ON SQUARE-INTEGRABLE REPRESENTATIONS OF CLASSICAL p-adic GROUPS II

Μέγεθος: px
Εμφάνιση ξεκινά από τη σελίδα:

Download "ON SQUARE-INTEGRABLE REPRESENTATIONS OF CLASSICAL p-adic GROUPS II"

Transcript

1 REPRESENTATION THEORY An Electronic Journal of the American Mathematical Society Volume 4, Pages (February 23, 2000) S (00) ON SQUARE-INTEGRABLE REPRESENTATIONS OF CLASSICAL p-adic GROUPS II CHRIS JANTZEN Abstract. In this paper, we continue our study of non-supercuspidal discrete series for the classical groups Sp(2n, F ), SO(2n + 1,F), where F is p-adic. 1. Introduction This paper is a continuation of [Jan4]. Suppose ρ is an irreducible, unitary supercuspidal representation of GL n (F )and σ an irreducible, supercuspidal representation of S r (F )=Sp 2r (F )orso 2r+1 (F ). Then, if ρ = ρ (self-contragredient), we have Ind( detn x ρ σ) is irreducible for all x R. Otherwise, there is a unique α 0 such that Ind( det n x ρ σ) is reducible for x = ±α and irreducible for all x R \{±α} (cf. [Sil2], [Tad5]). Assuming certain conjectures, Mœglin [Mœ2] and Zhang [Zh] have shown that α 1 2 Z (also, cf. [M-R], [Re], [Sha1], [Sha2]). In [Jan4], we showed that the problem of classifying non-supercuspidal discrete series could be reduced to classifying discrete series with supercuspidal support in S((ρ, α); σ) ={ν x ρ, ν x ρ} x α+z {σ} (though most of this result is from [Jan2] and [Tad4]). In [Jan4], to an irreducible representation π supported on S((ρ, α); σ), we associated an element χ 0 (π) in the minimal Jacquet module for π (minimal meaning with respect to the smallest parabolic subgroup admitting a nonzero Jacquet module for π). We note that there is an order on the components of the (semisimplified) minimal Jacquet module of π such that χ 0 (π) is minimal with respect to. From χ 0 (π), we can read off a representation δ 0 (π) havingtheform δ 0 (π) =δ([ν b1 ρ, ν a1 ρ]) δ([ν b k ρ]) σ, where δ([ν bi ρ, ν ai ρ]) denotes the generalized Steinberg (for a general linear group) which has ν ai ρ ν ai 1 ρ ν bi ρ as its (minimal) Jacquet module. We have π Ind(δ 0 (π)). Further, one can determine whether π is square-integrable just from δ 0 (π). Our goal is then to constrain the candidates for δ 0 for square-integrable representations, thereby constraining where one needs to look for square-integrable representations. Of course, the ultimate goal is exhaustion: if the constraints are strong enough, every possible δ 0 listed will actually occur as δ 0 (π) for some discrete series π. That is the case for the theorem below when ρ =1 F, σ =1 S0(F ) (cf. [Mœ1]); we expect it holds in general. Received by the editors July 28, 1999 and, in revised form, October 18, Mathematics Subject Classification. Primary 22E c 2000 American Mathematical Society

2 128 CHRIS JANTZEN In a sense, our point-of-view may be regarded as an extension of that of [B-Z] and [Zel]. Their descriptions of representations of general linear groups are given in combinatorial terms, with the combinatorial descriptions essentially independent of which particular supercuspidal representations appear in the support. Here, we work with discrete series for symplectic and odd-orthogonal groups in an analogous setting; while our results are not independent of the ρ, σ which appear, they depend on ρ, σ only insofar as ρ, σ determine α. (Roughly speaking, α = 1 for the GL counterpart, so this issue does not arise.) Thus, from our point-of-view, α is essentially treated as input data, with discrete series characterizations as the output. The main result in this paper is the following: Theorem 1.1. Suppose ρ is an irreducible, unitary, supercuspidal representation of GL n (F ) and σ an irreducible, supercuspidal representation of S r (F ). Also, suppose Ind( det n x ρ σ) is reducible for x = ±α and irreducible for x R \{±α}, where α 0 has α 1 2Z.Ifπis an irreducible, square-integrable representation supported on S((ρ, α); σ) with δ 0 (π) =δ([ν b1 ρ, ν a1 ρ]) δ([ν b k ρ]) σ, the following conditions must be satisfied: 1. basic conditions: (a) a k a k 1 a 1 > 0. (b) a i >b i for all i. (c) There is a β with α +1 β>0 such that each of { β, β 1,..., α} appears exactly once among b 1,b 2,...,b k and there are no other negative b i s. 2. a i β for all i; b i β 1 for all b i a 1,a 2,...,a k,b 1,b 2,...,b k are all distinct. 4. We do not have a i >a j >b i >b j for any i>j. Remarks Coupled with Theorem of [Jan4], Theorem 1.1 above applies to any irreducible, square-integrable representation if we assume the half-integrality hypothesis in Theorem 1.1 holds in general. That is, for any such (ρ, σ), the value of x R, x 0 which makes Ind( det x ρ σ) reducible (if any) lies in 1 2Z. (The half-integrality hypothesis follows from [Mœ2] or [Zh] if certain conjectures are assumed.) 2. We make two remarks concerning converse directions. (a) If π is an irreducible representation and δ 0 (π) =δ([ν b1 ρ, ν a1 ρ]) δ([ν b k ρ]) σ has a 1,...,a k,b 1,...,b k satisfying 1 4 of Theorem 1.1, then π is square-integrable. (In fact, condition 1 of Theorem 1.1 is sufficient; cf. Theorem 4.2.1, [Jan4].) (b) If a 1,...,a k,b 1,...,b k satisfy 1 4 of Theorem 1.1, one can ask whether there exists an irreducible representation π having δ 0 (π)=δ([ν b1 ρ, ν a1 ρ]) δ([ν b k ρ]) σ. We expect that this is the case, but it is not proved. (Note that in the case where a k > >a 1 >b 1 > >b k, such a π is constructed in section 7; in the case where a k >b k >a k 1 > b k 1 > >a 1 >b 1, the existence of such a π follows from the results in [Tad5].) 3. We note that we can have irreducible, square-integrable representations π 1,π 2 which have π 1 = π2 but δ 0 (π 1 )=δ 0 (π 2 ) (cf. Theorem 7.7; examples may also be easily obtained from [Tad5]).

3 SQUARE-INTEGRABLE REPRESENTATIONS OF CLASSICAL p-adic GROUPS II 129 The basic proof is by induction on the parabolic rank (cf [B-W]) of the supercuspidal support; the whole paper is essentially one large inductive proof. Thus, we establish Theorem 1.1 and complementary results (Theorem 7.7 and Corollaries 7.8 and 7.9) for a fixed parabolic rank p.r. assuming they all hold when the parabolic rank of the supercuspidal support is less than p.r. We note that if δ 0 (π) =δ([ν b1 ρ, ν a1 ρ]) δ([ν b k ρ]) σ, then the supercuspidal support of π has parabolic rank (a 1 + b 1 +1)+ +(a k + b k + 1). To avoid repeating the hypotheses of Theorem 1.1 in the rest of the theorems in this paper, let us simply let (H) denote the hypotheses of Theorem 1.1. We now discuss the contents section by section. The next section reviews notation and some background results. We also record Theorem 2.4, which says that the first condition in Theorem 1.1 holds (it is a combination of results from [Jan4]). In the third section, we verify condition 2 of the theorem. The basic idea in proving this is to show that should condition 2 fail, there must be something in the Jacquet module of π lower than the χ 0 (π) we started with, a contradiction. For the most part, this can be done with a simple string of embeddings and equivalences; there is one subtler case which occupies most of this section. The fourth and fifth sections verify that conditions 3 and 4 hold. Conditions 3 and 4 are essentially conditions on pairs of segments, and are treated as such. In the fourth section, we assume that b i >a i 1 for some i. Roughly speaking, such segments can be removed from consideration. What remains then has lower parabolic rank and the inductive hypothesis allows us to finish this case. The fifth section deals with the case where b i a i 1 for all i. Here we need to show conditions 3 and 4 directly. The basic idea is to show that if 3 or 4 fails, then π may be embedded into an induced representation where the inducing representation is lower than δ 0 (π), a contradiction. To do this, we use Jacquet module arguments to compare certain induced representations. We note that in the case where σ is generic, condition 3 may be deduced from [Mu]. In order to do the comparisons for section 5, we need to know the existence of square-integrable representations with certain properties. The last two sections are geared toward this. In the sixth section, we prove a technical result which is needed in the seventh section. In the seventh section, we show the existence of square-integrable representations with certain specific δ 0 s, in particular, δ 0 having b i <a i 1 for all i. This is done by analyzing the induced representation δ([ν b1 ρ, ν a1 ρ]) δ([ν b2 ρ, ν a2 ρ],...,[ν b k, where δ([ν b2 ρ, ν a2 ρ],...,[ν b k is a square-integrable representation (by the inductive hypothesis, already constructed) which has δ([ν b2 ρ, ν a2 ρ]) δ([ν b k ρ]) σ as its δ Notation and preliminaries We retain the notation introduced in section 1.2, [Jan4] and will forgo reviewing it here. We will also forgo reviewing certain standard results already discussed in [Jan4]. The Langlands classification and Casselman criteria for S r (F ) ([B-W], [Sil1], [Cas] and [Tad1]) are discussed in section 3.1, [Jan4]. The Langlands classification for GL n (F ) is discussed in section 2.4, [Jan4]. For clarity, we use L (resp. L) when the Langlands subrepresentation is for a general linear group (resp., classical group). We briefly review some of the notation and results from the remainder of [Jan4], and we introduce a few additional items needed for this paper.

4 130 CHRIS JANTZEN Let ρ be an irreducible unitary supercuspidal representation of GL n (F ), σ an irreducible supercuspidal representation of S r (F ). Also, suppose ν x ρ σ is reducible for x = ±α and irreducible for x R \{±α}, whereα 0hasα 1 2 Z. Definition 2.1. Let π be an irreducible representation supported on S((ρ, α); σ). Set { X(π) = χ s min (π) χ = ν x1 ρ ν xm ρ σ has x x m minimal for s min (π) Then, let χ 0 (π) X(π), which is minimal in the lexicographic ordering. The following is Lemma 4.1.2, [Jan4] with a minor change of notation. (As in section 5, [Jan4], it is more convenient to incorporate a negative into the b i s.) Lemma 2.2. χ 0 (π) has the form χ 0 (π) =(ν a1 ρ ν a1 1 ρ ν b1 ρ) (ν a k ρ ν ak 1 ρ ν b k ρ) σ, with a 1 a 2 a k. Definition 2.3. With notation as above, if χ 0 (π) =(ν a1 ρ ν a1 1 ρ ν b1 ρ) (ν a k ρ ν ak 1 ρ ν b k ρ) σ, set δ 0 (π) =δ([ν b1 ρ, ν a1 ρ]) δ([ν b k ρ]) σ. We also recall that if χ 0 (π),δ 0 (π) as above, M = GL (a1 b 1+1)n(F ) GL (ak b k +1)n(F ) S r (F ), then π i GM (δ 0 (π)) (cf. Lemma 4.1.4, [Jan4]). We also note the following, a consequence of Theorem of [Jan4], Lemma of [Jan4], and Lemma of [Jan4]. Theorem 2.4. Suppose (H). Then, condition 1 in Theorem 1.1 holds. We now introduce a bit of notation which will be used in the rest of this paper. In Definition 2.1 above, let t.e.(χ 0 (π)) denote the minimal value of x 1 + +x m which arises. More generally, we let t.e.(π) =t.e.(δ 0 (π)) denote the same value. Implicit in Definition 2.1 is an ordering: χ 1 χ 2 if t.e.(χ 1 ) >t.e.(χ 2 )ort.e.(χ 1 )=t.e.(χ 2 ) and χ 1 is lexicographically higher than χ 2. We extend this ordering as follows: π 1 π 2 if χ 0 (π 1 ) χ 0 (π 2 ). (Typically, π 1 and π 2 will be representations of Levi factors of standard parabolic subgroups.) We introduce one other piece of notation. At times, it will be convenient to write s app when the Jacquet module of S m is taken with respect to the appropriate standard parabolic subgroup. This will only be used when what constitutes the appropriate parabolic subgroup is clear from context. We now recall Definition of [Jan4]. It uses the Jacquet module structures developed in [Tad2]. Since a summary of the notation and results needed from [Tad2] was given in section 3.1 of [Jan4], we will forgo a discussion here. Definition 2.5. Suppose τ is an irreducible representation of GL m (F )andπ a representation of S n (F ). Write µ (π) = i m i ξ i θ i, }.

5 SQUARE-INTEGRABLE REPRESENTATIONS OF CLASSICAL p-adic GROUPS II 131 where ξ i θ i is irreducible and m i is its multiplicity. Let I τ = {i ξ i = τ}. Weset µ τ (π) = i I τ m i ξ i θ i = i I τ m i τ θ i. Similarly, if ξ is a representation of GL r (F )and let J τ = {j ξ (1) j M (ξ) = j n j ξ (1) j ξ (2) j, = τ}. Weset Mτ (ξ) = n j ξ (1) j ξ (2) j = n j τ ξ (2) j J τ j J τ It will be convenient to extend this definition to the case where τ = τ 1 τ j. In this case, we write s τ1 τ j rather than µ τ 1 τ j (since s τ1 τ j s app not µ ). We close with a lemma which will expedite certain calculations. Lemma Suppose η δ([ν b1 ρ, ν a1 ρ]) δ([ν b k. Assume τ is an irreducible representation of a general linear group such that {ν b1 ρ,...,ν b k ρ, ν a1 ρ,...,ν a k ρ} supp(τ) =. Then for any representation ξ of a general linear group, we have µ τ (ξ η) =M τ (ξ) (1 η). 2. Suppose {x Mν ρ(ξ) 0} support(τ) =. Then, x µ τ (ξ η) =(1 ξ) µ τ (η). Further, we note that Mν ρ(τ) 0if and only if the following holds: either x r min (τ) contains a term of the form ν x ρ... or r min (τ) contains a term of the form ν x ρ. Proof. Write M (ξ) = i τ (1) i τ (2) i and µ (η) = j τ j θ j. Then, µ (ξ η) = τ (1) i τ j τ (2) i θ j. i,j Observe that if τ j 1, we must have r min (τ j ) = h να h ρ... with α h {b 1,...,b k,a 1,...,a k } for all h. But, for τ j to contribute to µ τ,wemusthave ν α h ρ supp(τ). Since this is not the case, we must have τ j = 1. Part 1 of the lemma follows. The proof of 2 is straightforward and essentially the same as that of Lemma 5.2.2, [Jan4]; we omit it. 3. Condition 2 The main result in this section is Theorem 3.7, which says that condition 2 in Theorem 1.1 must be satisfied. This condition is interesting only when β 2. We note that when β>2, the proof is fairly straightforward. The case β = 2 occupies most of this section. In this section, we begin the inductive proof. Note that by the inductive hypothesis, we may assume Theorem 1.1, Theorem 7.7 (and Corollaries 7.8 and 7.9) hold when the parabolic rank of the supercuspidal support is less than p.r. By Theorem 2.4, we actually know that condition 1 of Theorem 1.1 holds in general. j.

6 132 CHRIS JANTZEN Suppose π is an irreducible subquotient of δ([ν b1 ρ, ν a1 ρ]) δ([ν b k ρ]) σ. For c 0, let m(c) denote the number of copies of ν ±c ρ which appear in an element of s min (π). We note that m(c) is well-defined and independent of the particular subquotient π (it is simply a matter of supercuspidal support). Set n + (c) =numberofa i or b i which equal c, n (c) =numberofa i or b i which equal c. For c =0,wesimplywriten(0). If δ 0 (π) =δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b k ρ, ν a k ρ]) σ, we similarly define n + (c) =numberofa i or b i which equal c, n (c) =numberofa i or b i which equal c. Again, for c = 0, we simply write n (0). The lemma below then follows from supercuspidal support considerations (it is essentially an extension of Proposition 5.3.2, [Jan4]): Lemma 3.1. For c 1, we have the following: 1. For c>1, n + (c 1) n (c) =m(c 1) m(c) =n +(c 1) n (c). 2. For c =1, n (0) n (1) = 2m(0) m(1) = n(0) n (1). (At times, it will be convenient to let n + (0) = n(0), n +(0) = n (0) and just write n + (c 1) n (c) =n +(c 1) n (c) for c 1.) Proof. Consider the contribution of δ([ν bi ρ, ν ai ρ]). Suppose a i,b i 0. Then δ([ν bi ρ, ν ai ρ]) contributes 1 to both n + (a i ) n (a i +1) andn + (b i ) n (b i +1). Also, δ([ν bi ρ, ν ai ρ]) contributes 1 to both m(a i ) m(a i +1) (resp., 2m(0) m(1) if a i =0)andm(b i ) m(b i +1) (resp., 2m(0) m(1) if b i =0). Thecasesa i 0,b i < 0 and a i < 0,b i 0 may be done similarly (a i,b i < 0 is not possible). Combining these and the corresponding observations for δ([ν b i ρ, ν a i ρ]) gives the lemma. We now note the following observation: Lemma 3.2. Suppose π is an irreducible representation with δ 0 (π) =δ([ν b1 ρ, ν a1 ρ]) δ([ν b k ρ]) σ. Then, there is an irreducible representation θ having δ 0 (θ) =δ([ν b2 ρ, ν a2 ρ]) δ([ν b k ρ]) σ. Proof. Choose θ irreducible such that 1. µ (π) δ([ν b1 ρ, ν a1 ρ]) θ, 2. s app (θ) δ([ν b2 ρ, ν a2 ρ]) δ([ν b k ρ]) σ. Since s app (π) δ([ν b1 ρ, ν a1 ρ]) δ 0 (θ), we see that δ 0 (θ) δ([ν b2 ρ, ν a2 ρ]) δ([ν b k ρ]) σ. The lemma then follows immediately.

7 SQUARE-INTEGRABLE REPRESENTATIONS OF CLASSICAL p-adic GROUPS II 133 Lemma 3.3. If β>2, then condition 2 of Theorem 1.1 holds. Proof. By the preceding lemma, Theorem 4.2.1, [Jan4], and the inductive hypothesis, [ν b2 ρ, ν a2 ρ],...,[ν b k ρ] satisfy the conditions of Theorem 1.1. If b 1 < 0, we must have b 1 = β by Remark 4.4.3, [Jan4], and therefore a 1 β as well. Thus, the only way condition 2 could fail for δ 0 (π) istohaveβ 2 b 1 0 (noting that if a 1 β 1, then b 1 β 2). In particular, this also means β { b 2,..., b k,α+1}, and therefore b i β 1fori 2. First, we check that a 1 β 1. If not, we would have δ([ν b1 ρ, ν a1 ρ]) δ([ν bi ρ, ν ai ρ]) irreducible for all i 2(ifb i < 0, then a i b i β>a 1 +1; if b i 0, then a i >b i β 1 a 1 >b 1 ). Then, we could commute δ([ν b1 ρ, ν a1 ρ]) to the right to get π δ([ν b1 ρ, ν a1 ρ]) δ([ν b2 ρ, ν a2 ρ]) δ([ν b3 ρ, ν a3 ρ]) δ([ν b k = δ([ν b2 ρ, ν a2 ρ]) δ([ν b1 ρ, ν a1 ρ]) δ([ν b3 ρ, ν a3 ρ]) δ([ν b k. = δ([ν b2 ρ, ν a2 ρ]) δ([ν b3 ρ, ν a3 ρ]) δ([ν b k ρ]) δ([ν b1 ρ, ν a1 = δ([ν b2 ρ, ν a2 ρ]) δ([ν b3 ρ, ν a3 ρ]) δ([ν b k ρ]) δ([ν a1 ρ, ν b1, by the irreducibility of δ([ν a1 ρ, ν b1 (cf. Theorem 13.2, [Tad3]). However, by Frobenius reciprocity, this contradicts the minimality of δ 0 (π) (justbyt.e. considerations). Thus, a 1 β 1. For the moment, let us assume α Z. Then, π δ([ν b1 ρ, ν a1 ρ]) δ([ν b2 ρ, ν a2 ρ]) δ([ν b k δ([ρ, ν a1 ρ]) δ([ν b1 ρ, ν 1 ρ]) δ([ν b2 ρ, ν a2 ρ]) δ([ν b k. Now, δ([ν b1 ρ, ν 1 ρ]) δ([ν bi ρ, ν ai ρ]) is irreducible for all i 2(ifb i < 0, then a i b i β > 1; if b i 0, then a i >b i β 1 >b 1 ). Thus, commuting δ([ν b1 ρ, ν 1 ρ]) to the right, we get π δ([ρ, ν a1 ρ]) δ([ν b2 ρ, ν a2 ρ]) δ([ν b k ρ]) δ([ν b1 ρ, ν 1 = δ([ρ, ν a1 ρ]) δ([ν b2 ρ, ν a2 ρ]) δ([ν b k ρ]) δ([νρ,ν b1 by the irreducibility of δ([ν b1 ρ, ν 1 (since b 1 <α). Since δ([ν bi ρ, ν ai ρ]) δ([νρ,ν b1 ρ]) is irreducible for all i 2(ifb i < 0, then a i b i β>b 1 +1; if b i 0, then a i >b 1 and b i β 1 > 1), we have π δ([ρ, ν a1 ρ]) δ([νρ,ν b1 ρ]) δ([ν b2 ρ, ν a2 ρ]) δ([ν b k = δ([νρ,ν b1 ρ]) δ([ρ, ν a1 ρ]) δ([ν b2 ρ, ν a2 ρ]) δ([ν b k δ([νρ,ν b1 ρ]) δ([ν β 1 ρ, ν a1 ρ]) δ([ρ, ν β 2 ρ]) δ([ν b2 ρ, ν a2 ρ]) δ([ν b k

8 134 CHRIS JANTZEN (noting that a 1 >β 2 b 1 ). Now, the same sort of argument as above, applied to δ([ρ, ν β 2 ρ]) this time, gives π δ([νρ,ν b1 ρ]) δ([ν β 1 ρ, ν a1 ρ]) δ([ν b2 ρ, ν a2 ρ]) δ([ν b k ρ]) δ([ρ, ν β 2 = δ([νρ,ν b1 ρ]) δ([ν β 1 ρ, ν a1 ρ]) δ([ν b2 ρ, ν a2 ρ]) δ([ν b k ρ]) δ([ν β+2 ρ, = δ([νρ,ν b1 ρ]) δ([ν β+2 ρ, ρ]) δ([ν β 1 ρ, ν a1 ρ]) δ([ν b2 ρ, ν a2 ρ]) δ([ν b k, noting that β>2 is required for the irreducibility of δ([ν β 1 ρ, ν a1 ρ]) δ([ν β+2 ρ, ρ]). By Frobenius reciprocity and the Casselman criteria, this contradicts the squareintegrability of π. The case α Z is essentially the same. Lemma 3.4. Suppose δ 0 (π) =δ([ρ, ν a1 ρ]) δ([ν 2 ρ, ν a2 ρ]) δ([ν α ρ, ν aα ρ]) σ. Then we must have a 1 <a 2. Proof. Of course, we automatically have a α > >a 2 a 1. Thus we only need to show that a 1 a 2. Suppose a 1 = a 2. Then, π δ([ρ, ν a1 ρ]) δ([ν 2 ρ, ν a1 ρ]) δ([ν 3 ρ, ν a3 ρ]) δ([ν α ρ, ν aα = δ([ν 2 ρ, ν a1 ρ]) δ([ρ, ν a1 ρ]) δ([ν 3 ρ, ν a3 ρ]) δ([ν α ρ, ν aα δ([ν 2 ρ, ν a1 ρ]) δ([ν 2 ρ, ν a1 ρ]) δ([ρ, νρ]) δ([ν 3 ρ, ν a3 ρ]) δ([ν α ρ, ν aα = δ([ν 2 ρ, ν a1 ρ]) δ([ν 2 ρ, ν a1 ρ]) δ([ν 3 ρ, ν a3 ρ]) δ([ν α ρ, ν aα ρ]) δ([ρ, ν = δ([ν 2 ρ, ν a1 ρ]) δ([ν 2 ρ, ν a1 ρ]) δ([ν 3 ρ, ν a3 ρ]) δ([ν α ρ, ν aα ρ]) δ([ν 1 ρ,, contradicting the minimality of δ 0 (π) (t.e. considerations are enough). Lemma 3.5. Suppose k = α and a α > >a 1 > 0. Suppose π δ([ρ, ν a1 ρ]) δ([ν 2 ρ, ν a2 ρ],...,[ν α ρ, ν aα ρ]; σ) (noting that the inducing representation exists by Theorem 7.7 and the inductive hypothesis). Then, π ρ δ([νρ,ν a1 ρ], [ν 2 ρ, ν a2 ρ],...,[ν α ρ, ν aα ρ]; σ). Proof. Let π = δ([ρ, ν a1 ρ]) δ([ν 2 ρ, ν a2 ρ],...,[ν α ρ, ν aα ρ]; σ) and π = ρ δ([νρ,ν a1 ρ], [ν 2 ρ, ν a2 ρ],...,[ν α ρ, ν aα ρ]; σ) π = ρ δ([νρ,ν a1 ρ]) δ([ν 2 ρ, ν a2 ρ],...,[ν α ρ, ν aα ρ]; σ).

9 SQUARE-INTEGRABLE REPRESENTATIONS OF CLASSICAL p-adic GROUPS II 135 Then, π,π π. Next, we claim µ δ([ρ,ν a 1 ρ]) (π )=µ δ([ρ,ν a 1 ρ]) (π )=µ δ([ρ,ν a 1 ρ]) (π ) =2 δ([ρ, ν a1 ρ]) δ([ν 2 ρ, ν a2 ρ],...,[ν α ρ, ν aα ρ]; σ). For π,π, this follows from Lemma 2.6. For π, it is enough to show µ (π ) 2 δ([ρ, ν a1 ρ]) δ([ν 2 ρ, ν a2 ρ],...,[ν α ρ, ν aα ρ]; σ) (since π π ). This follows immediately from the observation that M (ρ) 2ρ 1andµ (δ([νρ,ν a1 ρ]) δ([ν 2 ρ, ν a2 ρ],...,[ν α ρ, ν aα ρ]; σ)) δ([νρ,ν a1 ρ]) δ([ν 2 ρ, ν a2 ρ],...,[ν α ρ, ν aα ρ]; σ). By Frobenius reciprocity, one then has that π must be a subquotient, and therefore (by unitarity) a subrepresentation, of π, as needed. Corollary 3.6. Suppose π δ([ρ, ν a1 ρ]) δ([ν 2 ρ, ν a2 ρ],...,[ν α ρ, ν aα ρ]; σ) with π irreducible and a α > >a 2 >a 1 > 0. Then, either δ 0 (π) =ρ δ([νρ,ν a1 ρ]) δ([ν 2 ρ, ν a2 ρ]) δ([ν α ρ, ν aα ρ]) σ or there are terms of the form ν x ρ with x<0 which appear in χ 0 (π). Proof. We consider three cases based upon Lemma 3.1. Case 1: n (1) = 1. We note the following: n (x) = 0 for all x>α. The proof of this is fairly simple: for x>α, ν x ρ does not appear as a lower segment end by Lemma 4.4.1, [Jan4]. Also, ν x ρ does not appear as an upper segment end since s min (π) ν y ρ... has y {b 1,...,b k,a 1,...,a k }, which implies c 1 α, and therefore c i α for all i. (Alternatively, the assumption that no negative exponents occur in χ 0 (π) rules out the possibility that ν x ρ appears as an upper segment end.) Now, from n (0) n (1) = n(0) n (1), we see that since n(0) = 1, n (1) = 0, and n (1) = 1, we must have n (0) = 2. For c>1, we have n (c) =n (c), which implies n +(c 1) = n + (c 1). Thus, δ 0 (π) =δ([ν d1 ρ, ν c1 ρ]) δ([ν dα+1 ρ, ν cα+1 ρ]) σ with d 1,...,d α+1,c 1,...,c α+1 equal to α,..., 1, 0, 0,a 1,...,a α up to permutation. In order to have no negative exponents in χ 0 (π), we must have α,..., 1 as d i s and a 1,...,a α as c i s. This accounts for everything but the two zeros, one of which must be the remaining d i, the other the remaining c i.thus, δ 0 (π) =δ([ν d1 ρ, ρ]) δ([ν d2 ρ, ν a1 ρ]) δ([ν dα+1 ρ, ν aα ρ]) σ. Next, observe that we must have d 1 =0;ifd 1 = 1, we do not have n (1) = 1. By Lemma 3.2, there is an irreducible θ having δ 0 (θ) =δ([ν d2 ρ, ν a1 ρ]) δ([ν dα+1 ρ, ν aα ρ]) σ. By Theorem 4.2.1, [Jan4], θ is square-integrable. By the inductive hypothesis, the conditions of Theorem 1.1 are satisfied. The only way this can happen (cf. conditions 3 and 4 of that theorem) is if d 2 > >d α+1.thusd 2 = 1,...,d α+1 = α. Thus δ 0 (π) =ρ δ([νρ,ν a1 ρ]) δ([ν α ρ, ν aα ρ]) σ

10 136 CHRIS JANTZEN is the only possibility remaining in Case 1. Case 2: n (1) = 0 and n (2) = 1. By Lemma 3.1, one easily gets δ 0 (π) =δ([ν d1 ρ, ν c1 ρ]) δ([ν d2 ρ, ν c2 ρ]) δ([ν dα ρ, ν cα ρ]) σ with d 1,...,d α,c 1,...,c α equal to α,..., 2, 0,a 1,...,a α up to permutation. Again, to avoid picking up any negative exponents, we must have a 1,...,a α as c i s, and therefore α,..., 2, 0asd i s. Thus, δ 0 (π) =δ([ν d1 ρ, ν a1 ρ]) δ([ν d2 ρ, ν a2 ρ]) δ([ν dα ρ, ν aα ρ]) σ. As in Case 1, there is an irreducible θ with δ 0 (θ) =δ([ν d2 ρ, ν a2 ρ]) δ([ν dα ρ, ν aα ρ]) σ. As in Case 1, θ is square-integrable and the inductive assumption and Theorem 1.1 tell us we must have d 2 > >d α. Further, condition 2 of Theorem 1.1 also tells us we cannot have one of d 2,...,d α equal to 0. Thus the only possibility remaining in Case 2 is δ 0 (π) =δ([ρ, ν a1 ρ]) δ([ν 2 ρ, ν a2 ρ]) δ([ν α ρ, ν aα ρ]) σ. However, by Lemma 3.5, this cannot occur. Thus Case 2 can be eliminated. Case 3: n (1) = n (2) = 0. In this case, let β be the minimal value having n (β )=1. Ifn (x) = 0 for all x, we take β = α+1. Then, n (1) = = n (β 1) = 0, n (β )= = n (α) =1, and n (x) = 0 for all x>α.sincen (1) = n (1) = 0, n (0) n (1) = n(0) n (1) tells us n (0) = n(0) = 1. Next, we check that a x = x for 1 x β 2. First, n + (1) n (2) = n + (1) n (2) has n (2) = 0 and n (2) = 1. Since n + (1) 1andn + (1) 0, we see that n + (1) = 1 and n +(1) = 0. To have n + (1) = 1, we must have a 1 =1. We iterate this argument until we get to the following: n + (β 2) n (β 1) = n + (β 2) n (β 1) has n (β 1) = 0 and n (β 1) = 1. Therefore, n + (β 2) = 1 and n +(β 2) = 0. For n + (β 2) = 1, we must have a β 2 = β 2. Finally, observe that for x β,wehaven (x) = n (x). Therefore, since n +(x 1) n (x) =n + (x 1) n (x), we have n +(y) =n + (y) fory β 1. Thus, if δ 0 (π) =δ([ν d1 ρ, ν c1 ρ]) δ([ν 2 ρ, ν a2 ρ]) δ([ν dj ρ, ν cj ρ]) σ, we must have d 1,...,d j,c 1,...,c j equal to α,..., β, 0,a β 1,...,a α up to permutation. In order that no negative exponents appear, a β 1,...,a α must be c i s and therefore α,..., β, 0mustbed i s. An argument like that used in the previous case then allows us to conclude that δ 0 (π) =δ([ρ, ν a β 1 ρ]) δ([ν β ρ, ν a β ρ]) δ([ν α ρ, ν aα ρ]) σ. However, since β > 2, this is in contradiction to Theorem 2.4. Thus we can also eliminate Case 3. The corollary now follows. Theorem 3.7. Suppose (H). Suppose that Theorems 1.1 and 7.7 (and Corollaries 7.8 and 7.9) hold when the parabolic rank of the supercuspidal support is less than p.r. Then, condition 2 in Theorem 1.1 holds.

11 SQUARE-INTEGRABLE REPRESENTATIONS OF CLASSICAL p-adic GROUPS II 137 Proof. By Lemma 3.3, we may assume β =2(whichmeansα must be an integer). As in the proof of Lemma 3.3 (first paragraph), if b 1 < 0, we have b 1 = β and can easily deduce that condition 2 of Theorem 1.1 holds. Also, as in the proof of Lemma 3.3 (first paragraph), if b 1 0, the only way condition 2 of Theorem 1.1 could fail is if β 2 b 1 0, i.e., b 1 = 0. Therefore, all we need to do is show b 1 0. So, let us suppose b 1 = 0. Then, δ 0 (π) =δ([ρ, ν a1 ρ]) δ([ν b2 ρ, ν a2 ρ]) δ([ν b k ρ]) σ. Next, let [ν d1 ρ, ν c1 ρ],...,[ν d k ρ, ν c k ρ] denote a permutation of [ν b1 ρ, ν a1 ρ],...,[ν b k ρ] satisfying d 1 d k. To make it unique, note that only d i =0 can occur for more than one value of i, and then only for two. If we should have b i =0forsomei>1, then [ρ, ν ai ρ] should appear before [ρ, ν a1 ρ] in the above list. We claim that π δ([ν d1 ρ, ν c1 ρ]) δ([ν d k ρ, ν c k which we may write as π δ([ν d1 ρ, ν c1 ρ]) δ([ν dj ρ, ν cj ρ]) δ([ρ, ν a1 ρ]) δ([ν 2 ρ, ν cj+2 ρ]) δ([ν α ρ, ν c k. Suppose [ν d1 ρ, ν c1 ρ]=[ν bj ρ, ν aj ρ]. Observe that for i<j, so [ν bi ρ, ν ai ρ] [ν bj ρ, ν aj ρ] δ([ν bi ρ, ν ai ρ]) δ([ν d1 ρ, ν c1 ρ]) = δ([ν d1 ρ, ν c1 ρ]) δ([ν bi ρ, ν ai ρ]) (irreducible). Thus, we may commute δ([ν d1 ρ, ν c1 ρ]) to the left: π δ([ν b1 ρ, ν a1 ρ]) δ([ν bj 2 ρ, ν aj 2 ρ]) δ([ν bj 1 ρ, ν aj 1 ρ]) δ([ν d1 ρ, ν c1 ρ]) δ([ν bj+1 ρ, ν aj+1 ρ]) δ([ν b k = δ([ν b1 ρ, ν a1 ρ]) δ([ν bj 2 ρ, ν aj 2 ρ]) δ([ν d1 ρ, ν c1 ρ]) δ([ν bj 1 ρ, ν aj 1 ρ]) δ([ν bj+1 ρ, ν aj+1 ρ]) δ([ν b k. = δ([ν d1 ρ, ν c1 ρ]) δ([ν b1 ρ, ν a1 ρ]) δ([ν bj 2 ρ, ν aj 2 ρ]) δ([ν bj 1 ρ, ν aj 1 ρ]) δ([ν bj+1 ρ, ν aj+1 ρ]) δ([ν b k. Iterating this argument, we next commute δ([ν d2 ρ, ν c2 ρ]) to the left, etc., and we get the result claimed. Now, by Lemma 5.5, [Jan2], π δ([ν d1 ρ, ν c1 ρ]) δ([ν dj ρ, ν cj ρ]) δ([ρ, ν a1 ρ]) δ([ν 2 ρ, ν cj+2 ρ]) δ([ν α ρ, ν c k π δ([ν d1 ρ, ν c1 ρ]) δ([ν dj ρ, ν cj ρ]) θ for some irreducible θ δ([ρ, ν a1 ρ]) δ([ν 2 ρ, ν cj+2 ρ]) δ([ν α ρ, ν c k. Since s app (π) δ([ν d1 ρ, ν c1 ρ]) δ([ν dj ρ, ν cj ρ]) δ 0 (θ)

12 138 CHRIS JANTZEN we must have t.e.(δ 0 (θ)) t.e.(δ([ρ, ν a1 ρ]) δ([ν 2 ρ, ν cj+2 ρ]) δ([ν α ρ, ν c k ρ]) σ). By Corollary 3.6, δ 0 (θ) =ρ δ([νρ,ν a1 ρ]) δ([ν 2 ρ, ν cj+2 ρ]) δ([ν α ρ, ν c k ρ]) σ. Then, π δ([ν d1 ρ, ν c1 ρ]) δ([ν dj ρ, ν cj ρ]) θ δ([ν d1 ρ, ν c1 ρ]) δ([ν dj ρ, ν cj ρ]) ρ δ([νρ,ν a1 ρ]) δ([ν 2 ρ, ν cj+2 ρ]) δ([ν α ρ, ν c k = ρ δ([ν d1 ρ, ν c1 ρ]) δ([ν dj ρ, ν cj ρ]) δ([νρ,ν a1 ρ]) δ([ν 2 ρ, ν cj+2 ρ]) δ([ν α ρ, ν c k, contradicting the square-integrability of π. The theorem follows. 4. Conditions 3 and 4 the first part of proof In this section, we begin the proof that conditions 3 and 4 in Theorem 1.1 hold. We will view conditions 3 and 4 of Theorem 1.1 as conditions on pairs of segments. By Lemma 3.2, Theorem 4.2.1, [Jan4], and the inductive hypothesis, we know that [ν bi ρ, ν ai ρ], [ν bj ρ, ν aj ρ] satisfy these conditions when i, j > 1withi j. Thus, our goal is to show [ν b1 ρ, ν a1 ρ], [ν bi ρ, ν ai ρ] satisfy these conditions for i>1. In this section, we address the case where b i >a 1 for some i 2. For any such i, we have a i >b i >a 1 >b 1, so conditions 3 and 4 of Theorem 1.1. certainly hold for [ν b1 ρ, ν a1 ρ], [ν bi ρ, ν ai ρ]. Roughly speaking, we show that there is an irreducible representation θ, whereδ 0 (θ) is essentially δ 0 (π) withsuchδ([ν bi ρ, ν ai ρ]) removed. We can then use the inductive hypothesis to show δ 0 (θ) satisfies conditions 3. and 4. of Theorem 1.1, which is enough to finish this case. Let us take a moment to review where we stand with respect to the inductive hypothesis. Of course, we may assume Theorems 1.1 and 7.7 (and Corollaries 7.8 and 7.9) hold when the parabolic rank of the supercuspidal support is less than p.r. In addition, we may assume that condition 1 of Theorem 1.1 holds in general (Theorem 2.4) and that condition 2 of Theorem 1.1 holds when the parabolic rank of the supercuspidal support is equal to p.r. (Theorem 3.7 and the inductive hypothesis). The proof of the following proposition is essentially the same as that used in the proof of Theorem 3.7; we will not go through the details. Proposition 4.1. Suppose δ 0 (π) =δ([ν b1 ρ, ν a1 ρ]) δ([ν b k ρ]) σ. Then there exist (unique) l, m, l + m +1=k and permutation [ν b 1 ρ, ν a 1 ρ],...,[ν b l ρ, ν a l ρ], [ν b 1 ρ, ν a1 ρ], [ν b 1 ρ, ν a 1 ρ],...,[ν b m ρ, ν a m ρ] of [ν b1 ρ, ν a1 ρ],...,[ν b k ρ] satisfying the following conditions: 1. b 1 >b 2 > >b l >a 1, 2. a 1 <a 2 < <a m with b i a 1 for all i. Furthermore, π δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b l ρ, ν a l ρ]) δ([ν b 1 ρ, ν a1 ρ]) δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b m ρ, ν a m.

13 SQUARE-INTEGRABLE REPRESENTATIONS OF CLASSICAL p-adic GROUPS II 139 The proof for conditions 3 and 4 of Theorem 1.1 will be broken into two parts we address the case l 1 here; the case l = 0 will be addressed in the next section. So, for the remainder of this section, we assume l 1. Observe that by Lemma 3.2, Theorem 4.2.1, [Jan4], and the inductive hypothesis, conditions 3 and 4 of Theorem 1.1 hold for [ν b 1 ρ, ν a construction, we cannot have a i >b i >a j a m >a m 1 > >a 1 >b 1 >b >b j 1 ρ],...,[ν b m ρ, ν a m ρ]. By for i>j. Therefore, we have 2 > >b m. Then, by Lemma 5.5, [Jan2], π δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b l ρ, ν a l ρ]) δ([ν b 1 ρ, ν a1 ρ]) δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b m 1 ρ, ν a m 1 ρ]) (δ([ν b m ρ, ν a m ) π δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b l ρ, ν a l ρ]) δ([ν b 1 ρ, ν a1 ρ]) δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b m 1 ρ, ν a m 1 ρ]) ξ for some irreducible ξ δ([ν b m ρ, ν a m. By the inductive hypothesis, ξ is one of the representations from Theorem 7.7; by t.e. (and Frobenius reciprocity), to avoid contradicting the minimality of δ 0 (π), we must have ξ = δ([ν b m ρ, ν a m ρ]; σ)t. Then, π δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b l ρ, ν a l ρ]) δ([ν b 1 ρ, ν a1 ρ]) δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b m 2 ρ, ν a m 2 ρ]) (δ([ν b m 1 ρ, ν a m 1 ρ]) δ([ν b m ρ, ν a m ρ]; σ)t ) π δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b l ρ, ν a l ρ]) δ([ν b 1 ρ, ν a1 ρ]) δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b m 1 ρ, ν a m 1 ρ]) ξ for some irreducible ξ δ([ν b m 2 ρ, ν a m 2 ρ]) δ([ν b m ρ, ν a m ρ]; σ)t. Again, by the inductive hypothesis, ξ is one of the representations from Theorem 7.7; again, by t.e. considerations, we get ξ = δ([ν b m 1 ρ, ν a m 1 ρ], [ν b m ρ, ν a m ρ]; σ)t. Iterating this argument, we get π δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b l ρ, ν a l ρ]) θ for some irreducible θ δ([ν b1 ρ, ν a1 ρ]) δ([ν b 1 ρ, ν a 1 ρ],...,[ν b m ρ, ν a m ρ]; σ)t. Lemma 4.2. With notation as above, s δ([ν b (π) =c 1 ρ,ν a 1 ρ]) δ([ν b l ρ,ν a l ρ]) δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b l ρ, ν a l ρ]) θ for some nonzero integer c. Proof. First, by Lemma 5.5, [Jan2], there is an irreducible θ l δ([ν b l ρ, ν a l ρ]) θ such that π δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b l 1 ρ, ν a l 1 ρ]) θ l. Iterating this argument, we obtain θ l 1,θ l 2,...,θ 2 such that each is irreducible, θ i δ([ν b i ρ, ν a i ρ]) θ i+1,andπ δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b i 1 ρ, ν a i 1 ρ]) θ i. First, consider δ([ν b 1 ρ, ν a 1 ρ]) θ 2. We claim µ δ([ν b 1 ρ,ν a 1 ρ]) (δ([ν b 1 ρ, ν a 1 ρ]) θ 2 )=2 δ([ν b 1 ρ, ν a 1 ρ]) θ 2.

14 140 CHRIS JANTZEN Write µ (θ 2 )= h τ h ξ h. Then, a 1 +1 a 1 +1 µ (δ([ν b 1 ρ, ν a 1 ρ]) θ 2 )= δ([ν i+1 ρ, ν b 1 ρ]) h i= b 1 j=i δ([ν j ρ, ν a 1 ρ]) τh δ([ν i ρ, ν j 1 ρ]) ξ h. We now need an observation (the importance of which was underscored for the author by [Mœ3]). Suppose δ([ν bs ρ, ν as ρ]) appears to the left of δ([ν b 1 ρ, ν a 1 ρ]) in the original δ 0 (π). For s = 1, we have a 1 > b 1 > a 1 > b 1 by construction. For s > 1, we have that [ν b 1 ρ, ν a 1 ρ], [ν b s ρ, ν as ρ] satisfy the conditions of Theorem 1.1 (by Lemma 3.2, Theorem 4.2.1, [Jan4], and the inductive hypothesis). Therefore, either a 1 >b 1 >a s >b s or a 1 >a s >b s >b 1;wecanrule out the latter since b 1 is known to be larger than b s by construction. Now, suppose δ([ν bs ρ, ν as ρ]) appears to the right of δ([ν b 1 ρ, ν a 1 ρ]) in the original δ0 (π). Again, since [ν b 1 ρ, ν a 1 ρ], [ν b s ρ, ν as ρ] satisfy the conditions of Theorem 1.1, either a s >b s >a 1 >b 1 or a s >a 1 >b 1 >b s; we can rule out the former since b 1 is known to be larger than b s by construction. In particular, the observation we need is the following: there is no element of {a 1,...,a k,b 1,...,b k }\{b 1,a 1} in the interval [b 1,a 1 ]. Therefore, to contribute to µ,wemusthavej δ([ν b 1 ρ,ν a 1 ρ]) b 1 +1 in the displayed equation above. Suppose j = b Then, τ h = δ([ν b 1 ρ, ν i ρ]). If i b 1 + 1, then by central character considerations, θ 2 ν i ρ ν i 1 ρ ν b 1 ρ ξh π δ([ν b 1 ρ, ν a 1 ρ]) ν i ρ ν i 1 ρ ν b1 ρ ξ h = ν i ρ ν i 1 ρ ν b 1 ρ δ([ν b 1 ρ, ν a 1 ρ]) ξh, contradicting the square-integrability of π. Thus,i = b 1 +1, so τ h =1andξ h = θ 2. This contributes one copy of δ([ν b 1 ρ, ν a 1 ρ]) θ 2 (and nothing else) to µ δ([ν b 1 ρ,ν a 1 ρ]) (δ([ν b 1 ρ, ν a 1 ρ]) θ 2 ). If j b 1 + 1, then we must have i = b 1 (since νb 1 ρ appear only once in r min (δ([ν b 1 ρ, ν a 1 ρ]))). A similar argument tells us we must again have τh =1, ξ h = θ 2. So, j = b 1 contributes a second copy of δ([ν b 1 ρ, ν a 1 ρ]) θ 2 to µ δ([ν b 1 ρ,ν a 1 ρ]) (δ([ν b 1 ρ, ν a 1 ρ]) θ 2 ). Thus, µ δ([ν b 1 ρ,ν a 1 ρ]) (δ([ν b 1 ρ, ν a 1 ρ]) θ 2 )=2 δ([ν b 1 ρ, ν a 1 ρ]) θ 2, as claimed. Let us briefly discuss the first iteration of this argument. Our claim is that µ δ([ν b 2 ρ,ν a 2 ρ]) (δ([ν b 2 ρ, ν a 2 ρ]) θ 3 )=2 δ([ν b 2 ρ, ν a 2 ρ]) θ 3. The analogue to the observation above is that there is nothing in {a 1,...,a k,b 1,...,b k }\{a 1,b 1} which is strictly between b 2 +1 and a 2. Again, let us write

15 SQUARE-INTEGRABLE REPRESENTATIONS OF CLASSICAL p-adic GROUPS II 141 µ (θ 3 )= h τ h ξ h. Then, µ (δ([ν b 2 ρ, ν a 2 ρ]) θ 3 )= a 2 +1 a 2 +1 δ([ν i+1 ρ, ν b 2 ρ]) h i= b j=i 2 δ([ν j ρ, ν a 2 ρ]) τh δ([ν i ρ, ν j 1 ρ]) ξ h. Again, to contribute to µ,wemusthavej δ([ν b 2 ρ,ν a 2 ρ]) b Ifj = b 2 +1,we get i = b 2 +1asbefore: ifi b 2 + 1, then central character considerations tell us θ 3 ν i ρ ν i 1 ρ ν b 1 ρ ξh π δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b 2 ρ, ν a 2 ρ]) ν i ρ ν i 1 ρ ν b 2 ρ ξh = ν i ρ ν i 1 ρ ν b 2 ρ δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b 2 ρ, ν a 2 ρ]) ξh, which contradicts the square-integrability of π. Wheni = b 2 + 1, we get a contribution of one copy of δ([ν b 2 ρ, ν a 2 ρ]) θ 3 to µ δ([ν b 2 ρ,ν a 2 ρ]) (δ([ν b 2 ρ, ν a 2 ρ]) θ 3 ). If j < b 2 + 1, then we must have i = b 2 and τ h = δ([ν b 2 ρ, ν j 1 ρ]). The same argument as before tells us we must have j = b 2. Therefore, τ h = 1 and ξ h = θ 3, giving a contribution of one more copy of δ([ν b 2 ρ, ν a 2 ρ]) θ 3 to µ δ([ν b 2 ρ,ν a 2 ρ]) (δ([ν b 2 ρ, ν a 2 ρ]) θ 3 ), as claimed. The lemma follows by iterating. It follows immediately from properties of Jacquet functors that there is an irreducible representation θ satisfying 1. s app (π) δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b l ρ, ν a l ρ]) θ, 2. s app (θ ) δ([ν b1 ρ, ν a1 ρ]) δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b m ρ, ν a m ρ]) σ. The corollary below then follows immediately from Lemma 4.2. Corollary 4.3. With notation as above, θ = θ. Theorem 4.4. Suppose (H). Suppose that Theorems 1.1 and 7.7 (and Corollaries 7.8 and 7.9) are proved when the parabolic rank of the supercuspidal support is less than p.r. and that condition 2 of Theorem 1.1 is also proved when the parabolic rank of the supercuspidal support is equal to p.r. Then, if l 1, conditions 3 and 4 of Theorem 1.1 hold (l as in Proposition 4.1). Proof. By the inductive hypothesis, Lemma 3.2, and Theorem of [Jan4], we have that [ν bi ρ, ν ai ρ], [ν b k ρ, ν aj ρ] satisfy conditions 3 and 4 of Theorem 1.1 when i>j 2. Therefore, it is enough to show that [ν b1 ρ, ν a1 ρ], [ν bi ρ, ν ai ρ], i 2 satisfy conditions 3 and 4. Further, by construction [ν b i ρ, ν a i ρ], [ν b 1 ρ, ν a1 ρ] satisfy conditions 3 and 4 Thus, it remains to show that [ν b1 ρ, ν a1 ρ], [ν b i ρ, ν a i ρ] satisfy conditions 3 and 4. To show [ν b1 ρ, ν a1 ρ], [ν b i ρ, ν a i ρ] satisfy conditions 3 and 4, it is enough (by the inductive hypothesis) to show the following: δ 0 (θ) =δ([ν b1 ρ, ν a1 ρ]) δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b m ρ, ν a m ρ]) σ (θ as above). The theorem then follows immediately. Consider the permutation of [ν b 1 ρ, ν a 1 ρ],...,[ν b l ρ, ν a l ρ] where a 1,...,a l appear in increasing order of size. Since it is this permutation we need for the

16 142 CHRIS JANTZEN remainder of this proof, let us abuse notation somewhat and simply denote it by [ν b 1 ρ, ν a 1 ρ],...,[ν b l ρ, ν a l ρ], where a 1 < <a l. Then, π δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b l ρ, ν a l ρ]) δ([ν b 1 ρ, ν a1 ρ]) δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b m ρ, ν a m (as in the proof of Theorem 3.7, any δ([ν b i ρ, ν a i ρ]), δ([ν b i+1 ρ, ν a i+1 ρ]) which has a i+1 <a i can be transposed by an irreducibility argument; a sequence of such transpositions suffices to rearrange terms in the order described). First, observe that by construction, δ 0 (θ ) δ([ν b1 ρ, ν a1 ρ]) δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b m ρ, ν a m ρ]) σ. Also, since we see that s app (π) δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b l ρ, ν a l ρ]) δ0 (θ ), t.e.(δ 0 (θ )) t.e.(δ([ν b1 ρ, ν a1 ρ]) δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b m ρ, ν a m ρ]) σ). Therefore, t.e.(δ 0 (θ ))=t.e.(δ([ν b1 ρ, ν a1 ρ]) δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b m ρ, ν a m ρ]) σ). Now, suppose δ 0 (θ )=δ([ν d1 ρ, ν c1 ρ]) δ([ν cs ρ, ν cs ρ]) σ. If we assume δ 0 (θ ) δ([ν b1 ρ, ν a1 ρ]) δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b m ρ, ν a m ρ]) σ, wemusthaveδ([ν d1 ρ, ν c1 ρ]) δ([ν cs ρ, ν cs ρ]) σ lower than δ 0 (θ ) δ([ν b1 ρ, ν a1 ρ]) δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b m ρ, ν a m ρ]) σ lexicographically. Suppose δ([ν d1 ρ, ν c1 ρ]) is less than δ([ν b1 ρ, ν a1 ρ]) lexicographically. Then, a j >b j >a 1 c 1.Ifd 1 b l,thenδ([ν b j ρ, ν a j ρ]) δ([ν d 1 ρ, ν c1 ρ]) is irreducible for all j. By Corollary 4.3, π δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b l ρ, ν a l ρ]) δ([ν d 1 ρ, ν c1 ρ]) δ([ν ds ρ, ν cs = δ([ν d1 ρ, ν c1 ρ]) δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b l ρ, ν a l ρ]) δ([ν ds ρ, ν cs, contradicting the minimality of δ 0 (π). If d 1 >b l,then δ([ν b l ρ, ν a l ρ]) δ([ν d 1 ρ, ν c1 ρ]) = δ([ν b l ρ, ν c 1 ρ]) δ([ν d1 ρ, ν a l ρ]) + L(δ([ν d1 ρ, ν c1 ρ]),δ([ν b l ρ, ν a l ρ])) (noting that c 1 {b 1,b 1,...,b m,a 1,a 1,...,a m} forces c 1 > b l ). Therefore, by Lemma 5.5, [Jan4], π δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b l 1 ρ, ν a l 1 ρ]) δ([ν b l ρ, ν c 1 ρ]) δ([ν d1 ρ, ν a l ρ]) δ([ν d 2 ρ, ν c2 ρ]) δ([ν ds ρ, ν cs

17 SQUARE-INTEGRABLE REPRESENTATIONS OF CLASSICAL p-adic GROUPS II 143 or π δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b l 1 ρ, ν a l 1 ρ]) δ([ν d 1 ρ, ν c1 ρ]) δ([ν b l ρ, ν a l ρ]) δ([ν d 2 ρ, ν c2 ρ]) δ([ν ds ρ, ν cs. In the first case, we now argue as we did when d 1 b l to obtain π δ([ν b l ρ, ν c 1 ρ]) δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b l 1 ρ, ν a l 1 ρ]) δ([ν d1 ρ, ν a l ρ]) δ([ν d 2 ρ, ν c2 ρ]) δ([ν ds ρ, ν cs, again contradicting the minimality of δ 0 (π). In the second case, we can iterate the argument above: if d 1 b l 1, we can again commute ρ, ν δ([ν d1 c1 ρ]) forward to get a contradiction. If d 1 >b l 1, we can again look at δ([ν b l 1 ρ, ν b l 1 ρ]) δ([ν d1 ρ, ν c1 ρ]) and repeat the argument above. In either case, we eventually get π δ([ν x ρ, ν c1 ρ])..., with x = d 1 or b j for some j. In any case, we contradict the minimality of δ 0(π). Therefore, we could not have had δ([ν d1 ρ, ν c1 ρ]) less than δ([ν b1 ρ, ν a1 ρ]) lexicographically. Thus c 1 = a 1 and d 1 = b 1. Since δ([ν d1 ρ, ν c1 ρ]) = δ([ν b1 ρ, ν a1 ρ]), let us now suppose δ([ν d2 ρ, ν c2 ρ]) is lexicographically lower than δ([ν b 1 ρ, ν a 1 ρ]). We can easily see that π δ([ν b1 ρ, ν a1 ρ]) δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b l ρ, ν a l ρ]) δ([ν d2 ρ, ν c2 ρ]) δ([ν ds ρ, ν cs. Now, take i such that a i <a 1 <a i+1 (subject to the obvious convention if a 1 <a 1 or a 1 > a l ). Then, [ν b j ρ, ν a j ρ] = [ν b j+1 ρ, ν aj+1 ρ]for1 j i. Further, [ν b 1 ρ, ν a 1 ρ]=[ν b i+2 ρ, ν ai+2 ρ]. Repeating the argument above, we get π δ([ν b1 ρ, ν a1 ρ]) δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b i ρ, ν a i ρ]) δ([ν x ρ, ν c2 ρ])... = δ([ν b1 ρ, ν a1 ρ]) δ([ν b2 ρ, ν a2 ρ]) δ([ν bi+1 ρ, ν ai+1 ρ]) δ([ν x ρ, ν c2 ρ])... for x = d 2 or b j some j>i. Again, this contradicts the minimality of δ 0(π). Thus we must have d 2 = b 1 and c 2 = a 1. We can now iterate this argument to see that δ 0 (θ )=δ([ν b1 ρ, ν a1 ρ]) δ([ν b 1 ρ, ν a 1 ρ]) δ([ν b m ρ, ν a m ρ]) σ, as claimed. This finishes the proof. 5. Conditions 3 and 4 the second part of proof In this section, we address the second part of the proof of conditions 3 and 4, the case where l = 0. The basic idea here is this: we assume δ 0 (π) fails to satisfy these conditions, then show that π can be embedded into another induced representation, one which corresponds to a δ 0 lower than assumed. Let us review where we stand with respect to the inductive hypothesis. Of course, we may assume Theorems 1.1 and 7.7 (and Corollaries 7.8 and 7.9) hold when the parabolic rank of the supercuspidal support is less than p.r. In addition, we may assume that condition 1 of Theorem 1.1 holds in general (Theorem 2.4) and

18 144 CHRIS JANTZEN that condition 2 of Theorem 1.1 holds when the parabolic rank of the supercuspidal support is equal to p.r. (Theorem 3.7 and the inductive hypothesis). We are also free to assume that conditions 3 and 4 of Theorem 1.1 hold when the parabolic rank of the supercuspidal support is equal to p.r. and l 1, but we do not actually use this in dealing with the case l =0. Again, by Lemma 3.2, Theorem of [Jan4], and the inductive hypothesis, [ν b2 ρ, ν a2 ρ],...,[ν b k ρ] satisfy the conditions of Theorem 1.1. Therefore, for i>j 2, we must have either a i >b i >a j >b j or a i >a j >b j >b i. However, to have l = 0, the former cannot occur. Thus, a i >a j >b j >b i for i, j 2. Therefore, a k > >a 2 >b 2 > >b k. Further, we must have b 2 a 1. By construction, we have a 1 a 2. We now show that the inequality is strict. Lemma 5.1. a 1 a 2. Proof. Suppose not. Let us use a 1 for both a 1 and a 2.Since π δ([ν b1 ρ, ν a1 ρ]) δ([ν b2 ρ, ν a1 ρ]) δ([ν b3 ρ, ν a3 ρ]) δ([ν b k = δ([ν b2 ρ, ν a1 ρ]) δ([ν b1 ρ, ν a1 ρ]) δ([ν b3 ρ, ν a3 ρ]) δ([ν b k, we see that in order for δ 0 (π) to be minimal lexicographically, we must have b 1 b 2. First, suppose b 1 < 0. By Theorem 2.4, we have b 1 b 2. Then, π δ([ν b1 ρ, ν a1 ρ]) δ([ν b2 ρ, ν a1 ρ]) δ([ν b3 ρ, ν a3 ρ]) δ([ν b k = δ([ν b2 ρ, ν a1 ρ]) δ([ν b1 ρ, ν a1 ρ]) δ([ν b3 ρ, ν a3 ρ]) δ([ν b k δ([ν b2 ρ, ν a1 ρ]) δ([ν b1+1 ρ, ν a1 ρ]) ν b1 ρ δ([ν b3 ρ, ν a3 ρ]) δ([ν b k = δ([ν b2 ρ, ν a1 ρ]) δ([ν b1+1 ρ, ν a1 ρ]) δ([ν b3 ρ, ν a3 ρ]) δ([ν b k ρ]) ν b1 ρ σ = δ([ν b2 ρ, ν a1 ρ]) δ([ν b1+1 ρ, ν a1 ρ]) δ([ν b3 ρ, ν a3 ρ]) δ([ν b k ρ]) ν b1 ρ σ, where ν b1 ρ δ([ν bi ρ, ν ai ρ]) is irreducible for i 3since b 1 < b i 1. This contradicts the minimality of δ 0 (π) (by Frobenius reciprocity and total exponent considerations). Thus, we are now free to assume b 1 0. Next,wenotethatifb 2 < 0, then by Theorem 3.7 we have b 1 b 2 1=β 1. Thus (whether b 2 < 0orb 2 0), Theorem 7.7 and the inductive hypothesis allow us to define π t = δ([ν b1 ρ, ν a1 ρ]) δ([ν b2 ρ, ν a1 ρ], [ν b3 ρ, ν a3 ρ],...,[ν b k, π t = δ([ν a1 ρ, ν a1 ρ]) δ([ν b2 ρ, ν b1 ρ], [ν b3 ρ, ν a3 ρ],...,[ν b k, π t = δ([ν b1 ρ, ν a1 ρ]) δ([ν b1+1 ρ, ν a1 ρ]) δ([ν b2 ρ, ν b1 ρ], [ν b3 ρ, ν a3 ρ],...,[ν b k

19 SQUARE-INTEGRABLE REPRESENTATIONS OF CLASSICAL p-adic GROUPS II 145 (noting that [ν b2 ρ, ν b1 ρ]= is possible). By Corollary 7.8, we see that π t πt. By Lemma 5.4, [BDK], πt = ρ, ν δ([ν b1 a1 ρ]) δ([ν a1 ρ, ν b1 1 ρ]) δ([ν b2 ρ, ν b1 ρ], [ν b3 ρ, ν a3 ρ],...,[ν b k, so π t π t as well. It follows fairly easily from Lemma 2.6 that µ δ([ν b 1 +1 ρ,ν a 1 ρ]) δ([ν b 1 +1 ρ,ν a 1 ρ]) (π t )=µ δ([ν b 1 +1 ρ,ν a 1 ρ]) δ([ν b 1 +1 ρ,ν a 1 ρ]) (π t) = δ([ν b1+1 ρ, ν a1 ρ]) δ([ν b1+1 ρ, ν a1 ρ]) δ([ν b1 ρ, ν b1 ρ]) δ([ν b2 ρ, ν b1 ρ], [ν b3 ρ, ν a3 ρ],...,[ν b k Since π π t,wehaveµ (π) 0. Now, δ([ν b 1 ρ,ν a 1 ρ]) δ([ν b 2 ρ,ν a 1 ρ]) m ρ, ν δ([ν b 1 +1 ρ,ν a 1 ρ]) δ([ν b 1 +1 ρ,ν a 1 ρ]) (δ([ν b1 a1 ρ]) δ([ν b2 ρ, ν a1 ρ])) 0 (π) 0. µ δ([ν b 1 +1 ρ,ν a 1 ρ]) δ([ν b 1 +1 ρ,ν a 1 ρ]) From above, this implies π π t. By unitarity, π π t, contradicting the squareintegrability of π. Thuswecouldnothavehada 1 = a 2, as claimed. If b 1 >b 2,wehavea k > >a 2 >a 1 >b 1 >b 2 > >b k, which certainly satisfies conditions 3 and 4 of Theorem 1.1. So, let us assume b 1 b 2. Then, a 2 >a 1 b 2 b 1. Consider the possibility that b 2 = a 1 ; write a 1 for both. Observe that π δ([ν b1 ρ, ν a1 ρ]) δ([ν a1 ρ, ν a2 ρ]) δ([ν b3 ρ, ν a3 ρ],...,[ν b k = δ([ν a1 ρ, ν a2 ρ]) δ([ν b1 ρ, ν a1 ρ]) δ([ν b3 ρ, ν a3 ρ],...,[ν b k π δ([ν a1 ρ, ν a2 ρ]) θ, for some irreducible θ δ([ν b1 ρ, ν a1 ρ]) δ([ν b3 ρ, ν a3 ρ],...,[ν b k (Lemma 5.5, [Jan2]). Now, consider µ δ([ν a 1 ρ,ν a 2 ρ]) (π). Write µ (θ) = h τ h θ h. Then, µ (δ([ν a1 ρ, ν a2 ρ]) θ) = a 2+1 a 2+1 [δ([ν i+1 ρ, ν a1 ρ]) h i= a 1 j=i δ([ν j ρ, ν a2 ρ]) τ h ] δ([ν i ρ, ν j 1 ρ]) θ h. To contribute to µ ρ. There are three possible δ([ν a 1 ρ,ν a 2 ρ]),wemustpickupaν a1 sources. If the ν a1 ρ comes from δ([ν j ρ, ν a2 ρ]), we must have j = a 1.Theni = a 1 and τ h = 1, giving a contribution of one copy of δ([ν a1 ρ, ν a2 ρ]) θ. Suppose δ([ν i+1 ρ, ν a1 ρ]) contributes the ν a1 ρ. Then, i = a By Lemma 2.6, we then have τ h = 1. Therefore, j = a Thus we get a contribution of one more copy of δ([ν a1 ρ, ν a2 ρ]) θ. Finally, suppose τ h contributes the ν a1 ρ. Since anything in r min (τ h ) has the form ν x ρ... with x {b 1,b 3,...,b k,a 1,a 3,...,a k },wesee that r min (τ h )mustcontainsomethingoftheform(ν x ρ ν x 1 ρ ν a1 ρ)... for such an x. Further, since a 3,...,a k >a 2,wemusthavex a 1. Now, we have that s min (θ) contains something of the form (ν x ρ ν x 1 ρ ν a1 ρ)... Therefore, by central character considerations, θ ν x ρ ν x 1 ρ ν a1 ρ θ

20 146 CHRIS JANTZEN for some irreducible θ. Then, π δ([ν a1 ρ, ν a2 ρ]) θ δ([ν a1 ρ, ν a2 ρ]) ν x ρ ν x 1 ρ ν a1 ρ θ = ν x ρ ν x 1 ρ ν a1 ρ δ([ν a1 ρ, ν a2 ρ]) θ. However, by Frobenius reciprocity, this contradicts the square-integrability of π. Thus we cannot have τ h contributing the ν a1 ρ. It follows immediately that we must have µ δ([ν a 1 ρ,ν a 2 ρ]) (π) =c ρ, ν δ([ν a1 a2 ρ]) θ for some c 2. On the other hand, from properties of the Jacquet functors, we know that there is an irreducible θ such that 1. µ (π) δ([ν a1 ρ, ν a2 ρ]) θ, 2. s app (θ ) δ([ν b1 ρ, ν a1 ρ]) δ([ν b3 ρ, ν a3 ρ]) δ([ν b k ρ]) σ. From the above calculations, it follows that θ = θ. We now claim δ 0 (θ) = δ([ν b1 ρ, ν a1 ρ]) δ([ν b3 ρ, ν a3 ρ]) δ([ν b k ρ]) σ. As the argument is similar to the proof of Theorem 4.4 (but easier), we will be somewhat brief. First, it follows easily from θ = θ that t.e.(δ 0 (θ)) = t.e.(δ([ν b1 ρ, ν a1 ρ]) δ([ν b3 ρ, ν a3 ρ]) δ([ν b k ρ]) σ). Suppose δ 0 (θ) =δ([ν d1 ρ, ν c1 ρ]) δ([ν d2 ρ, ν c2 ρ]) δ([ν d h ρ, ν c h ρ]) σ is lower than δ([ν b1 ρ, ν a1 ρ]) δ([ν b3 ρ, ν a3 ρ]) δ([ν b k ρ]) σ lexicographically. Observe that δ([ν a1 ρ, ν a2 ρ]) δ([ν d1 ρ, ν c1 ρ]) is either irreducible or is equal to δ([ν a1 ρ, ν c1 ρ]) δ([ν d1 ρ, ν a2 ρ]) + L(δ([ν d1 ρ, ν c1 ρ]),δ([ν a1 ρ, ν a2 ρ])). In any case, we get either π δ([ν a1 ρ, ν a2 ρ]) δ([ν d1 ρ, ν c1 ρ]) δ([ν d2 ρ, ν c2 ρ])... (Lemma 5.5, [Jan2]) π δ([ν d1 ρ, ν c1 ρ]) δ([ν a1 ρ, ν a2 ρ]) δ([ν d2 ρ, ν c2 ρ])... or δ([ν a1 ρ, ν c1 ρ]) δ([ν d1 ρ, ν a2 ρ]) δ([ν d2 ρ, ν c2 ρ])... Whichever holds, if [ν d1 ρ, ν c1 ρ] is lexicographically lower than [ν b1 ρ, ν a1 ρ]weget a contradiction to the minimality of δ 0 (π). Therefore, d 1 = b 1 and c 1 = a 1. But, in that case, π δ([ν a1 ρ, ν a2 ρ]) δ([ν b1 ρ, ν a1 ρ]) δ([ν d2 ρ, ν c2 ρ])... = δ([ν b1 ρ, ν a1 ρ]) δ([ν a1 ρ, ν a2 ρ]) δ([ν d2 ρ, ν c2 ρ])..., so δ([ν d2 ρ, ν c2 ρ]) δ([ν d h ρ, ν c h ρ]) σ cannot be lexicographically lower than δ([ν b3 ρ, ν a3 ρ]) δ([ν b k ρ]) σ. The claim follows. By Theorem 4.2.1, [Jan4], θ is square-integrable; by the inductive hypothesis, the conditions of Theorem 1.1 must be satisfied. Since b 3 <b 2 a 1 (from the assumption l =0), we can easily conclude that a k > >a 3 >a 1 >b 1 >b 3 > >b k.wearenow ready to prove the following:

21 SQUARE-INTEGRABLE REPRESENTATIONS OF CLASSICAL p-adic GROUPS II 147 Lemma 5.2. a 1 >b 2. Proof. Suppose b 2 = a 1.Observethat π δ([ν b1 ρ, ν a1 ρ]) δ([ν a1 ρ, ν a2 ρ]) δ([ν b3 ρ, ν a3 ρ],...,[ν b k = δ([ν a1 ρ, ν a2 ρ]) δ([ν b1 ρ, ν a1 ρ]) δ([ν b3 ρ, ν a3 ρ],...,[ν b k (Lemma 5.5, [Jan2]) π δ([ν a1 ρ, ν a2 ρ]) θ for some irreducible θ δ([ν b1 ρ, ν a1 ρ]) δ([ν b3 ρ, ν a3 ρ],...,[ν b k By Theorem 7.7 (and the inductive hypothesis), we know the possibilities for δ 0 (θ). Since s app (π) δ([ν a1 ρ, ν a2 ρ]) δ 0 (θ), we see that only θ = δ([ν b1 ρ, ν a1 ρ], [ν b3 ρ, ν a3 ρ],...,[ν b k does not contradict the minimality of δ 0 (π). Note that if k>2, we have t = t automatically. (If k = 2, we could argue that t = t, but it is not needed.) Since b 1 >b 3, we may now define π t = ρ, ν δ([ν a1 a2 ρ]) δ([ν b1 ρ, ν a1 ρ], [ν b3 ρ, ν a3 ρ],...,[ν b k, π t = ρ, ν δ([ν a1 a1 ρ]) δ([ν b1 ρ, ν a2 ρ], [ν b3 ρ, ν a3 ρ],...,[ν b k, πt = ρ, ν δ([ν a1 a1 ρ]) δ([ν a1+1 ρ, ν a2 ρ]) δ([ν b1 ρ, ν a1 ρ], [ν b3 ρ, ν a3 ρ],...,[ν b k. By Corollary 7.8 (and the inductive hypothesis), π t,π t π t. By Lemma 2.6 (and Corollary 7.8) µ δ([ν a 1 +1 ρ,ν a 2 ρ]) (π t )=µ δ([ν a 1 +1 ρ,ν a 2 ρ]) (π t )=µ δ([ν a 1 +1 ρ,ν a 2 ρ]) (π t ) = δ([ν a1+1 ρ, ν a2 ρ]) δ([ν a1 ρ, ν a1 ρ]) δ([ν b1 ρ, ν a1 ρ], [ν b3 ρ, ν a3 ρ],...,[ν b k. Now, µ δ([ν a 1 ρ,ν a 2 ρ]) (π) 0 implies µ δ([ν a 1 +1 ρ,ν a 2 ρ]) (π) 0. Therefore, π π t ; by unitarity, π π t. However, this contradicts the square-integrability of π (or minimality of δ 0 (π)). Thus, we could not have had b 2 = a 1. We now show the following: Lemma 5.3. Assuming a 2 >a 1 >b 2 b 1, we have b 1 b 2. Proof. Suppose b 1 = b 2. By Theorem 2.4, we must have b 1 = b 2 0. As in the beginning of section 4 (the argument immediately preceding the statement of Lemma 4.2), if δ 0 (π) is as assumed, we have π δ([ν b1 ρ, ν a1 ρ]) δ([ν b1 ρ, ν a2 ρ], [ν b3 ρ, ν a3 ρ],...,[ν b k, for some t. By Theorem 7.7 and the inductive hypothesis, we may define π t = δ([ν b1 ρ, ν a1 ρ]) δ([ν b1 ρ, ν a2 ρ], [ν b3 ρ, ν a3 ρ],...,[ν b k, π t = δ([ν b1 ρ, ν b1 ρ]) δ([ν a1 ρ, ν a2 ρ], [ν b3 ρ, ν a3 ρ],...,[ν b k, πt = δ([ν b1+1 ρ, ν a1 ρ]) δ([ν b1 ρ, ν b1 ρ]) δ([ν b1 ρ, ν a2 ρ], [ν b3 ρ, ν a3 ρ],...,[ν b k

Every set of first-order formulas is equivalent to an independent set

Every set of first-order formulas is equivalent to an independent set Every set of first-order formulas is equivalent to an independent set May 6, 2008 Abstract A set of first-order formulas, whatever the cardinality of the set of symbols, is equivalent to an independent

Διαβάστε περισσότερα

2 Composition. Invertible Mappings

2 Composition. Invertible Mappings Arkansas Tech University MATH 4033: Elementary Modern Algebra Dr. Marcel B. Finan Composition. Invertible Mappings In this section we discuss two procedures for creating new mappings from old ones, namely,

Διαβάστε περισσότερα

EE512: Error Control Coding

EE512: Error Control Coding EE512: Error Control Coding Solution for Assignment on Finite Fields February 16, 2007 1. (a) Addition and Multiplication tables for GF (5) and GF (7) are shown in Tables 1 and 2. + 0 1 2 3 4 0 0 1 2 3

Διαβάστε περισσότερα

C.S. 430 Assignment 6, Sample Solutions

C.S. 430 Assignment 6, Sample Solutions C.S. 430 Assignment 6, Sample Solutions Paul Liu November 15, 2007 Note that these are sample solutions only; in many cases there were many acceptable answers. 1 Reynolds Problem 10.1 1.1 Normal-order

Διαβάστε περισσότερα

Ordinal Arithmetic: Addition, Multiplication, Exponentiation and Limit

Ordinal Arithmetic: Addition, Multiplication, Exponentiation and Limit Ordinal Arithmetic: Addition, Multiplication, Exponentiation and Limit Ting Zhang Stanford May 11, 2001 Stanford, 5/11/2001 1 Outline Ordinal Classification Ordinal Addition Ordinal Multiplication Ordinal

Διαβάστε περισσότερα

ORDINAL ARITHMETIC JULIAN J. SCHLÖDER

ORDINAL ARITHMETIC JULIAN J. SCHLÖDER ORDINAL ARITHMETIC JULIAN J. SCHLÖDER Abstract. We define ordinal arithmetic and show laws of Left- Monotonicity, Associativity, Distributivity, some minor related properties and the Cantor Normal Form.

Διαβάστε περισσότερα

Homework 3 Solutions

Homework 3 Solutions Homework 3 Solutions Igor Yanovsky (Math 151A TA) Problem 1: Compute the absolute error and relative error in approximations of p by p. (Use calculator!) a) p π, p 22/7; b) p π, p 3.141. Solution: For

Διαβάστε περισσότερα

Lecture 2. Soundness and completeness of propositional logic

Lecture 2. Soundness and completeness of propositional logic Lecture 2 Soundness and completeness of propositional logic February 9, 2004 1 Overview Review of natural deduction. Soundness and completeness. Semantics of propositional formulas. Soundness proof. Completeness

Διαβάστε περισσότερα

Example Sheet 3 Solutions

Example Sheet 3 Solutions Example Sheet 3 Solutions. i Regular Sturm-Liouville. ii Singular Sturm-Liouville mixed boundary conditions. iii Not Sturm-Liouville ODE is not in Sturm-Liouville form. iv Regular Sturm-Liouville note

Διαβάστε περισσότερα

Concrete Mathematics Exercises from 30 September 2016

Concrete Mathematics Exercises from 30 September 2016 Concrete Mathematics Exercises from 30 September 2016 Silvio Capobianco Exercise 1.7 Let H(n) = J(n + 1) J(n). Equation (1.8) tells us that H(2n) = 2, and H(2n+1) = J(2n+2) J(2n+1) = (2J(n+1) 1) (2J(n)+1)

Διαβάστε περισσότερα

ST5224: Advanced Statistical Theory II

ST5224: Advanced Statistical Theory II ST5224: Advanced Statistical Theory II 2014/2015: Semester II Tutorial 7 1. Let X be a sample from a population P and consider testing hypotheses H 0 : P = P 0 versus H 1 : P = P 1, where P j is a known

Διαβάστε περισσότερα

ON JACQUET MODULES OF REPRESENTATIONS OF SEGMENT TYPE

ON JACQUET MODULES OF REPRESENTATIONS OF SEGMENT TYPE ON JACQUET MODULES OF REPRESENTATIONS OF SEGMENT TYPE IVAN MATIĆ AND MARKO TADIĆ Abstract. Let G n denote either the group Sp(n, F ) or SO(2n + 1, F ) over a local nonarchimedean field F. We study representations

Διαβάστε περισσότερα

4.6 Autoregressive Moving Average Model ARMA(1,1)

4.6 Autoregressive Moving Average Model ARMA(1,1) 84 CHAPTER 4. STATIONARY TS MODELS 4.6 Autoregressive Moving Average Model ARMA(,) This section is an introduction to a wide class of models ARMA(p,q) which we will consider in more detail later in this

Διαβάστε περισσότερα

CHAPTER 25 SOLVING EQUATIONS BY ITERATIVE METHODS

CHAPTER 25 SOLVING EQUATIONS BY ITERATIVE METHODS CHAPTER 5 SOLVING EQUATIONS BY ITERATIVE METHODS EXERCISE 104 Page 8 1. Find the positive root of the equation x + 3x 5 = 0, correct to 3 significant figures, using the method of bisection. Let f(x) =

Διαβάστε περισσότερα

Finite Field Problems: Solutions

Finite Field Problems: Solutions Finite Field Problems: Solutions 1. Let f = x 2 +1 Z 11 [x] and let F = Z 11 [x]/(f), a field. Let Solution: F =11 2 = 121, so F = 121 1 = 120. The possible orders are the divisors of 120. Solution: The

Διαβάστε περισσότερα

Nowhere-zero flows Let be a digraph, Abelian group. A Γ-circulation in is a mapping : such that, where, and : tail in X, head in

Nowhere-zero flows Let be a digraph, Abelian group. A Γ-circulation in is a mapping : such that, where, and : tail in X, head in Nowhere-zero flows Let be a digraph, Abelian group. A Γ-circulation in is a mapping : such that, where, and : tail in X, head in : tail in X, head in A nowhere-zero Γ-flow is a Γ-circulation such that

Διαβάστε περισσότερα

Congruence Classes of Invertible Matrices of Order 3 over F 2

Congruence Classes of Invertible Matrices of Order 3 over F 2 International Journal of Algebra, Vol. 8, 24, no. 5, 239-246 HIKARI Ltd, www.m-hikari.com http://dx.doi.org/.2988/ija.24.422 Congruence Classes of Invertible Matrices of Order 3 over F 2 Ligong An and

Διαβάστε περισσότερα

derivation of the Laplacian from rectangular to spherical coordinates

derivation of the Laplacian from rectangular to spherical coordinates derivation of the Laplacian from rectangular to spherical coordinates swapnizzle 03-03- :5:43 We begin by recognizing the familiar conversion from rectangular to spherical coordinates (note that φ is used

Διαβάστε περισσότερα

Section 8.3 Trigonometric Equations

Section 8.3 Trigonometric Equations 99 Section 8. Trigonometric Equations Objective 1: Solve Equations Involving One Trigonometric Function. In this section and the next, we will exple how to solving equations involving trigonometric functions.

Διαβάστε περισσότερα

Uniform Convergence of Fourier Series Michael Taylor

Uniform Convergence of Fourier Series Michael Taylor Uniform Convergence of Fourier Series Michael Taylor Given f L 1 T 1 ), we consider the partial sums of the Fourier series of f: N 1) S N fθ) = ˆfk)e ikθ. k= N A calculation gives the Dirichlet formula

Διαβάστε περισσότερα

SCITECH Volume 13, Issue 2 RESEARCH ORGANISATION Published online: March 29, 2018

SCITECH Volume 13, Issue 2 RESEARCH ORGANISATION Published online: March 29, 2018 Journal of rogressive Research in Mathematics(JRM) ISSN: 2395-028 SCITECH Volume 3, Issue 2 RESEARCH ORGANISATION ublished online: March 29, 208 Journal of rogressive Research in Mathematics www.scitecresearch.com/journals

Διαβάστε περισσότερα

3.4 SUM AND DIFFERENCE FORMULAS. NOTE: cos(α+β) cos α + cos β cos(α-β) cos α -cos β

3.4 SUM AND DIFFERENCE FORMULAS. NOTE: cos(α+β) cos α + cos β cos(α-β) cos α -cos β 3.4 SUM AND DIFFERENCE FORMULAS Page Theorem cos(αβ cos α cos β -sin α cos(α-β cos α cos β sin α NOTE: cos(αβ cos α cos β cos(α-β cos α -cos β Proof of cos(α-β cos α cos β sin α Let s use a unit circle

Διαβάστε περισσότερα

Fractional Colorings and Zykov Products of graphs

Fractional Colorings and Zykov Products of graphs Fractional Colorings and Zykov Products of graphs Who? Nichole Schimanski When? July 27, 2011 Graphs A graph, G, consists of a vertex set, V (G), and an edge set, E(G). V (G) is any finite set E(G) is

Διαβάστε περισσότερα

Lecture 2: Dirac notation and a review of linear algebra Read Sakurai chapter 1, Baym chatper 3

Lecture 2: Dirac notation and a review of linear algebra Read Sakurai chapter 1, Baym chatper 3 Lecture 2: Dirac notation and a review of linear algebra Read Sakurai chapter 1, Baym chatper 3 1 State vector space and the dual space Space of wavefunctions The space of wavefunctions is the set of all

Διαβάστε περισσότερα

Problem Set 3: Solutions

Problem Set 3: Solutions CMPSCI 69GG Applied Information Theory Fall 006 Problem Set 3: Solutions. [Cover and Thomas 7.] a Define the following notation, C I p xx; Y max X; Y C I p xx; Ỹ max I X; Ỹ We would like to show that C

Διαβάστε περισσότερα

Phys460.nb Solution for the t-dependent Schrodinger s equation How did we find the solution? (not required)

Phys460.nb Solution for the t-dependent Schrodinger s equation How did we find the solution? (not required) Phys460.nb 81 ψ n (t) is still the (same) eigenstate of H But for tdependent H. The answer is NO. 5.5.5. Solution for the tdependent Schrodinger s equation If we assume that at time t 0, the electron starts

Διαβάστε περισσότερα

SCHOOL OF MATHEMATICAL SCIENCES G11LMA Linear Mathematics Examination Solutions

SCHOOL OF MATHEMATICAL SCIENCES G11LMA Linear Mathematics Examination Solutions SCHOOL OF MATHEMATICAL SCIENCES GLMA Linear Mathematics 00- Examination Solutions. (a) i. ( + 5i)( i) = (6 + 5) + (5 )i = + i. Real part is, imaginary part is. (b) ii. + 5i i ( + 5i)( + i) = ( i)( + i)

Διαβάστε περισσότερα

On Square-Integrable Representations of Classical p-adic Groups

On Square-Integrable Representations of Classical p-adic Groups Canad. J. Math. Vol. 52 3, 2000 pp. 539 581 On Square-Integrable Representations of Classical p-adic Groups Chris Jantzen Abstract. In this paper, we use Jacquet module methods to study the problem of

Διαβάστε περισσότερα

Math221: HW# 1 solutions

Math221: HW# 1 solutions Math: HW# solutions Andy Royston October, 5 7.5.7, 3 rd Ed. We have a n = b n = a = fxdx = xdx =, x cos nxdx = x sin nx n sin nxdx n = cos nx n = n n, x sin nxdx = x cos nx n + cos nxdx n cos n = + sin

Διαβάστε περισσότερα

forms This gives Remark 1. How to remember the above formulas: Substituting these into the equation we obtain with

forms This gives Remark 1. How to remember the above formulas: Substituting these into the equation we obtain with Week 03: C lassification of S econd- Order L inear Equations In last week s lectures we have illustrated how to obtain the general solutions of first order PDEs using the method of characteristics. We

Διαβάστε περισσότερα

HOMEWORK 4 = G. In order to plot the stress versus the stretch we define a normalized stretch:

HOMEWORK 4 = G. In order to plot the stress versus the stretch we define a normalized stretch: HOMEWORK 4 Problem a For the fast loading case, we want to derive the relationship between P zz and λ z. We know that the nominal stress is expressed as: P zz = ψ λ z where λ z = λ λ z. Therefore, applying

Διαβάστε περισσότερα

MINIMAL CLOSED SETS AND MAXIMAL CLOSED SETS

MINIMAL CLOSED SETS AND MAXIMAL CLOSED SETS MINIMAL CLOSED SETS AND MAXIMAL CLOSED SETS FUMIE NAKAOKA AND NOBUYUKI ODA Received 20 December 2005; Revised 28 May 2006; Accepted 6 August 2006 Some properties of minimal closed sets and maximal closed

Διαβάστε περισσότερα

6.3 Forecasting ARMA processes

6.3 Forecasting ARMA processes 122 CHAPTER 6. ARMA MODELS 6.3 Forecasting ARMA processes The purpose of forecasting is to predict future values of a TS based on the data collected to the present. In this section we will discuss a linear

Διαβάστε περισσότερα

Chapter 6: Systems of Linear Differential. be continuous functions on the interval

Chapter 6: Systems of Linear Differential. be continuous functions on the interval Chapter 6: Systems of Linear Differential Equations Let a (t), a 2 (t),..., a nn (t), b (t), b 2 (t),..., b n (t) be continuous functions on the interval I. The system of n first-order differential equations

Διαβάστε περισσότερα

Second Order Partial Differential Equations

Second Order Partial Differential Equations Chapter 7 Second Order Partial Differential Equations 7.1 Introduction A second order linear PDE in two independent variables (x, y Ω can be written as A(x, y u x + B(x, y u xy + C(x, y u u u + D(x, y

Διαβάστε περισσότερα

A Note on Intuitionistic Fuzzy. Equivalence Relation

A Note on Intuitionistic Fuzzy. Equivalence Relation International Mathematical Forum, 5, 2010, no. 67, 3301-3307 A Note on Intuitionistic Fuzzy Equivalence Relation D. K. Basnet Dept. of Mathematics, Assam University Silchar-788011, Assam, India dkbasnet@rediffmail.com

Διαβάστε περισσότερα

SOME PROPERTIES OF FUZZY REAL NUMBERS

SOME PROPERTIES OF FUZZY REAL NUMBERS Sahand Communications in Mathematical Analysis (SCMA) Vol. 3 No. 1 (2016), 21-27 http://scma.maragheh.ac.ir SOME PROPERTIES OF FUZZY REAL NUMBERS BAYAZ DARABY 1 AND JAVAD JAFARI 2 Abstract. In the mathematical

Διαβάστε περισσότερα

Matrices and Determinants

Matrices and Determinants Matrices and Determinants SUBJECTIVE PROBLEMS: Q 1. For what value of k do the following system of equations possess a non-trivial (i.e., not all zero) solution over the set of rationals Q? x + ky + 3z

Διαβάστε περισσότερα

Lecture 13 - Root Space Decomposition II

Lecture 13 - Root Space Decomposition II Lecture 13 - Root Space Decomposition II October 18, 2012 1 Review First let us recall the situation. Let g be a simple algebra, with maximal toral subalgebra h (which we are calling a CSA, or Cartan Subalgebra).

Διαβάστε περισσότερα

6.1. Dirac Equation. Hamiltonian. Dirac Eq.

6.1. Dirac Equation. Hamiltonian. Dirac Eq. 6.1. Dirac Equation Ref: M.Kaku, Quantum Field Theory, Oxford Univ Press (1993) η μν = η μν = diag(1, -1, -1, -1) p 0 = p 0 p = p i = -p i p μ p μ = p 0 p 0 + p i p i = E c 2 - p 2 = (m c) 2 H = c p 2

Διαβάστε περισσότερα

Pacific Journal of Mathematics

Pacific Journal of Mathematics Pacific Journal of Mathematics REDUCIBILITY OF STANDARD REPRESENTATIONS GORAN MUIĆ Volume 222 No. 1 November 2005 PACIFIC JOURNAL OF MATHEMATICS Vol. 222, No. 1, 2005 REDUCIBILITY OF STANDARD REPRESENTATIONS

Διαβάστε περισσότερα

Solutions to Exercise Sheet 5

Solutions to Exercise Sheet 5 Solutions to Eercise Sheet 5 jacques@ucsd.edu. Let X and Y be random variables with joint pdf f(, y) = 3y( + y) where and y. Determine each of the following probabilities. Solutions. a. P (X ). b. P (X

Διαβάστε περισσότερα

On a four-dimensional hyperbolic manifold with finite volume

On a four-dimensional hyperbolic manifold with finite volume BULETINUL ACADEMIEI DE ŞTIINŢE A REPUBLICII MOLDOVA. MATEMATICA Numbers 2(72) 3(73), 2013, Pages 80 89 ISSN 1024 7696 On a four-dimensional hyperbolic manifold with finite volume I.S.Gutsul Abstract. In

Διαβάστε περισσότερα

Srednicki Chapter 55

Srednicki Chapter 55 Srednicki Chapter 55 QFT Problems & Solutions A. George August 3, 03 Srednicki 55.. Use equations 55.3-55.0 and A i, A j ] = Π i, Π j ] = 0 (at equal times) to verify equations 55.-55.3. This is our third

Διαβάστε περισσότερα

Other Test Constructions: Likelihood Ratio & Bayes Tests

Other Test Constructions: Likelihood Ratio & Bayes Tests Other Test Constructions: Likelihood Ratio & Bayes Tests Side-Note: So far we have seen a few approaches for creating tests such as Neyman-Pearson Lemma ( most powerful tests of H 0 : θ = θ 0 vs H 1 :

Διαβάστε περισσότερα

Approximation of distance between locations on earth given by latitude and longitude

Approximation of distance between locations on earth given by latitude and longitude Approximation of distance between locations on earth given by latitude and longitude Jan Behrens 2012-12-31 In this paper we shall provide a method to approximate distances between two points on earth

Διαβάστε περισσότερα

Απόκριση σε Μοναδιαία Ωστική Δύναμη (Unit Impulse) Απόκριση σε Δυνάμεις Αυθαίρετα Μεταβαλλόμενες με το Χρόνο. Απόστολος Σ.

Απόκριση σε Μοναδιαία Ωστική Δύναμη (Unit Impulse) Απόκριση σε Δυνάμεις Αυθαίρετα Μεταβαλλόμενες με το Χρόνο. Απόστολος Σ. Απόκριση σε Δυνάμεις Αυθαίρετα Μεταβαλλόμενες με το Χρόνο The time integral of a force is referred to as impulse, is determined by and is obtained from: Newton s 2 nd Law of motion states that the action

Διαβάστε περισσότερα

Bounding Nonsplitting Enumeration Degrees

Bounding Nonsplitting Enumeration Degrees Bounding Nonsplitting Enumeration Degrees Thomas F. Kent Andrea Sorbi Università degli Studi di Siena Italia July 18, 2007 Goal: Introduce a form of Σ 0 2-permitting for the enumeration degrees. Till now,

Διαβάστε περισσότερα

Lecture 15 - Root System Axiomatics

Lecture 15 - Root System Axiomatics Lecture 15 - Root System Axiomatics Nov 1, 01 In this lecture we examine root systems from an axiomatic point of view. 1 Reflections If v R n, then it determines a hyperplane, denoted P v, through the

Διαβάστε περισσότερα

Econ 2110: Fall 2008 Suggested Solutions to Problem Set 8 questions or comments to Dan Fetter 1

Econ 2110: Fall 2008 Suggested Solutions to Problem Set 8  questions or comments to Dan Fetter 1 Eon : Fall 8 Suggested Solutions to Problem Set 8 Email questions or omments to Dan Fetter Problem. Let X be a salar with density f(x, θ) (θx + θ) [ x ] with θ. (a) Find the most powerful level α test

Διαβάστε περισσότερα

Statistical Inference I Locally most powerful tests

Statistical Inference I Locally most powerful tests Statistical Inference I Locally most powerful tests Shirsendu Mukherjee Department of Statistics, Asutosh College, Kolkata, India. shirsendu st@yahoo.co.in So far we have treated the testing of one-sided

Διαβάστε περισσότερα

Partial Differential Equations in Biology The boundary element method. March 26, 2013

Partial Differential Equations in Biology The boundary element method. March 26, 2013 The boundary element method March 26, 203 Introduction and notation The problem: u = f in D R d u = ϕ in Γ D u n = g on Γ N, where D = Γ D Γ N, Γ D Γ N = (possibly, Γ D = [Neumann problem] or Γ N = [Dirichlet

Διαβάστε περισσότερα

The Simply Typed Lambda Calculus

The Simply Typed Lambda Calculus Type Inference Instead of writing type annotations, can we use an algorithm to infer what the type annotations should be? That depends on the type system. For simple type systems the answer is yes, and

Διαβάστε περισσότερα

Jesse Maassen and Mark Lundstrom Purdue University November 25, 2013

Jesse Maassen and Mark Lundstrom Purdue University November 25, 2013 Notes on Average Scattering imes and Hall Factors Jesse Maassen and Mar Lundstrom Purdue University November 5, 13 I. Introduction 1 II. Solution of the BE 1 III. Exercises: Woring out average scattering

Διαβάστε περισσότερα

9.09. # 1. Area inside the oval limaçon r = cos θ. To graph, start with θ = 0 so r = 6. Compute dr

9.09. # 1. Area inside the oval limaçon r = cos θ. To graph, start with θ = 0 so r = 6. Compute dr 9.9 #. Area inside the oval limaçon r = + cos. To graph, start with = so r =. Compute d = sin. Interesting points are where d vanishes, or at =,,, etc. For these values of we compute r:,,, and the values

Διαβάστε περισσότερα

Inverse trigonometric functions & General Solution of Trigonometric Equations. ------------------ ----------------------------- -----------------

Inverse trigonometric functions & General Solution of Trigonometric Equations. ------------------ ----------------------------- ----------------- Inverse trigonometric functions & General Solution of Trigonometric Equations. 1. Sin ( ) = a) b) c) d) Ans b. Solution : Method 1. Ans a: 17 > 1 a) is rejected. w.k.t Sin ( sin ) = d is rejected. If sin

Διαβάστε περισσότερα

Tridiagonal matrices. Gérard MEURANT. October, 2008

Tridiagonal matrices. Gérard MEURANT. October, 2008 Tridiagonal matrices Gérard MEURANT October, 2008 1 Similarity 2 Cholesy factorizations 3 Eigenvalues 4 Inverse Similarity Let α 1 ω 1 β 1 α 2 ω 2 T =......... β 2 α 1 ω 1 β 1 α and β i ω i, i = 1,...,

Διαβάστε περισσότερα

Tempered representations for classical p-adic groups

Tempered representations for classical p-adic groups manuscripta math. 145, 319 387 014) Springer-Verlag Berlin Heidelberg 014 Chris Jantzen Tempered representations for classical p-adic groups Received: 30 December 01 / Revised: 1 March 014 Published online:

Διαβάστε περισσότερα

2. Let H 1 and H 2 be Hilbert spaces and let T : H 1 H 2 be a bounded linear operator. Prove that [T (H 1 )] = N (T ). (6p)

2. Let H 1 and H 2 be Hilbert spaces and let T : H 1 H 2 be a bounded linear operator. Prove that [T (H 1 )] = N (T ). (6p) Uppsala Universitet Matematiska Institutionen Andreas Strömbergsson Prov i matematik Funktionalanalys Kurs: F3B, F4Sy, NVP 2005-03-08 Skrivtid: 9 14 Tillåtna hjälpmedel: Manuella skrivdon, Kreyszigs bok

Διαβάστε περισσότερα

Συστήματα Διαχείρισης Βάσεων Δεδομένων

Συστήματα Διαχείρισης Βάσεων Δεδομένων ΕΛΛΗΝΙΚΗ ΔΗΜΟΚΡΑΤΙΑ ΠΑΝΕΠΙΣΤΗΜΙΟ ΚΡΗΤΗΣ Συστήματα Διαχείρισης Βάσεων Δεδομένων Φροντιστήριο 9: Transactions - part 1 Δημήτρης Πλεξουσάκης Τμήμα Επιστήμης Υπολογιστών Tutorial on Undo, Redo and Undo/Redo

Διαβάστε περισσότερα

ω ω ω ω ω ω+2 ω ω+2 + ω ω ω ω+2 + ω ω+1 ω ω+2 2 ω ω ω ω ω ω ω ω+1 ω ω2 ω ω2 + ω ω ω2 + ω ω ω ω2 + ω ω+1 ω ω2 + ω ω+1 + ω ω ω ω2 + ω

ω ω ω ω ω ω+2 ω ω+2 + ω ω ω ω+2 + ω ω+1 ω ω+2 2 ω ω ω ω ω ω ω ω+1 ω ω2 ω ω2 + ω ω ω2 + ω ω ω ω2 + ω ω+1 ω ω2 + ω ω+1 + ω ω ω ω2 + ω 0 1 2 3 4 5 6 ω ω + 1 ω + 2 ω + 3 ω + 4 ω2 ω2 + 1 ω2 + 2 ω2 + 3 ω3 ω3 + 1 ω3 + 2 ω4 ω4 + 1 ω5 ω 2 ω 2 + 1 ω 2 + 2 ω 2 + ω ω 2 + ω + 1 ω 2 + ω2 ω 2 2 ω 2 2 + 1 ω 2 2 + ω ω 2 3 ω 3 ω 3 + 1 ω 3 + ω ω 3 +

Διαβάστε περισσότερα

Math 446 Homework 3 Solutions. (1). (i): Reverse triangle inequality for metrics: Let (X, d) be a metric space and let x, y, z X.

Math 446 Homework 3 Solutions. (1). (i): Reverse triangle inequality for metrics: Let (X, d) be a metric space and let x, y, z X. Math 446 Homework 3 Solutions. (1). (i): Reverse triangle inequalit for metrics: Let (X, d) be a metric space and let x,, z X. Prove that d(x, z) d(, z) d(x, ). (ii): Reverse triangle inequalit for norms:

Διαβάστε περισσότερα

TMA4115 Matematikk 3

TMA4115 Matematikk 3 TMA4115 Matematikk 3 Andrew Stacey Norges Teknisk-Naturvitenskapelige Universitet Trondheim Spring 2010 Lecture 12: Mathematics Marvellous Matrices Andrew Stacey Norges Teknisk-Naturvitenskapelige Universitet

Διαβάστε περισσότερα

CRASH COURSE IN PRECALCULUS

CRASH COURSE IN PRECALCULUS CRASH COURSE IN PRECALCULUS Shiah-Sen Wang The graphs are prepared by Chien-Lun Lai Based on : Precalculus: Mathematics for Calculus by J. Stuwart, L. Redin & S. Watson, 6th edition, 01, Brooks/Cole Chapter

Διαβάστε περισσότερα

Section 9.2 Polar Equations and Graphs

Section 9.2 Polar Equations and Graphs 180 Section 9. Polar Equations and Graphs In this section, we will be graphing polar equations on a polar grid. In the first few examples, we will write the polar equation in rectangular form to help identify

Διαβάστε περισσότερα

( y) Partial Differential Equations

( y) Partial Differential Equations Partial Dierential Equations Linear P.D.Es. contains no owers roducts o the deendent variables / an o its derivatives can occasionall be solved. Consider eamle ( ) a (sometimes written as a ) we can integrate

Διαβάστε περισσότερα

ANSWERSHEET (TOPIC = DIFFERENTIAL CALCULUS) COLLECTION #2. h 0 h h 0 h h 0 ( ) g k = g 0 + g 1 + g g 2009 =?

ANSWERSHEET (TOPIC = DIFFERENTIAL CALCULUS) COLLECTION #2. h 0 h h 0 h h 0 ( ) g k = g 0 + g 1 + g g 2009 =? Teko Classes IITJEE/AIEEE Maths by SUHAAG SIR, Bhopal, Ph (0755) 3 00 000 www.tekoclasses.com ANSWERSHEET (TOPIC DIFFERENTIAL CALCULUS) COLLECTION # Question Type A.Single Correct Type Q. (A) Sol least

Διαβάστε περισσότερα

Homomorphism in Intuitionistic Fuzzy Automata

Homomorphism in Intuitionistic Fuzzy Automata International Journal of Fuzzy Mathematics Systems. ISSN 2248-9940 Volume 3, Number 1 (2013), pp. 39-45 Research India Publications http://www.ripublication.com/ijfms.htm Homomorphism in Intuitionistic

Διαβάστε περισσότερα

Chapter 3: Ordinal Numbers

Chapter 3: Ordinal Numbers Chapter 3: Ordinal Numbers There are two kinds of number.. Ordinal numbers (0th), st, 2nd, 3rd, 4th, 5th,..., ω, ω +,... ω2, ω2+,... ω 2... answers to the question What position is... in a sequence? What

Διαβάστε περισσότερα

Exercises 10. Find a fundamental matrix of the given system of equations. Also find the fundamental matrix Φ(t) satisfying Φ(0) = I. 1.

Exercises 10. Find a fundamental matrix of the given system of equations. Also find the fundamental matrix Φ(t) satisfying Φ(0) = I. 1. Exercises 0 More exercises are available in Elementary Differential Equations. If you have a problem to solve any of them, feel free to come to office hour. Problem Find a fundamental matrix of the given

Διαβάστε περισσότερα

Areas and Lengths in Polar Coordinates

Areas and Lengths in Polar Coordinates Kiryl Tsishchanka Areas and Lengths in Polar Coordinates In this section we develop the formula for the area of a region whose boundary is given by a polar equation. We need to use the formula for the

Διαβάστε περισσότερα

SOLVING CUBICS AND QUARTICS BY RADICALS

SOLVING CUBICS AND QUARTICS BY RADICALS SOLVING CUBICS AND QUARTICS BY RADICALS The purpose of this handout is to record the classical formulas expressing the roots of degree three and degree four polynomials in terms of radicals. We begin with

Διαβάστε περισσότερα

PARTIAL NOTES for 6.1 Trigonometric Identities

PARTIAL NOTES for 6.1 Trigonometric Identities PARTIAL NOTES for 6.1 Trigonometric Identities tanθ = sinθ cosθ cotθ = cosθ sinθ BASIC IDENTITIES cscθ = 1 sinθ secθ = 1 cosθ cotθ = 1 tanθ PYTHAGOREAN IDENTITIES sin θ + cos θ =1 tan θ +1= sec θ 1 + cot

Διαβάστε περισσότερα

Section 7.6 Double and Half Angle Formulas

Section 7.6 Double and Half Angle Formulas 09 Section 7. Double and Half Angle Fmulas To derive the double-angles fmulas, we will use the sum of two angles fmulas that we developed in the last section. We will let α θ and β θ: cos(θ) cos(θ + θ)

Διαβάστε περισσότερα

Main source: "Discrete-time systems and computer control" by Α. ΣΚΟΔΡΑΣ ΨΗΦΙΑΚΟΣ ΕΛΕΓΧΟΣ ΔΙΑΛΕΞΗ 4 ΔΙΑΦΑΝΕΙΑ 1

Main source: Discrete-time systems and computer control by Α. ΣΚΟΔΡΑΣ ΨΗΦΙΑΚΟΣ ΕΛΕΓΧΟΣ ΔΙΑΛΕΞΗ 4 ΔΙΑΦΑΝΕΙΑ 1 Main source: "Discrete-time systems and computer control" by Α. ΣΚΟΔΡΑΣ ΨΗΦΙΑΚΟΣ ΕΛΕΓΧΟΣ ΔΙΑΛΕΞΗ 4 ΔΙΑΦΑΝΕΙΑ 1 A Brief History of Sampling Research 1915 - Edmund Taylor Whittaker (1873-1956) devised a

Διαβάστε περισσότερα

Coefficient Inequalities for a New Subclass of K-uniformly Convex Functions

Coefficient Inequalities for a New Subclass of K-uniformly Convex Functions International Journal of Computational Science and Mathematics. ISSN 0974-89 Volume, Number (00), pp. 67--75 International Research Publication House http://www.irphouse.com Coefficient Inequalities for

Διαβάστε περισσότερα

New bounds for spherical two-distance sets and equiangular lines

New bounds for spherical two-distance sets and equiangular lines New bounds for spherical two-distance sets and equiangular lines Michigan State University Oct 8-31, 016 Anhui University Definition If X = {x 1, x,, x N } S n 1 (unit sphere in R n ) and x i, x j = a

Διαβάστε περισσότερα

Solution Series 9. i=1 x i and i=1 x i.

Solution Series 9. i=1 x i and i=1 x i. Lecturer: Prof. Dr. Mete SONER Coordinator: Yilin WANG Solution Series 9 Q1. Let α, β >, the p.d.f. of a beta distribution with parameters α and β is { Γ(α+β) Γ(α)Γ(β) f(x α, β) xα 1 (1 x) β 1 for < x

Διαβάστε περισσότερα

Chapter 6: Systems of Linear Differential. be continuous functions on the interval

Chapter 6: Systems of Linear Differential. be continuous functions on the interval Chapter 6: Systems of Linear Differential Equations Let a (t), a 2 (t),..., a nn (t), b (t), b 2 (t),..., b n (t) be continuous functions on the interval I. The system of n first-order differential equations

Διαβάστε περισσότερα

ΚΥΠΡΙΑΚΗ ΕΤΑΙΡΕΙΑ ΠΛΗΡΟΦΟΡΙΚΗΣ CYPRUS COMPUTER SOCIETY ΠΑΓΚΥΠΡΙΟΣ ΜΑΘΗΤΙΚΟΣ ΔΙΑΓΩΝΙΣΜΟΣ ΠΛΗΡΟΦΟΡΙΚΗΣ 19/5/2007

ΚΥΠΡΙΑΚΗ ΕΤΑΙΡΕΙΑ ΠΛΗΡΟΦΟΡΙΚΗΣ CYPRUS COMPUTER SOCIETY ΠΑΓΚΥΠΡΙΟΣ ΜΑΘΗΤΙΚΟΣ ΔΙΑΓΩΝΙΣΜΟΣ ΠΛΗΡΟΦΟΡΙΚΗΣ 19/5/2007 Οδηγίες: Να απαντηθούν όλες οι ερωτήσεις. Αν κάπου κάνετε κάποιες υποθέσεις να αναφερθούν στη σχετική ερώτηση. Όλα τα αρχεία που αναφέρονται στα προβλήματα βρίσκονται στον ίδιο φάκελο με το εκτελέσιμο

Διαβάστε περισσότερα

Μηχανική Μάθηση Hypothesis Testing

Μηχανική Μάθηση Hypothesis Testing ΕΛΛΗΝΙΚΗ ΔΗΜΟΚΡΑΤΙΑ ΠΑΝΕΠΙΣΤΗΜΙΟ ΚΡΗΤΗΣ Μηχανική Μάθηση Hypothesis Testing Γιώργος Μπορμπουδάκης Τμήμα Επιστήμης Υπολογιστών Procedure 1. Form the null (H 0 ) and alternative (H 1 ) hypothesis 2. Consider

Διαβάστε περισσότερα

1. Introduction and Preliminaries.

1. Introduction and Preliminaries. Faculty of Sciences and Mathematics, University of Niš, Serbia Available at: http://www.pmf.ni.ac.yu/filomat Filomat 22:1 (2008), 97 106 ON δ SETS IN γ SPACES V. Renuka Devi and D. Sivaraj Abstract We

Διαβάστε περισσότερα

Reminders: linear functions

Reminders: linear functions Reminders: linear functions Let U and V be vector spaces over the same field F. Definition A function f : U V is linear if for every u 1, u 2 U, f (u 1 + u 2 ) = f (u 1 ) + f (u 2 ), and for every u U

Διαβάστε περισσότερα

ΕΛΛΗΝΙΚΗ ΔΗΜΟΚΡΑΤΙΑ ΠΑΝΕΠΙΣΤΗΜΙΟ ΚΡΗΤΗΣ. Ψηφιακή Οικονομία. Διάλεξη 7η: Consumer Behavior Mαρίνα Μπιτσάκη Τμήμα Επιστήμης Υπολογιστών

ΕΛΛΗΝΙΚΗ ΔΗΜΟΚΡΑΤΙΑ ΠΑΝΕΠΙΣΤΗΜΙΟ ΚΡΗΤΗΣ. Ψηφιακή Οικονομία. Διάλεξη 7η: Consumer Behavior Mαρίνα Μπιτσάκη Τμήμα Επιστήμης Υπολογιστών ΕΛΛΗΝΙΚΗ ΔΗΜΟΚΡΑΤΙΑ ΠΑΝΕΠΙΣΤΗΜΙΟ ΚΡΗΤΗΣ Ψηφιακή Οικονομία Διάλεξη 7η: Consumer Behavior Mαρίνα Μπιτσάκη Τμήμα Επιστήμης Υπολογιστών Τέλος Ενότητας Χρηματοδότηση Το παρόν εκπαιδευτικό υλικό έχει αναπτυχθεί

Διαβάστε περισσότερα

= λ 1 1 e. = λ 1 =12. has the properties e 1. e 3,V(Y

= λ 1 1 e. = λ 1 =12. has the properties e 1. e 3,V(Y Stat 50 Homework Solutions Spring 005. (a λ λ λ 44 (b trace( λ + λ + λ 0 (c V (e x e e λ e e λ e (λ e by definition, the eigenvector e has the properties e λ e and e e. (d λ e e + λ e e + λ e e 8 6 4 4

Διαβάστε περισσότερα

The challenges of non-stable predicates

The challenges of non-stable predicates The challenges of non-stable predicates Consider a non-stable predicate Φ encoding, say, a safety property. We want to determine whether Φ holds for our program. The challenges of non-stable predicates

Διαβάστε περισσότερα

Space-Time Symmetries

Space-Time Symmetries Chapter Space-Time Symmetries In classical fiel theory any continuous symmetry of the action generates a conserve current by Noether's proceure. If the Lagrangian is not invariant but only shifts by a

Διαβάστε περισσότερα

Intuitionistic Fuzzy Ideals of Near Rings

Intuitionistic Fuzzy Ideals of Near Rings International Mathematical Forum, Vol. 7, 202, no. 6, 769-776 Intuitionistic Fuzzy Ideals of Near Rings P. K. Sharma P.G. Department of Mathematics D.A.V. College Jalandhar city, Punjab, India pksharma@davjalandhar.com

Διαβάστε περισσότερα

Strain gauge and rosettes

Strain gauge and rosettes Strain gauge and rosettes Introduction A strain gauge is a device which is used to measure strain (deformation) on an object subjected to forces. Strain can be measured using various types of devices classified

Διαβάστε περισσότερα

LECTURE 2 CONTEXT FREE GRAMMARS CONTENTS

LECTURE 2 CONTEXT FREE GRAMMARS CONTENTS LECTURE 2 CONTEXT FREE GRAMMARS CONTENTS 1. Developing a grammar fragment...1 2. A formalism that is too strong and too weak at the same time...3 3. References...4 1. Developing a grammar fragment The

Διαβάστε περισσότερα

Homomorphism of Intuitionistic Fuzzy Groups

Homomorphism of Intuitionistic Fuzzy Groups International Mathematical Forum, Vol. 6, 20, no. 64, 369-378 Homomorphism o Intuitionistic Fuzz Groups P. K. Sharma Department o Mathematics, D..V. College Jalandhar Cit, Punjab, India pksharma@davjalandhar.com

Διαβάστε περισσότερα

ΚΥΠΡΙΑΚΗ ΕΤΑΙΡΕΙΑ ΠΛΗΡΟΦΟΡΙΚΗΣ CYPRUS COMPUTER SOCIETY ΠΑΓΚΥΠΡΙΟΣ ΜΑΘΗΤΙΚΟΣ ΔΙΑΓΩΝΙΣΜΟΣ ΠΛΗΡΟΦΟΡΙΚΗΣ 6/5/2006

ΚΥΠΡΙΑΚΗ ΕΤΑΙΡΕΙΑ ΠΛΗΡΟΦΟΡΙΚΗΣ CYPRUS COMPUTER SOCIETY ΠΑΓΚΥΠΡΙΟΣ ΜΑΘΗΤΙΚΟΣ ΔΙΑΓΩΝΙΣΜΟΣ ΠΛΗΡΟΦΟΡΙΚΗΣ 6/5/2006 Οδηγίες: Να απαντηθούν όλες οι ερωτήσεις. Ολοι οι αριθμοί που αναφέρονται σε όλα τα ερωτήματα είναι μικρότεροι το 1000 εκτός αν ορίζεται διαφορετικά στη διατύπωση του προβλήματος. Διάρκεια: 3,5 ώρες Καλή

Διαβάστε περισσότερα

Second Order RLC Filters

Second Order RLC Filters ECEN 60 Circuits/Electronics Spring 007-0-07 P. Mathys Second Order RLC Filters RLC Lowpass Filter A passive RLC lowpass filter (LPF) circuit is shown in the following schematic. R L C v O (t) Using phasor

Διαβάστε περισσότερα

Abstract Storage Devices

Abstract Storage Devices Abstract Storage Devices Robert König Ueli Maurer Stefano Tessaro SOFSEM 2009 January 27, 2009 Outline 1. Motivation: Storage Devices 2. Abstract Storage Devices (ASD s) 3. Reducibility 4. Factoring ASD

Διαβάστε περισσότερα

Quadratic Expressions

Quadratic Expressions Quadratic Expressions. The standard form of a quadratic equation is ax + bx + c = 0 where a, b, c R and a 0. The roots of ax + bx + c = 0 are b ± b a 4ac. 3. For the equation ax +bx+c = 0, sum of the roots

Διαβάστε περισσότερα

5. Choice under Uncertainty

5. Choice under Uncertainty 5. Choice under Uncertainty Daisuke Oyama Microeconomics I May 23, 2018 Formulations von Neumann-Morgenstern (1944/1947) X: Set of prizes Π: Set of probability distributions on X : Preference relation

Διαβάστε περισσότερα

Areas and Lengths in Polar Coordinates

Areas and Lengths in Polar Coordinates Kiryl Tsishchanka Areas and Lengths in Polar Coordinates In this section we develop the formula for the area of a region whose boundary is given by a polar equation. We need to use the formula for the

Διαβάστε περισσότερα

k A = [k, k]( )[a 1, a 2 ] = [ka 1,ka 2 ] 4For the division of two intervals of confidence in R +

k A = [k, k]( )[a 1, a 2 ] = [ka 1,ka 2 ] 4For the division of two intervals of confidence in R + Chapter 3. Fuzzy Arithmetic 3- Fuzzy arithmetic: ~Addition(+) and subtraction (-): Let A = [a and B = [b, b in R If x [a and y [b, b than x+y [a +b +b Symbolically,we write A(+)B = [a (+)[b, b = [a +b

Διαβάστε περισσότερα

Fourier Series. MATH 211, Calculus II. J. Robert Buchanan. Spring Department of Mathematics

Fourier Series. MATH 211, Calculus II. J. Robert Buchanan. Spring Department of Mathematics Fourier Series MATH 211, Calculus II J. Robert Buchanan Department of Mathematics Spring 2018 Introduction Not all functions can be represented by Taylor series. f (k) (c) A Taylor series f (x) = (x c)

Διαβάστε περισσότερα

Lecture 21: Properties and robustness of LSE

Lecture 21: Properties and robustness of LSE Lecture 21: Properties and robustness of LSE BLUE: Robustness of LSE against normality We now study properties of l τ β and σ 2 under assumption A2, i.e., without the normality assumption on ε. From Theorem

Διαβάστε περισσότερα