# Erratum: Corrigendum to “A characterization of metacirculants” (Journal of Combinatorial Theory. Series A (2013) 1 (39–48)(S0097316512001148)(10.1016/j.jcta.2012.06.010))

Cai Heng Li, Shu Jiao Song

Research output: Contribution to journalComment/debate

2 Citations (Scopus)

### Abstract

A graph is called a weak-metacirculant if it contains a metacyclic vertex-transitive automorphism group, introduced by Marušič and Šparl [5]. It is a generalization of the concept of matecirculant (introduced in [1]), which has a metacyclic vertex-transitive automorphism group with certain restrictive condition. A natural question is whether a weak-metacirculant is actually a metacirculant, see [5]. The answer to this question is negative: an infinitely family of examples is given in [4, Theorem 1.3], which are tetravalent Cayley graphs of non-split metacyclic p-groups, with p odd prime. However, there is, unfortunately, a mistake in [4, Theorem 1.3] which claimed that such Cayley graphs could be edge-transitive. The following theorem corrects the mistake and confirms the statement that such graphs are weak-metacirculants but not metacirculants is still valid. Theorem 1 Let p be an odd prime, and let G be a non-split metacyclic p-group. Let Γ be a connected Cayley graph of G of valency at most 2(p−1). Then AutΓ=G, and Γ is not edge-transitive; in particular, Γ is a weak-metacirculant but not a metacirculant. Proof Let Γ=Cay(G,S). Then Γ is a weak-metacirculant. Let A=AutΓ, let α be the vertex of Γ corresponding to the identity of G. By [3, Corollary 1.2], A=G:Aut(G,S), where Aut(G,S)=〈σ∈Aut(G)|Sσ=S〉≤Aut(G). Since G is a non-split metacyclic p-group, Aut(G) is a p-group by [2]. Let S={s1,s1−1,s2−1,…,sk,sk−1}, where k≤p−1. Suppose that Aut(G,S)≠1. Then there exists an element σ∈Aut(G,S) of order p. As |S|=2k<2p, it implies that sjσi=sj−1 for some integers i and j, which is not possible as p is odd. Thus Aα=Aut(G,S)=1, and A=G. In particular, Γ is not edge-transitive. Since G is non-split, Γ is not a metacirculant, see [4] or [5].  □ It is remarkable that, for a non-split metacyclic p-group G with p being odd and prime, all connected Cayley graphs of G of valency at most 2(p−1) are always so-called graphical regular representations of the group G, and are not edge-transitive.

Original language English 344-345 2 Journal of Combinatorial Theory. Series A 146 https://doi.org/10.1016/j.jcta.2016.09.005 Published - 1 Feb 2017

### Fingerprint

P-groups
Cayley Graph
Odd
Series
Vertex-transitive
Theorem
Automorphism Group
Connected graph
Graph in graph theory
Corollary
Valid
Imply
Integer
Vertex of a graph

### Cite this

@article{d238fb60d4ba446a8efa3df474532b08,
title = "Erratum: Corrigendum to “A characterization of metacirculants” (Journal of Combinatorial Theory. Series A (2013) 1 (39–48)(S0097316512001148)(10.1016/j.jcta.2012.06.010))",
abstract = "A graph is called a weak-metacirculant if it contains a metacyclic vertex-transitive automorphism group, introduced by Marušič and Šparl [5]. It is a generalization of the concept of matecirculant (introduced in [1]), which has a metacyclic vertex-transitive automorphism group with certain restrictive condition. A natural question is whether a weak-metacirculant is actually a metacirculant, see [5]. The answer to this question is negative: an infinitely family of examples is given in [4, Theorem 1.3], which are tetravalent Cayley graphs of non-split metacyclic p-groups, with p odd prime. However, there is, unfortunately, a mistake in [4, Theorem 1.3] which claimed that such Cayley graphs could be edge-transitive. The following theorem corrects the mistake and confirms the statement that such graphs are weak-metacirculants but not metacirculants is still valid. Theorem 1 Let p be an odd prime, and let G be a non-split metacyclic p-group. Let Γ be a connected Cayley graph of G of valency at most 2(p−1). Then AutΓ=G, and Γ is not edge-transitive; in particular, Γ is a weak-metacirculant but not a metacirculant. Proof Let Γ=Cay(G,S). Then Γ is a weak-metacirculant. Let A=AutΓ, let α be the vertex of Γ corresponding to the identity of G. By [3, Corollary 1.2], A=G:Aut(G,S), where Aut(G,S)=〈σ∈Aut(G)|Sσ=S〉≤Aut(G). Since G is a non-split metacyclic p-group, Aut(G) is a p-group by [2]. Let S={s1,s1−1,s2−1,…,sk,sk−1}, where k≤p−1. Suppose that Aut(G,S)≠1. Then there exists an element σ∈Aut(G,S) of order p. As |S|=2k<2p, it implies that sjσi=sj−1 for some integers i and j, which is not possible as p is odd. Thus Aα=Aut(G,S)=1, and A=G. In particular, Γ is not edge-transitive. Since G is non-split, Γ is not a metacirculant, see [4] or [5].  □ It is remarkable that, for a non-split metacyclic p-group G with p being odd and prime, all connected Cayley graphs of G of valency at most 2(p−1) are always so-called graphical regular representations of the group G, and are not edge-transitive.",
author = "Li, {Cai Heng} and Song, {Shu Jiao}",
year = "2017",
month = "2",
day = "1",
doi = "10.1016/j.jcta.2016.09.005",
language = "English",
volume = "146",
pages = "344--345",
journal = "Journal of Combinatorial Theory Series A",
issn = "0021-9800",

}

In: Journal of Combinatorial Theory. Series A, Vol. 146, 01.02.2017, p. 344-345.

Research output: Contribution to journalComment/debate

TY - JOUR

T1 - Erratum

T2 - Corrigendum to “A characterization of metacirculants” (Journal of Combinatorial Theory. Series A (2013) 1 (39–48)(S0097316512001148)(10.1016/j.jcta.2012.06.010))

AU - Li, Cai Heng

AU - Song, Shu Jiao

PY - 2017/2/1

Y1 - 2017/2/1

N2 - A graph is called a weak-metacirculant if it contains a metacyclic vertex-transitive automorphism group, introduced by Marušič and Šparl [5]. It is a generalization of the concept of matecirculant (introduced in [1]), which has a metacyclic vertex-transitive automorphism group with certain restrictive condition. A natural question is whether a weak-metacirculant is actually a metacirculant, see [5]. The answer to this question is negative: an infinitely family of examples is given in [4, Theorem 1.3], which are tetravalent Cayley graphs of non-split metacyclic p-groups, with p odd prime. However, there is, unfortunately, a mistake in [4, Theorem 1.3] which claimed that such Cayley graphs could be edge-transitive. The following theorem corrects the mistake and confirms the statement that such graphs are weak-metacirculants but not metacirculants is still valid. Theorem 1 Let p be an odd prime, and let G be a non-split metacyclic p-group. Let Γ be a connected Cayley graph of G of valency at most 2(p−1). Then AutΓ=G, and Γ is not edge-transitive; in particular, Γ is a weak-metacirculant but not a metacirculant. Proof Let Γ=Cay(G,S). Then Γ is a weak-metacirculant. Let A=AutΓ, let α be the vertex of Γ corresponding to the identity of G. By [3, Corollary 1.2], A=G:Aut(G,S), where Aut(G,S)=〈σ∈Aut(G)|Sσ=S〉≤Aut(G). Since G is a non-split metacyclic p-group, Aut(G) is a p-group by [2]. Let S={s1,s1−1,s2−1,…,sk,sk−1}, where k≤p−1. Suppose that Aut(G,S)≠1. Then there exists an element σ∈Aut(G,S) of order p. As |S|=2k<2p, it implies that sjσi=sj−1 for some integers i and j, which is not possible as p is odd. Thus Aα=Aut(G,S)=1, and A=G. In particular, Γ is not edge-transitive. Since G is non-split, Γ is not a metacirculant, see [4] or [5].  □ It is remarkable that, for a non-split metacyclic p-group G with p being odd and prime, all connected Cayley graphs of G of valency at most 2(p−1) are always so-called graphical regular representations of the group G, and are not edge-transitive.

AB - A graph is called a weak-metacirculant if it contains a metacyclic vertex-transitive automorphism group, introduced by Marušič and Šparl [5]. It is a generalization of the concept of matecirculant (introduced in [1]), which has a metacyclic vertex-transitive automorphism group with certain restrictive condition. A natural question is whether a weak-metacirculant is actually a metacirculant, see [5]. The answer to this question is negative: an infinitely family of examples is given in [4, Theorem 1.3], which are tetravalent Cayley graphs of non-split metacyclic p-groups, with p odd prime. However, there is, unfortunately, a mistake in [4, Theorem 1.3] which claimed that such Cayley graphs could be edge-transitive. The following theorem corrects the mistake and confirms the statement that such graphs are weak-metacirculants but not metacirculants is still valid. Theorem 1 Let p be an odd prime, and let G be a non-split metacyclic p-group. Let Γ be a connected Cayley graph of G of valency at most 2(p−1). Then AutΓ=G, and Γ is not edge-transitive; in particular, Γ is a weak-metacirculant but not a metacirculant. Proof Let Γ=Cay(G,S). Then Γ is a weak-metacirculant. Let A=AutΓ, let α be the vertex of Γ corresponding to the identity of G. By [3, Corollary 1.2], A=G:Aut(G,S), where Aut(G,S)=〈σ∈Aut(G)|Sσ=S〉≤Aut(G). Since G is a non-split metacyclic p-group, Aut(G) is a p-group by [2]. Let S={s1,s1−1,s2−1,…,sk,sk−1}, where k≤p−1. Suppose that Aut(G,S)≠1. Then there exists an element σ∈Aut(G,S) of order p. As |S|=2k<2p, it implies that sjσi=sj−1 for some integers i and j, which is not possible as p is odd. Thus Aα=Aut(G,S)=1, and A=G. In particular, Γ is not edge-transitive. Since G is non-split, Γ is not a metacirculant, see [4] or [5].  □ It is remarkable that, for a non-split metacyclic p-group G with p being odd and prime, all connected Cayley graphs of G of valency at most 2(p−1) are always so-called graphical regular representations of the group G, and are not edge-transitive.

UR - http://www.scopus.com/inward/record.url?scp=84997327222&partnerID=8YFLogxK

U2 - 10.1016/j.jcta.2016.09.005

DO - 10.1016/j.jcta.2016.09.005

M3 - Comment/debate

VL - 146

SP - 344

EP - 345

JO - Journal of Combinatorial Theory Series A

JF - Journal of Combinatorial Theory Series A

SN - 0021-9800

ER -