewidth]{fig1.eps}
72: \end{center}
73: \caption{a) Two isospectral billiards with a triangular base shape.
74: b) A pair of isospectral billiards constructed from the same unfolding rules 
75: as a).}
76: \label{billards}
77: \end{figure}
78: Further examples of isospectral billiards were obtained by applying 
79: Sunada's theorem to reflexion groups in the 
80: hyperbolic upper-half plane \cite{BusConDoySem94}: examples of
81: billiards made of 7, 13, 15 or 21 triangular tiles were given.
82: Later on, it was shown that the transplantation 
83: method produces exactly the same examples and provides 
84: a proof for iso-length-spectrality of these billiards
85: (i.e. the set of lengths of periodic orbits is identical) \cite{OkaShu01}.
86: A similar transplantation
87: technique, together with results on Green functions of polygonal billiards,
88: was applied to show that there is no such equality of the lengths for
89: diffractive orbits (that is, the lines joining two vertices); a
90: more involved relation was found \cite{Gir04}.\\
91: 
92: The purpose of this letter is to show that all known examples of
93: isospectral pairs, given by \cite{BusConDoySem94} or \cite{OkaShu01}, 
94: can be obtained by considering finite projective spaces
95: (FPS): the transplantation map between two isospectral billiards 
96: can be taken to be the incidence matrix of a FPS. This construction 
97: from FPSs indicates that the deep origin of isospectrality in 
98: Euclidean spaces is point-block duality. It also leads to a
99: relation between Green functions, and to a correspondence 
100: between diffractive orbits in isospectral billiards.
101: 
102: 
103: 
104: %=================================================================
105: %================    Isospectral billiards     ===================
106: %=================================================================
107: \subsection*{Isospectral billiards.} 
108: All known isospectral billiards can be obtained by unfolding 
109: triangle-shaped tiles \cite{BusConDoySem94, OkaShu01}.
110: The way the tiles are unfolded can be specified by three 
111: permutation matrices $M^{(\mu)}$, $1\leq \mu\leq 3$, associated to the 
112: three sides of the triangle: $M^{(\mu)}_{ij}=1$ if tiles $i$ and $j$
113: are glued by their side $\mu$ (and $M^{(\mu)}_{ii}=1$ if the side $\mu$
114: of tile $i$ is the boundary of the billiard), 
115: and 0 otherwise \cite{OkaShu01, Tha04, Gir04}.
116: Following \cite{OkaShu01}, one can sum up the action of the $M^{(\mu)}$
117: in a graph with coloured edges: each copy of the base 
118: tile is associated to a vertex, and vertices $i$ and $j$, $i\neq j$,
119: are linked by an edge of colour $\mu$ if and only if $M^{(\mu)}_{ij}=1$ 
120: (see \ref{graphs}).
121: \begin{figure}[ht]
122: \begin{center}
123: \includegraphics[width=0.66\linewidth]{fig2.eps}
124: \end{center}
125: \caption{The graphs corresponding to a pair of isospectral billiards: 
126: if we label the sides of the triangle by $\mu=1,2,3$, the unfolding rule
127: by symmetry with respect to side $\mu$ can be represented by edges made 
128: of $\mu$ braids in the graph. From a given pair of graphs, one can construct
129: infinitely many pairs of isospectral billiards by applying the unfolding
130: rules to any triangle. Note that a different labeling of the tiles would just
131: induce a permutation of the labelings of points and blocks in the Fano plane.}
132: \label{graphs}
133: \end{figure}
134: In the same way, in the second member of the pair, the tiles are 
135: unfolded according to
136: permutation matrices $N^{(\mu)}$, $1\leq \mu\leq 3$. Two billiards are
137: said to be transplantable if there exists an invertible matrix $T$
138: (the {\it transplantation matrix})
139: such that $\forall\mu\ \ T M^{(\mu)}= N^{(\mu)} T$. One can show that
140: transplantability implies isospectrality (if the matrix $T$ is not 
141: merely a permutation matrix, in which case the two domains would just
142: have the same shape). The underlying 
143: idea is that if $\psi^{(1)}$ is
144: an eigenfunction of the first billiard and $\psi^{(1)}_i$ its restriction
145: to triangle $i$, then one can build an eigenfunction  $\psi^{(2)}$ 
146: of the second billiard by taking $\psi^{(2)}_i=\sum_j T_{ij}\psi^{(1)}_j$. 
147: Obviously $\psi^{(2)}$
148: verifies Schr\"odinger equation; it can be checked from the
149: commutation relations that the function is smooth at all edges 
150: of the triangles, and that boundary conditions at the boundary of 
151: the billiard are fulfilled \cite{OkaShu01}.\\
152: 
153: Suppose we want to construct a pair of isospectral billiards, starting
154: from any polygonal base shape. Our idea is to start from the 
155: transplantation matrix, and choose it in such a way that the existence
156: of commutation relations $T M^{(\mu)}= N^{(\mu)} T$ for some 
157: permutation matrices $M^{(\mu)}, N^{(\mu)}$ will be known {\it a priori}.
158: As we will see, this is the case if $T$ is taken to be
159: the incidence matrix of a FPS; the matrices $M^{(\mu)}$ and
160: $N^{(\mu)}$ are then permutations on the points and the hyperplanes
161: of the FPS.
162: 
163: 
164: 
165: 
166: 
167: %=================================================================
168: %================    Finite projective spaces  ===================
169: %=================================================================
170: \subsection*{Finite projective spaces.} 
171: For $n\geq 2$ and $q=p^h$  a power of a prime number, consider 
172: the $(n+1)$-dimensional vector space $\fq^{n+1}$, where
173: $\fq$ is the finite field of order $q$.
174: The {\it finite projective space} $PG(n,q)$ of {\it dimension} $n$
175: and {\it order} $q$ is the set of subspaces of $\fq^{n+1}$: 
176: the points of $PG(n,q)$ are the 1-dimensional subspaces of $\fq^{n+1}$, 
177: the lines of $PG(n,q)$ are the 2-dimensional subspaces of $\fq^{n+1}$,
178: and more generally $(d+1)$-dimensional subspaces of $\fq^{n+1}$ are 
179: called {\it $d$-spaces} of $PG(n,q)$; the $(n-1)$-spaces of $PG(n,q)$ 
180: are called hyperplanes or {\it blocks}. 
181: A $d$-space of $PG(n,q)$ contains $(q^{d+1}-1)/(q-1)$ points. In particular,
182: $PG(n,q)$ has $(q^{n+1}-1)/(q-1)$ points. It also has 
183: $(q^{n+1}-1)/(q-1)$ blocks \cite{Hir79}.
184: As an example, \ref{fano} shows the finite projective plane (FPP) 
185: of order $q=2$, or Fano plane, $PG(2,2)$. 
186: \begin{figure}[ht]
187: \begin{center}
188: \includegraphics[width=0.58\linewidth]{fig3.eps}
189: \caption{The Fano plane $PG(2,2)$ and its corresponding incidence
190: matrix $T$. The Fano plane has $(q^3-1)/(q-1)=7$ points 
191: and 7 lines; each line contains $q+1=3$ points and each point belongs 
192: to 3 lines. Any pair of points belongs to one and only one line.}
193: \label{fano}
194: \end{center}
195: \end{figure}
196: 
197: A $(N,k, \lambda)-${\it symmetric balanced incomplete block design} 
198: (SBIBD) is a set of
199: $N$ points, belonging to $N$ subsets (or {\it blocks});
200: each block contains $k$ points, in such a way that any two points belong 
201: to exactly $\lambda$ blocks, and each point is contained in 
202: $k$ different blocks \cite{DinSti92}. One can show that
203: $PG(n,q)$ is a $(N,k, \lambda)$-SBIBD with $N=(q^{n+1}-1)/(q-1)$,
204: $k=(q^{n}-1)/(q-1)$ and $\lambda=(q^{n-1}-1)/(q-1))$. 
205: For example, the Fano plane is a $(7,3,1)-$SBIBD.
206: The points and the blocks can be labeled from $0$ to $N-1$.
207: For any $(N,k, \lambda)-$SBIBD one can define an $N\times N$
208: {\it incidence matrix} 
209: $T$ describing to which block each point belongs. The entries $T_{ij}$ 
210: of the matrix are $T_{ij}=1$ if point $j$ belongs to line $i$, $0$ 
211: otherwise. The matrix $T$ verifies the relation
212: $T T^{t}=\lambda J+(N-k)\lambda/(k-1)I$, 
213: where $J$ is the $N\times N$ matrix with all entries equal to 1 
214: and $I$ the $N\times N$ identity matrix \cite{DinSti92}. 
215: In particular, the incidence matrix of $PG(n,q)$ verifies
216: \begin{equation}
217: \label{tt}
218: T T^{t}=\lambda J+(k-\lambda)I
219: \end{equation}
220: with $k$ and $\lambda$ as given above. For example, the incidence matrix 
221: of the Fano plane given in 
222: \ref{fano} corresponds to a labeling of the lines such that line 0
223: contains points 0,1,3, and line 1 contains points 1,2,4, etc.
224: 
225: A {\it collineation} of a FPS is a 
226: bijection that preserves incidence, that
227: is a permutation of the points that takes $d$-spaces to 
228: $d$-spaces (in particular, it takes blocks to blocks).
229: Any permutation $\sigma$ on the points
230: can be written as a $N\times N$ {\it permutation matrix} $M$ defined by 
231: $M_{i\sigma(i)}=1$ and the other entries equal to zero. If $M$ is a 
232: permutation matrix associated to a collineation, then there exists a 
233: permutation matrix $N$ such that
234: \begin{equation}
235: \label{commutation}
236: TM=NT.
237: \end{equation}
238: In words, (\ref{commutation}) means that permuting the 
239: columns of $T$ (i.e. the blocks of the
240: space) under $M$ is equivalent to permuting the rows of $T$ 
241: (i.e. the points of the space) under $N$.
242: The commutation relation (\ref{commutation}) is a 
243: related to an important feature of projective geometry, the so-called
244: ''principle of duality'' \cite{Hir79}. 
245: This principle states that for any theorem which is 
246: true in a FPS, the dual theorem obtained by exchanging 
247: the words ''point'' and ''block'' is also true. As we will see now, 
248: this symmetry between points and blocks in FPSs is the central reason 
249: which accounts for known pairs of isospectral billiards.\\
250: 
251: Let us consider a FPS $\calp$ with incidence matrix $T$.
252: To each block in $\calp$ we associate a tile in the first billiard, 
253: and to each point in $\calp$ we associate a tile in the second billiard.
254: If we choose permutations $M^{(\mu)}$ in the collineation 
255: group of $\calp$, then the commutation relation (\ref{commutation})
256: will ensure that there exist permutations $N^{(\mu)}$ verifying
257: $TM^{(\mu)}=N^{(\mu)}T$. These commutation relations imply 
258: transplantability, and thus isospectrality, of the billiards 
259: constructed from the graphs corresponding to $M^{(\mu)}$ and $N^{(\mu)}$.
260: 
261: If the base tile has $r$ sides, we need to choose $r$ 
262: elements $M^{(\mu)}$, $1\leq\mu\leq r$, in the 
263: collineation group of $\calp$. This choice is 
264: constrained by several factors. Since
265: $M^{(\mu)}$ represents the reflexion of a tile with respect to one of
266: its sides, it has to be of order 2 (i.e. an involution). 
267: In order that the billiards be connected, no point should be left out by 
268: the matrices $M^{(\mu)}$ (in other words, the graph associated to the
269: matrices $M^{(\mu)}$ should be connected). 
270: Finally, if we want the base tile to be of any shape, there should 
271: be no loop in the graph. We now need to characterize 
272: collineations of order 2.
273: 
274: 
275: 
276: 
277: 
278: %=================================================================
279: %===== Collineations of  FPSs  ===============
280: %=================================================================
281: \subsection*{Collineations of finite projective spaces.} 
282: Let $q=p^h$ be a power of a prime number.
283: Each point $P$ of $PG(n,q)$ is a 1-dimensional subspace of $\fq^{n+1}$, 
284: spanned by some vector $v$. We write $P=P(v)$.
285: 
286: An {\it automorphism} is a bijection of the points $P(v_i)$
287: of $PG(n,q)$ obtained 
288: by the action of an automorphism of $\fq$ on the coordinates 
289: of  the $v_i$. If $q=p^h$, the automorphisms of $\fq$ are
290: $t\mapsto t^{p^i}$, $0\leq i \leq h-1$.
291: 
292: A {\it projectivity} is a bijection of the points $P(v_i)$ 
293: of $PG(n,q)$ obtained by the action of a linear map $L$ on the $v_i$.
294: There are $q-1$ matrices $t L$, with $t\in\fq\setminus\{0\}$ and 
295: $L\in GL_{n+1}(\fq)$ (the group of $(n+1)\times (n+1)$ invertible
296: matrices with coefficients in $\fq$), yielding the same projectivity
297: $P(v_i)\mapsto P(L v_i)$.
298: %It can be proved that there always exist cyclic projectivities, that
299: %is projectivities which permute the $N$ points of $PG(n,q)$ in a single
300: %cycle. If $L$ is a linear map corresponding to such a projectivity, then
301: %one can choose a labeling of the points such that $P_i=P(L^i v)$, where
302: %$v=(1,0,...,0)\in \fq^{n+1}$. A line in $PG(n,q)$ is then a set of $q+1$
303: %points whose corresponding vectors are coplanar, and more
304: %generally a $d$-space in $PG(n,q)$ is a set of points whose
305: % corresponding vectors belong to a $(d+1)$-dimensional subspace of
306: %$\fq^{n+1}$. This way of labeling the points from $0$ to $N-1$ has the
307: %property that if $(d_0, d_1,..., d_{B-1})$ are the integers representing 
308: %the $B=(q^n-1)/(q-1)$ points of a block, then the $N$ blocks are the sets
309: %$(d_0+k, d_1+k,..., d_{B-1}+k)$ mod $N$, for $1\leq k\leq N-1$ \cite{Hir79}.
310: %For instance, points in the Fano plane $PG(2,2)$ can be labeled in such a
311: %way that its lines are given by $(k, k+1, k+3)$ mod $7$, $0\leq k\leq 6$.
312: %(see \ref{fano}).\\
313: 
314: The {\it Fundamental theorem of projective geometry } states
315: that any collineation of $PG(n,q)$ can be written as
316: the composition of a projectivity by an automorphism \cite{Hir79}.
317: The converse is obviously true since projectivities and automorphisms 
318: are collineations. Therefore the set of all collineations is obtained
319: by taking the composition of all the non-singular $(n+1)\times(n+1)$
320: matrices with coefficients in $\fq$ by all the $h$ automorphisms of $\fq$.
321: 
322: The collineation group of $PG(n, q)$ has 
323: $[h\prod_{k=0}^{n}(q^{n+1}-q^k)]/(q-1)$ elements, among which we only
324: want to consider elements of order 2. In the case of FPPs ($n=2$), 
325: there are various known properties characterizing
326: collineations of order 2. 
327: A {\it central collineation}, or {\it perspectivity}, is a collineation fixing
328: each line through a point $C$ (called the centre). By ''fixing'' we
329: mean that the line is invariant but the points can be permuted within the
330: line. One can show that the fixed points of a non-identical perspectivity
331: are the centre itself
332: and all points on a line (called the axis), while the fixed lines are
333: the axis and all lines through the centre. 
334: If the centre lies off the axis a perspectivity is called a 
335: {\it homology} (and has $q+2$ fixed points), whereas if the centre 
336: lies on the axis it is called an {\it elation} (and has $q+1$ fixed points).
337: Perspectivities in dimension $n=2$ have following properties \cite{Bon04}:
338: 
339: {\scshape Proposition 1.} A perspectivity of order two 
340: of a FPP of order $q$ is an elation or a 
341: homology according to whether $q$ is even or odd.
342: 
343: {\scshape Proposition 2.} A collineation of order two of a FPP
344:  of order $q$ is a perspectivity if $q$ 
345: is not a square; it is a collineation fixing all points and lines 
346: in a subplane if $q$ is a square (a subplane is a subset of points 
347: having all the properties of a FPP).
348: 
349: When the order of the FPP is a square, there is 
350: the following result \cite{Hir79}:
351: 
352: {\scshape Proposition 3.} $PG(2, q^2)$ can be partitioned into $q^2-q+1$
353: subplanes $PG(2, q)$.
354: 
355: 
356: 
357: 
358: 
359: 
360: 
361: %=================================================================
362: %=====       Generating isospectral pairs          ===============
363: %=================================================================
364: \subsection*{Generating isospectral pairs.} 
365: Let us assume we are looking for a pair of isospectral billiards with
366: $N=(q^3-1)/(q-1)$ copies of a base tile having the 
367: shape of a $r$-gon, $r\geq 3$. 
368: We need to find $r$ collineations of order 2 such that the associated graph is
369: connected and without loop. Such a graph connects $N$ vertices and thus
370:  requires $N-1$ edges. From propositions 1-3, we can deduce the 
371: number $s$ of fixed points of a collineation of order 2 for any FPP. 
372: Since a collineation is a permutation, it has a cycle decomposition as
373: a product of transpositions. For permutations of order 2 with $s$ fixed 
374: points, there are $(N-s)/2$ independent transpositions in this decomposition.
375: Each transposition is represented by an edge in the graph. As a consequence,
376:  $q$, $r$ and $s$ have to fulfill the following condition: 
377: $r(q^2+q+1-s)/2=q^2+q$. Let us examine the various cases.
378: 
379: {\it If $q$ is even and not a square},
380: propositions 1 and 2 imply that 
381: any collineation of order 2 is an elation and therefore has $q+1$
382: fixed points. Therefore, $q$ and $r$ are constrained by the relation
383: $r q^2/2=q^2+q$.
384: The only integer solution with $r\geq 3$ and $q\geq 2$ is $(r=3, q=2)$. 
385: These isospectral billiards  correspond to $PG(2,2)$ and will be made
386: of $N=7$ copies of a base triangle.
387: 
388: {\it If $q$ is odd and not a square},
389: propositions 1 and 2 imply that 
390: any collineation of order 2 is a homology and therefore has $q+2$
391: fixed points. Therefore, $q$ and $r$ are constrained by the relation
392: $r (q^2-1)/2=q^2+q$.
393: The only integer solution with $r\geq 3$ and $q\geq 2$ is $(r=3, q=3)$. 
394: These isospectral billiards correspond to $PG(2,3)$ and will be made
395: of $N=13$ copies of a base triangle.
396: 
397: {\it If $q=p^2$ is a square}, 
398: propositions 2 and 3 imply that 
399: any collineation of order 2 fixes all points in a subplane $PG(2,p)$
400: and therefore has $p^2+p+1$ fixed points. Therefore, $p$ and $r$ are 
401: constrained by the relation $r (p^4-p)/2=p^4+p^2$.
402: There is no integer solution with $r\geq 3$ and $q\geq 2$. However,
403: one can look for isospectral billiards with loops: this will require
404: the base tile to have a shape such that the loop  does not make the
405: copies of the tiles come on top of each other when unfolded. 
406: If we tolerate one loop
407: in the graph describing the isospectral billiards, then there are $N$ 
408: edges in the graph instead of $N-1$ and the equation for $p$ and $r$
409: becomes $r (p^4-p)/2=p^4+p^2+1$, which has the only integer solution
410: $(r=3, p=2)$. These isospectral billiards correspond to $PG(2,4)$
411: and will be made of $N=21$ copies of a base triangle.
412: 
413: We can now generate all possible pairs of isospectral billiards 
414: whose transplantation matrix is the incidence matrix of $PG(2,q)$, 
415: with  $r$ and $q$ restricted by the previous analysis. All pairs 
416: must have a triangular base shape ($r=3$).
417: $PG(2, 2)$ provides 3 pairs (made of 7 tiles), 
418: $PG(2, 3)$ provides 9 pairs (made of 13 tiles), 
419: $PG(2, 4)$ provides 1 pair (made of 21 tiles).
420: It turns out that the pairs obtained here are exactly those obtained by 
421: \cite{BusConDoySem94} and \cite{OkaShu01} by other methods.
422: Let us now consider spaces $PG(n,q)$ of higher dimensions. The smallest
423: one is $PG(3,2)$, which contains 15 points.
424: Since the base field for $PG(3,2)$ is $\fd$, the Fundamental 
425: Theorem of projective geometry states that the collineation group of 
426: $PG(3,2)$ is essentially the group $GL_4(\fd)$ of 
427: $4\times 4$ non-singular matrices with coefficients in $\fd$. 
428: Generating all possible graphs from the 316
429: elements of order 2 in $GL_4(\fd)$, we obtain four pairs of 
430: isospectral billiards with 15 triangular tiles, which completes
431: the list of all pairs found in \cite{BusConDoySem94} and \cite{OkaShu01}.\\
432: 
433: Our method explicitly gives the transplantation matrix $T$ for all these
434: pairs: each one is the incidence matrix of a FPS. 
435: The transplantation matrix explicitly provides the mapping between
436: eigenfunctions of both billiards. The inverse mapping is given by 
437: $T^{-1}=(1/q^{n-1})(T^{t}-(\lambda/k)J)$.
438: For all pairs, isospectrality can therefore be explained 
439: by the symmetry between points and blocks in FPSs.
440: We do not know if it is possible to find isospectral billiards for 
441: which isospectrality would not rely on this symmetry.
442: 
443: Our construction furthermore allows to generalize the results we obtained 
444: in \cite{Gir04}, where a relation between the Green functions of the 
445: billiards in an isospectral pair was derived. 
446: A similar relation can be found for all other pairs
447: obtained by point-block duality. Let $M^{(\mu)}$ and $N^{(\mu)}$
448: be the matrices describing the gluings of the tiles.
449: To any path $p$ going from a point to another on the first billiard, 
450: one can associate the sequence $(\mu_1,...,\mu_n)$, $1\leq\mu_i\leq 3$,
451: of edges of the triangle hit by the path. The matrix $M=\prod M^{(\mu_i)}$
452: is such that $M_{ij}=1$ if path $p$ can be drawn from $i$ to $j$.
453: If $N=T^{-1}MT$, relations (\ref{tt}) and (\ref{commutation}) imply  
454: $\sum_{k,l}T_{ik}T_{jl}M_{kl}=\lambda+(k-\lambda)N_{ij}$. 
455: Since Green functions can be written as a sum
456: over all paths, the relation between the Green functions $G^{(2)}(a,i;b,j)$ 
457: and $G^{(1)}(a,i';b,j')$ is
458: $\sum_{i',j'}T_{i,i'}T_{j,j'}G^{(1)}(a,i';b,j')=
459: (k-\lambda)G^{(2)}(a,i;b,j)+\lambda G^{t}(a;b)$, where $G^{t}$ is the 
460: Green function of the triangle, and a point $(a,i)$ is specified 
461: by a tile number $i$ and its position $a$ inside the tile
462: (see \cite{Gir04} for further detail).  
463: More precisely, the term $\sum_{k,l}T_{ik}T_{jl}M_{kl}$
464:  can be interpreted as the number of pairs of
465: tiles $(i,j)$ in the first billiard such that a path identical to
466: $p$ can go from $i$ to $j$.  
467: A given diffractive orbit going from $i$ to $j$ in the second 
468: billiard corresponds to a matrix $N$ such that $N_{ij}=1$: 
469: it is therefore constructed from a superposition
470: of $k=(q^n-1)/(q-1)$ identical diffractive orbits in the first billiard.
471: (Note that these results correspond to Neumann boundary conditions. It is
472: easy to obtain similar relations for Dirichlet
473: boundary conditions by
474: conjugating all matrices with a diagonal matrix $D$ with entries 
475: $D_{ii}=\pm 1$ according to whether tile $i$ is like the initial tile 
476: or like its mirror inverse.)
477: 
478: 
479: 
480: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
481: %                BIBLIOGRAPHICAL REFERENCES
482: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
483: 
484: 
485: \begin{thebibliography}{50}
486: 
487: \bibitem{Kac66}
488: Ka{\v c} M 1966 {\it Am. Math. Monthly} {\bf 73}  1
489: 
490: \bibitem{Be89}
491: B\'erard P 1989 {\it Ast\'erisque} {\bf 177-178}  127
492: 
493: \bibitem{Br88}
494: Brooks R 1988 {\it Am. Math. Mon.} {\bf 95}  823
495: 
496: \bibitem{GorWebWol92}
497: Gordon C, Webb D and Wolpert S 1992 {\it Invent. math.} {\bf 110}  1
498: 
499: \bibitem{Sun85}
500: Sunada T 1985 {\it Ann. Math.} {\bf 121} 169
501: 
502: \bibitem{Cha95}
503: Chapman S J 1995 {\it Am. Math. Monthly} {\bf 102}  124
504: 
505: \bibitem{Gor86}
506: Gordon C S 1986 {\it Contemp. Math.} {\bf 51}  63
507: 
508: \bibitem{Hir79}
509: Hirschfeld J W P 1979 {\it Projective geometries over finite fields}, 
510: Clarendon Press - Oxford
511: 
512: \bibitem{BusConDoySem94}
513: Buser P, Conway J, Doyle P and Semmler K-D 1994 
514: {\it Int. Math. Res. Not.} {\bf 9} 391
515: 
516: \bibitem{OkaShu01} 
517: Okada Y and Shudo A 2001 {\it J. Phys. A: Math. Gen.} {\bf 34} 5911
518: 
519: \bibitem{Tha04} Thain A 2004 {\it Eur. J. Phys.} {\bf 25} 633
520: 
521: \bibitem{Gir04}
522: Giraud O 2004 {\it J. Phys. A: Math. Gen.} {\bf 37} 2751
523: 
524: \bibitem{DinSti92} Dinitz J H and Stinson D R 1992, in {\it Contemporary 
525: Design Theory: A Collection of Surveys} (Ed. J. H. Dinitz and D. R. Stinson). 
526: New York: Wiley 1-12
527: 
528: \bibitem{Bon04}
529: Bonisoli A 2000 {\it Notes on Finite Geometries and its applications},
530: University of Ghent
531: 
532: 
533: \bibitem{Hug57}
534: Hughes D R 1957 {\it Ill. J. Math.} {\bf 1} 545
535: 
536: \end{thebibliography}
537: 
538: 
539: 
540: \end{document}
541: