Explore the full ebook collection and download it now at textbookfull.

com

Intelligent Distributed Computing X: Proceedings

of the 10th International Symposium on Intelligent
Distributed Computing – IDC 2016, Paris, France,
October 10-12 2016 1st Edition Costin Badica
https://textbookfull.com/product/intelligent-distributed-
computing-x-proceedings-of-the-10th-international-symposium-
on-intelligent-distributed-computing-idc-2016-paris-france-
october-10-12-2016-1st-edition-costin-badic/

OR CLICK HERE

DOWLOAD EBOOK

Browse and Get More Ebook Downloads Instantly at https://textbookfull.com

Click here to visit textbookfull.com and download textbook now
 Your digital treasures (PDF, ePub, MOBI) await
Download instantly and pick your perfect format...

Read anywhere, anytime, on any device!

Biota Grow 2C gather 2C cook Loucas

https://textbookfull.com/product/biota-grow-2c-gather-2c-cook-loucas/

textbookfull.com

Internet and Distributed Computing Systems 12th

International Conference IDCS 2019 Naples Italy October 10
12 2019 Proceedings Raffaele Montella
https://textbookfull.com/product/internet-and-distributed-computing-
systems-12th-international-conference-idcs-2019-naples-italy-
october-10-12-2019-proceedings-raffaele-montella/
textbookfull.com

Internet and Distributed Computing Systems 9th

International Conference IDCS 2016 Wuhan China September
28 30 2016 Proceedings 1st Edition Wenfeng Li
https://textbookfull.com/product/internet-and-distributed-computing-
systems-9th-international-conference-idcs-2016-wuhan-china-
september-28-30-2016-proceedings-1st-edition-wenfeng-li/
textbookfull.com

International Proceedings on Advances in Soft Computing

Intelligent Systems and Applications ASISA 2016 1st
Edition M. Sreenivasa Reddy
https://textbookfull.com/product/international-proceedings-on-
advances-in-soft-computing-intelligent-systems-and-applications-
asisa-2016-1st-edition-m-sreenivasa-reddy/
textbookfull.com
Progress in Advanced Computing and Intelligent Engineering
Proceedings of ICACIE 2016 Volume 2 1st Edition Khalid
Saeed
https://textbookfull.com/product/progress-in-advanced-computing-and-
intelligent-engineering-proceedings-of-icacie-2016-volume-2-1st-
edition-khalid-saeed/
textbookfull.com

Progress in Advanced Computing and Intelligent Engineering

Proceedings of ICACIE 2016 Volume 1 1st Edition Khalid
Saeed
https://textbookfull.com/product/progress-in-advanced-computing-and-
intelligent-engineering-proceedings-of-icacie-2016-volume-1-1st-
edition-khalid-saeed/
textbookfull.com

Advances in Visual Computing 12th International Symposium

ISVC 2016 Las Vegas NV USA December 12 14 2016 Proceedings
Part II 1st Edition George Bebis
https://textbookfull.com/product/advances-in-visual-computing-12th-
international-symposium-isvc-2016-las-vegas-nv-usa-
december-12-14-2016-proceedings-part-ii-1st-edition-george-bebis/
textbookfull.com

Advances in Visual Computing 12th International Symposium

ISVC 2016 Las Vegas NV USA December 12 14 2016 Proceedings
Part I 1st Edition George Bebis
https://textbookfull.com/product/advances-in-visual-computing-12th-
international-symposium-isvc-2016-las-vegas-nv-usa-
december-12-14-2016-proceedings-part-i-1st-edition-george-bebis/
textbookfull.com

Formal Techniques for Distributed Objects Components and

Systems 36th IFIP WG 6 1 International Conference FORTE
2016 Held as Part of the 11th International Federated
Conference on Distributed Computing Techniques DisCoTec
https://textbookfull.com/product/formal-techniques-for-distributed-
2016 Heraklion Crete Gr 1st Edition Elvira Albert
objects-components-and-systems-36th-ifip-wg-6-1-international-
conference-forte-2016-held-as-part-of-the-11th-international-
federated-conference-on-distributed-computi/
textbookfull.com
Studies in Computational Intelligence 678

Costin Badica
Amal El Fallah Seghrouchni
Aurélie Beynier · David Camacho
Cédric Herpson · Koen Hindriks
Paulo Novais Editors

Intelligent
Distributed
Computing X
Proceeding of the 10th International
Symposium on Intelligent Distributed
Computing – IDC 2016, Paris, France,
October 10–12 2016
Studies in Computational Intelligence

Volume 678

Series editor
Janusz Kacprzyk, Polish Academy of Sciences, Warsaw, Poland
e-mail: [email protected]
About this Series

The series “Studies in Computational Intelligence” (SCI) publishes new develop-

ments and advances in the various areas of computational intelligence—quickly and
with a high quality. The intent is to cover the theory, applications, and design
methods of computational intelligence, as embedded in the fields of engineering,
computer science, physics and life sciences, as well as the methodologies behind
them. The series contains monographs, lecture notes and edited volumes in
computational intelligence spanning the areas of neural networks, connectionist
systems, genetic algorithms, evolutionary computation, artificial intelligence,
cellular automata, self-organizing systems, soft computing, fuzzy systems, and
hybrid intelligent systems. Of particular value to both the contributors and the
readership are the short publication timeframe and the worldwide distribution,
which enable both wide and rapid dissemination of research output.

More information about this series at http://www.springer.com/series/7092

Costin Badica Amal El Fallah Seghrouchni
•

Aurélie Beynier David Camacho

•

Cédric Herpson Koen Hindriks

•

Paulo Novais
Editors

Intelligent Distributed
Computing X
Proceeding of the 10th International
Symposium on Intelligent Distributed
Computing – IDC 2016, Paris, France,
October 10–12 2016

123
Editors
Costin Badica David Camacho
Faculty of Automatics, Computer Science Universidad Autonoma de Madrid
and Electronics C. Francisco Tomas y Valiente, 11
University of Craiova Madrid
Craiova Spain
Romania
Cédric Herpson
Amal El Fallah Seghrouchni Sorbonne Universités
Sorbonne Universités UPMC Univ Paris 06, CNRS,
UPMC Univ Paris 06, CNRS, LIP6 UMR 7606
LIP6 UMR 7606 Paris
Paris France
France
Koen Hindriks
Aurélie Beynier Faculty of EEMCS
Sorbonne Universités Delft University of Technology
UPMC Univ Paris 06, CNRS, Delft, Zuid-Holland
LIP6 UMR 7606 The Netherlands
Paris
France Paulo Novais
Universidade do Minho
Campus of Gualtar
Braga
Portugal

ISSN 1860-949X ISSN 1860-9503 (electronic)

Studies in Computational Intelligence
ISBN 978-3-319-48828-8 ISBN 978-3-319-48829-5 (eBook)
DOI 10.1007/978-3-319-48829-5

Library of Congress Control Number: 2016954716

© Springer International Publishing AG 2017

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or
for any errors or omissions that may have been made.

Printed on acid-free paper

This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
 Preface

Intelligent Distributed Computing emerged as the result of the fusion and cross-
fertilization of ideas and research results in Intelligent Computing and Dis-
tributed Computing. Its main outcome was the development of the new gen-
eration of intelligent distributed systems, by combining methods and technol-
ogy from classical artificial intelligence, computational intelligence, multi-agent-
systems, and distributed systems.
The 10th Intelligent Distributed Computing IDC2016 continues the tradition
of the IDC Symposium Series that started 10 years ago as an initiative of two
research groups from:
1. Systems Research Institute, Polish Academy of Sciences, Warsaw, Poland
2. Software Engineering Department, University of Craiova, Craiova, Romania
The IDC Symposium welcomes submissions of original papers on all aspects
of intelligent distributed computing ranging from concepts and theoretical devel-
opments to advanced technologies and innovative applications. The symposium
aims to bring together researchers and practitioners involved in all aspects of
Intelligent Distributed Computing. IDC is interested in works that are relevant
for both Distributed Computing and Intelligent Computing, with scientific merit
in these areas.

This volume contains the proceedings of the 10th International Symposium

on Intelligent Distributed Computing, IDC2016. The symposium was hosted by
the Laboratoire d’Informatique de Paris 6 from the University Pierre and Marie
Curie, in Paris, France, between the 10th and the 12th of October, 2016.

The IDC2016 event comprised the main conference organized in eight ses-
sions: Dynamic Systems, Internet of Things, Security, Space-Based Coordination,
Behavioral Analysis, Optimization, Data Management and IC-Smart.
The proceedings book contains contributions, with 23 regular papers selected
from a total of 38 received submissions from 18 countries (counting the coun-
try of each co-author for each paper submitted). Each submission was carefully
reviewed by at least three members of the Program Committee. Acceptance
and publication were judged based on the relevance to the symposium topics,
clarity of presentation, originality and accuracy of results, and proposed solu-
tions. The acceptance rates were 60%, counting only regular papers. The con-
tributions published in this book address many topics related to theory and
applications of intelligent distributed computing including: cloud computing,
P2P networks, agent-based distributed simulation, ambient agents, smart and
context-driven environments, Internet of Things, network security, mobile com-
puting, Unmanned Vehicles, augmented physical reality, swarm computing, team
and social computing, constraints and optimization, and information fusion.

v
vi Preface

We would like to thank Janusz Kacprzyk, editor of Studies in Computational

Intelligence series and member of the Steering Committee, for his continuous
support and encouragement for the development of the IDC Symposium Series.
Also, we would like to thank the IDC2016 Program Committee members for
their work in promoting the event and refereeing submissions and also to all
colleagues who submitted their work to this event.
We deeply appreciate the efforts of our invited speakers Pr. Serge Haddad
from ENS Cachan, France, and Pr. Carlos COTTA from Universidad de Mlaga,
Spain, and thank them for their interesting lectures.
A special thanks also go to organizers of the special session IC-Smart : Amal
El Fallah Seghrouchni and Kenji Tei. Finally, we appreciate the efforts of local
organizers on behalf of the Laboratoire d’Informatique de Paris 6 (LIP6) from
the University Pierre and Marie Curie, Sorbonnes Universities, in Paris, France,
for hosting and organizing this event.

Craiova Costin Badica

Paris Amal El Fallah Seghrouchni
Paris Aurélie Beynier
Madrid David Camacho
Paris Cédric Herpson
Zuid-Holland Koen Hindriks
Braga Paulo Novais

July 2016
 Organization

Organizer
MultiAgent System team (SMA)
Laboratoire d’Informatique de Paris 6 (LIP6)
Sorbonne Universités, UPMC Univ Paris 06, France

General Chairs
Amal El Fallah Seghrouchni LIP6 - UPMC Sorbonne Universités, France
Costin Badica University of Craiova, Romania

Program Committee Chairs

Aurélie Beynier UPMC Sorbonne Universités, LIP6, France
David Camacho Universidad Autonoma de Madrid, Spain
Koen Hindriks Delft Robotics Institute, The Netherlands
Paulo Novais University of Minho, Portugal

Invited Speakers
Carlos Cotta Universidad de Malaga, Spain
Serge Haddad ENS Cachan, France

Program Comittee
Ajith Abraham Machine Intelligence Research Labs (MIR Labs)
Salvador Abreu JFLI-CNRS / LISP / CRI, University of Evora
Amparo Alonso-Betanzos University of A Corua
Ricardo Anacleto ISEP
Cesar Analide University of Minho
Razvan Andonie Central Washington University
Javier Bajo .Universidad Politécnica de Madrid
Nick Bassiliades Aristotle University of Thessaloniki
David Bednrek Charles University Prague
Doina Bein California State University, Fullerton
Gema Bello Orgaz Universidad Autonoma de Madrid
Nik Bessis Edge Hill University
Lars Braubach University of Hamburg
Dumitru Dan Burdescu University of Craiova
Giacomo Cabri Universit di Modena e Reggio Emilia
Davide Carneiro Universidade do Minho
Andre Carvalho USP
vii
viii Organization

Jose Carlos Castillo Montoya Universidad Carlos III de Madrid

Jen-Yao Chung IBM
Dorian Cojocaru University of Craiova
Rem Collier UCD
Phan Cong-Vinh Nguyen Tat Thanh University
Lus Correia University of Lisbon
Ängelo Costa Universidade do Minho
Paul Davidsson Malm University
Javier Del Ser Tecnalia Resaerch & Innovation
Giuseppe Di Fatta University of Reading
Amal El Fallah Seghrouchni UPMC Sorbonne Universités, LIP6, France
Vadim Ermolayev Zaporozhye National Univ.
Antonio Fernndez-Caballero Universidad de Castilla-La Mancha
Adina Magda Florea University Politehnica of Bucharest,
AI-MAS Laboratory
Giancarlo Fortino University of Calabria
Maria Ganzha University of Gdask
Antonio Gonzalez-Pardo Universidad Autonoma de Madrid
Bertha Guijarro-Berdias University of A Corua
Marjan Gusev UKIM University St Cyril and Methodius
Adnan Hashmi University of Lahore
Cédric Herpson UPMC Sorbonne Universités, LIP6, France
Dosam Hwang Yeungnam University
Barna Laszlo Iantovics Petru Maior University of Tg. Mures
Fuyuki Ishikawa National Institute of Informatics
Mirjana Ivanovic University of Novi Sad, Faculty of Sciences
Vicente Julian GTI-IA DSIC UPV
Jason Jung Chung-Ang University
Igor Kotenko St. Petersburg Institute for Informatics and
Automation
Dariusz Krol Wrocaw University of Technology
Florin Leon Technical University ”Gheorghe Asachi” of Iasi
Alessandro Longheu DIEEI - University of Catania
José Machado University of Minho, Computer Science and
Technology Centre
Ana Madureira Departamento de Engenharia Informtica
Giuseppe Mangioni DIEEI - University of Catania
Goreti Marreiros ISEP/IPP-GECAD
Ester Martinez-Martin Universitat Jaume I
Viviana Mascardi Department of Informatics, University of Genova
Ficco Massimo Second University of Naples (SUN)
Héctor Menéndez University Autonoma of Madrid
Organization ix

John-Jules Meyer Utrecht University

Paulo Moura Oliveira UTAD University
Grzegorz J Nalepa AGH University of Science and Technology
Jose Neves Universidade do Minho
David Obdrzalek Charles University in Prague
Andrea Omicini Alma Mater StudiorumUniversit di Bologna
Fernando Otero University of Kent
Juan Pavn Universidad Complutense de Madrid
Pawel Pawlewski Poznan University of Technology
Stefan-Gheorghe Pentiuc University Stefan cel Mare Suceava
Antonio Pereira Escola Superior de Tecnologia e Gesto do
IPLeiria
Dana Petcu West University of Timisoara
Florin Pop University Politehnica of Bucharest
Antonio Portilla-Figueras Universidad de Alcala
Maria Potop-Butucaru UPMC Sorbonne Universités, LIP6, Paris
Radu-Emil Precup Politehnica University of Timisoara
Maria D. R-Moreno Universidad de Alcala
Shahram Rahimi Southern Illinois University
Alessandro Ricci University of Bologna
Joel J.P.C. Rodrigues Instituto de Telecomunicaes, University of
Beira Interior
Domenico Rosaci DIMET Department, University Mediterranea
of Reggio Calabria
Sancho Salcedo-Sanz Universidad de Alcal
Corrado Santoro University of Catania - Dipartimento di
Matematica e Informatica
Ichiro Satoh National Institute of Informatics
Weiming Shen National Research Council
Fbio Silva Universidade do Minho
Safeeullah Soomro Indus University
Giandomenico Spezzano CNR-ICAR and University of Calabria
Stanimir Stoyanov University of Plovdiv ”Paisii Hilendarski”
Anna Toporkova National Research University Higher
School of Economics
Rainer Unland University of Duisburg-Essen, ICB
Salvatore Venticinque Seconda Universit di Napoli
Lucian Vintan ”Lucian Blaga” University of Sibiu
Martijn Warnier Delft University of Technology
Michal Wozniak Wroclaw University of Technology
Jakub Yaghob Charles University in Prague
Filip Zavoral Charles University in Prague
x Organization

Organizing Committee
Aurélie Beynier UPMC Sorbonne Universités, LIP6, France
Amal El Fallah Seghrouchni UPMC Sorbonne Universités, LIP6, France
Cédric Herpson UPMC Sorbonne Universités, LIP6, France

Sponsoring Institutions

Sorbonne Universités, UPMC Univ Paris 06

Laboratoire d’Informatique de Paris 6 (LIP6)
Faculté d’Ingénierie, UFR 919, UPMC
 Table of Contents

I Dynamic Systems

Adaptive Scaling Up/Down for Elasitc Clouds . . . . . . . . . . . . . . . . . . . . . . . . 3

Ichiro Satoh

A Dynamic Model to enhance the Distributed Discovery of SWs in P2P

Overlay Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Adel Boukhadra, Karima Benatchba and Amar Balla

Simulation of Dynamic Systems Using BDI Agents: Initial Steps . . . . . . . . 23

Amelia Bădică and Costin Bădică and Marius Brezovan

II Internet of Things
A Multi-Agent Middleware for the Deployment of Distributed
Applications in Smart Environments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Ferdinand Pierre, Cédric Dinont, Amal El Fallah Seghrouchni and
Patrick Taillibert

A Guidance of Ambient Agents Adapted to Opportunistic Situations . . . . 47

Ahmed-Chawki Chaouche, Jean-Michel Ilié and Djamel Eddine
Saı̈douni

Extended Context Patterns – A Visual Language for Context-Aware

Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Andrei Olaru and Adina Magda Florea

MDE4IoT: Supporting the Internet of Things with Model-Driven

Engineering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
Federico Ciccozzi and Romina Spalazzese

III Security
Detection of traffic anomalies in multi-service networks based on a
fuzzy logical inference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Igor Saenko, Sergey Ageev, and Igor Kotenko

Reconfiguration of RBAC schemes by genetic algorithms . . . . . . . . . . . . . . . 89

Igor Saenko and Igor Kotenko

String-based Malware Detection for Android Environments . . . . . . . . . . . . . 99

Alejandro Martı́n, Héctor D. Menéndez and David Camacho
xi
xii Table of Contents

IV Space-Based Coordination

Optimal Configuration Model of a Fleet of Unmanned Vehicles for

Interoperable Missions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Gabriella Gigante, Domenico Pascarella, Salvatore Luongo, Carlo
Di Benedetto, Angela Vozella, and Giuseppe Persechino

Spatial Tuples: Augmenting Physical Reality with Tuple Spaces . . . . . . . . . 121

Alessandro Ricci , Mirko Viroli, Andrea Omicini, Stefano Mariani,
Angelo Croatti and Danilo Poanini

Exploring unknown environments with multi-modal locomotion swarm . . . 131

Zedadra Ouarda, Jouandeau Nicolas, Seridi Hamid and Fortino
Giancarlo

V Behavioral Analysis

GroupTrust: Finding Trust-based Group Structures in Social Communities 143

Antonello Comi, Lidia Fotia, Fabrizio Messina, Domenico Rosaci
and Giuseppe M.L. Sarné

Non-intrusive Monitoring of Attentional Behavior in Teams . . . . . . . . . . . . 153

Davide Carneiro, Dalila Dures, Javier Bajo and Paulo Novais

A Speculative Computation Approach for Conflict Styles Assessment

with Incomplete Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Marco Gomes, Tiago Oliveira and Paulo Novais

Forming Classes in an e-Learning Social Network Scenario . . . . . . . . . . . . . 173

Pasquale De Meo, Lidia Fotia, Fabrizio Messina, Domenico Rosaci,
Giuseppe M. L. Sarné

VI Optimization

Scheduling Optimization in Grid with VO Stakeholders’ Preferences . . . . . 185

V. Toporkov, A. Toporkova, D. Yemelyanov, A. Bobchenkov and A.
Tselishchev

On the Application of Bio-inspired Heuristics for Network Routing

with Multiple QoS Constraints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
Miren Nekane Bilbao, Cristina Perfecto, Javier Del Ser, Xabier
Landa
Table of Contents xiii

Dealing with the Best Attachment Problem via Heuristics . . . . . . . . . . . . . . 205

M. Buzzanca, V. Carchiolo, A. Longheu, M. Malgeri and
G. Mangioni

VII Data Management

Towards Collaborative Sensing using Dynamic Intelligent Virtual Sensors 217
Radu-Casian Mihailescu, Jan Persson, Paul Davidsson, Ulrik
Eklund

Intelligent Data Metrics for Urban Driving with Data Fusion and
Distributed Machine Learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
Fbio Silva, Artur Quintas, Jason J. Jung, Paulo Novais and Cesar
Analide

A Probabilistic Sample Matchmaking Strategy for Imbalanced Data

Streams with Concept Drift . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
Jesus L. Lobo, Javier Del Ser, Miren Nekane Bilbao, Ibai Laña,
Sancho Salcedo Sanz

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247

 Part I

Dynamic Systems
 Adaptive Scaling Up/Down for Elastic Clouds

Ichiro Satoh

National Institute of Informatics

2-1-2 Hitotsubashi, Chiyoda-ku, Tokyo, Japan
[email protected]

Abstract. An approach for adapting distributed applications in response to changes

in user requirements and resource availability is presented. The notion of elastic-
ity enables capabilities and resources to be dynamically provisioned and released.
However, existing applications do not inherently support elastic capabilities and
resources. To solve this problem, we propose two novel functions: dynamic de-
ployment of components and dividing and merging components. The former en-
ables components to relocate themselves at new servers when provisioning the
servers and at remaining servers when deprovisioning servers, while the latter
enables the states of components to be divided, passed to other components, and
merged with other components in accordance with user-deÀned functions. We
constructed a middleware system for Java-based general-purpose software com-
ponents with the two functions because they are useful to adapt applications to
elasticity in cloud computing. The proposed system is useful because it enables
applications be operated with elastic capabilities and resources in cloud comput-
ing.

1 Introduction

Elasticity in cloud computing was originally deÀned in physics as a material property

with the capability of returning to its original state after a deformation. The concept of
elasticity has been applied to computing and is commonly considered to be one of the
central attributes of cloud computing. For example, the NIST deÀnition of cloud com-
puting [10] states that capabilities can be elastically provisioned and released, in some
cases automatically, to scale rapidly outward and inward in accordance with demand.
To the consumer, the capabilities available for provisioning often appear to be unlimited
and can be appropriated in any quantity at any time.
However, the conventional design and development of application software are not
able to adapt themselves to elastically provisioning and deprovisioning resources in
cloud computing. Furthermore, it is difÀcult to deprive parts of the computational re-
sources that such applications have already used. There have been a few attempts to
solve this problem. For example, Mesos [4] is a platform for sharing commodity clus-
ters between distributed data processing frameworks such as Hadoop and Spark. These
frameworks themselves are elastic in the sense that they have the ability to scale their
resources up or down, i.e., they can start using resources as soon as applications want
to acquire the resources or release the resources as soon as the applications do not need
them.

© Springer International Publishing AG 2017 3

C. Badica et al. (eds.), Intelligent Distributed Computing X,
Studies in Computational Intelligence 678, DOI 10.1007/978-3-319-48829-5_1
4 I. Satoh

We assume that applications are running on dynamic distributed systems, includ-

ing cloud computing environments, in the sense that computational resources avail-
able from the applications may be dynamically changed due to elasticity. We propose a
framework for enabling distributed applications to be adapted to changes in their avail-
able resources on elastic distributed systems as much as possible. The key ideas behind
the framework are the duplication and migration of running software components and
the integration of multiple same components into single components. To adapt dis-
tributed applications, which consist of software components, to elasticity in cloud com-
puting, the framework divides applications into some of the components and deploys
the components at servers, which are provisioned, and merges the components running
at servers, which are deprovisioned, into other components running at other available
servers. We are constructing a middleware system for adapting general-purpose soft-
ware components to changes in elastic cloud computing.

2 Related Work

Cloud computing environments allow for novel ways of efÀcient execution and manage-
ment of complex distributed systems, such as elastic resource provisioning and global
distribution of application components. Resource allocation management has been stud-
ied for several decades in various contexts in distributed systems, including cloud com-
puting. We focus here on only the most relevant work in the context of large-scale server
clusters and cloud computing in distributed systems. Several recent studies have ana-
lyzed cluster traces from Yahoo!, Google, and Facebook and illustrate the challenges
of scale and heterogeneity inherent in these modern data centers and workloads. Mesos
[4] splits the resource management and placement functions between a central resource
manager and multiple data processing frameworks such as Hadoop and Spark by us-
ing an offer-based mechanism. Resource allocation is performed in a central kernel and
master-slave architecture with a two-level scheduling system. With Mesos, reclaim of
resources is handled for unallocated capacity that is given to a framework. The Google
Borg system [11] is an example of a monolithic scheduler that supports both batch jobs
and long-running services. It provides a single RPC interface to support both types of
workload. Each Borg cluster consists of multiple cells, and it scales by distributing the
master functions among multiple processes and using multi-threading. YARN [14] is a
Hadoop-centric cluster manager. Each application has a manager that negotiates for the
resources it needs with a central resource manager. These systems assume the execu-
tion of particular applications, e.g., Hadoop and Spark, or can assign resources to their
applications before the applications start. In contrast, our framework enables running
applications to adapt themselves to changes in their available resources.
Several academic and commercial projects have explored attempts to create auto-
scaling applications. Most of them have used static mechanisms in the sense that they
are based on models to be deÀned and tuned at design time. For example, Tamura et al.
[13] proposed an approach to identify system viability zones that are deÀned as states
in which the system operation is not compromised and to verify whether the current
available resources can satisfy the validation at the development of the applications.
The variety of available resources with different characteristics and costs, variability and
Adaptive Scaling Up/Down for Elastic Clouds 5

unpredictability of workload conditions, and different effects of various conÀgurations

of resource allocations make the problem extremely hard if not impossible to solve
algorithmically at design time.
ReconÀguration of software systems at runtime to achieve speciÀc goals has been
studied by several researchers. For example, Jaeger et al. [6] introduced the notion of
self-organization to an object request broker and a publish / subscribe system. Lym-
beropoulos et al. [9] proposed a speciÀcation for adaptations based on their policy spec-
iÀcation, Ponder [1], but it was aimed at specifying management and security policies
rather than application-speciÀc processing and did not support the mobility of compo-
nents. Lupu and Sloman [8] described typical conÁicts between multiple adaptations
based on the Ponder language. Garlan et al. [3] presented a framework called Rainbow
that provided a language for specifying self-adaptation. The framework supported adap-
tive connections between operators of components that might be running on different
computers. They intended to adapt coordinations between existing software compo-
nents to changes in distributed systems, instead of increasing or decreasing the number
of components.
Most existing attempts have been aimed at provisioning of resources, e.g., the work
of Sharman at al. [12]. Therefore, there have been a few attempts to adapt applica-
tions to deprovisioned resources. Nevertheless, they explicitly or implicitly assume that
their target applications are initially constructed on the basis of master-slave and redun-
dant architectures. Several academic and commercial systems tried introducing live-
migration of virtual machines (VMs) into their systems, but they could not merge be-
tween applications, because they were running on different VMs.1 Jung et al.[7] have
focused on controllers that take into account the costs of system adaptation actions con-
sidering both the applications (e.g., the horizontal scaling) and the infrastructure (e.g.,
the live migration of virtual machines and virtual machine CPU allocation) concerns.
Thus, they differ from most cloud providers, which maintain a separation of concerns,
hiding infrastructural control decisions from cloud clients.

3 Approach

As mentioned in the Àrst section, elasticity, which is one of the most important features
of cloud computing, is the degree to which a system is able to adapt to workload changes
by provisioning and de-provisioning resources in an autonomic manner. Applications
need to adapt themselves to changes in their available resources due to elasticity.

3.1 Requirements

We will propose a framework to adapt applications to the provisioning and deprovi-

sioning of servers, which may be running on physical or virtual machines, and software
containers, such as Docker, by providing an additional layer of abstraction and automa-
tion of virtualization. Our framework assumes that each application consists of one or
1
Unlike private cloud environments, most commercial ones do not support live migrations be-
cause they tend to need wide-band and low-latency networks between servers.
6 I. Satoh

more software components that may be running on different computers. It has Àve re-
quirements.
– Supports elasticity: Elasticity allows applications to use more resources when
needed and fall back afterwards. Therefore, applications need to be adapted to dy-
namically increasing and decreasing their available resources.
– Separation of concerns: All software components should be deÀned indepen-
dently of our adaptation mechanism as much as possible. This will enable develop-
ers to concentrate on their own application-speciÀc processing.
– Self-adaptation: Distributed systems essentially lack a global view due to com-
munication latency between computers. Software components, which may be run-
ning on different computers, need to coordinate themselves to support their appli-
cations with partial knowledge about other computers.
– Non-centralized management: There is no central entity to control and coordinate
computers. Our adaptation should be managed without any centralized manage-
ment so that we can avoid any single points of failures and performance bottlenecks
to ensure reliability and scalability.
– General purpose: There are various applications running on a variety of dis-
tributed systems. Therefore, the framework should be implemented as a practical
middleware system to support general-purpose applications.
We assume that, before the existence of deprovisioning servers, the target cloud com-
puting environment can notify servers about the deprovisioning after a certain time.
Cloud computing environments can be classiÀed into three types: Infrastructure as a
Service (IaaS), Platform as a Service (PaaS), and Software as a Service (SaaS). The
framework is intended to be used in the second and third, but as much as possible it
does not distinguish between the two.

3.2 Adaptation for elasticity in cloud computing

To adapt applications to changes in their available resources due to elasticity, the frame-
work adapts the applications to provisioning and de-provisioning resources (Fig. 1).
– Adaptation to provisioning resources When provisioning servers, if a particular
component is busy and the servers can satisfy the requirement of that component,
the framework divides the component into two components and deploys one of
them at the servers, where the divided components have the same programs but their
internal data can be replicated or divided in accordance with application-speciÀc
data divisions.
– Adaptation to deprovisioning resources When deprovisioning servers, compo-
nents running on the servers are relocated to other servers that can satisfy the re-
quirements of the components. If other components whose programs are the same
as the former components co-exist on the latter servers, the framework instructs the
deployed components to be merged to the original components.
The Àrst and second adaptations need to deploy components at different computers.
Our framework introduces mobile agent technology. When migrating and duplicating
components, their internal states stored in their heap areas are transmitted to their des-
tinations and are replicated at their clones.
Adaptive Scaling Up/Down for Elastic Clouds 7

Partitioning function &

Adaptation of component to provisioning server

Step 1. Step 2. Duplication with Step 3. Migration

partitioning function

Heap space Heap space Heap space Heap space Heap space
Heap space
Program
gram C
Code
Program Code Program Code Program Code Program Code Program Code

Server 1 Server 2 Server 1 Server 2 Server 1 Server 2

Provisioned server

Adaptation of component to deprovisioning server

Merging components with
Step 1. Step 2. Step 3.
reduce function

Migration

Heap space Heap space Heap space Heap space Heap space
Heap space
Program Code Program Code
Program Code Program Code Program Code Program Code

Server 1 Server 2 Server 1 Server 2 Server 1 Server 2

Deprovisioned server (deprovisioned)
after certain time
Key Value
Key Value Key Value
e.g., Reduce function
(sum)
“A1” 50
“A2” 30 , “A2” 100
“A3” 200
= “A1” 50
“A2” 130
“A3” 200

Fig. 1. Adaptation to (de)provisioning servers

3.3 Data stores for dividing and merging components

The framework provides another data store for dividing and merging components. To
do this, it introduces two notions: key-value store (KVS) and reduce functions of the
MapReduce processing. The KVS offers a range of simple functions for manipulation
of unstructured data objects, called values, each of which is identiÀed by a unique key.
Such a KVS is implemented as an array of key and value pairs. Our framework provides
KVSs for components so that each component can maintain its internal state in its KVS.
Our KVSs are used to pass the internal data of components to other components and
to merge the internal data of components into their uniÀed data. The framework also
provides a mechanism to divide and merge components with their internal states stored
at KVSs by using MapReduce processing. MapReduce is a most typical modern com-
puting models for processing large data sets in distributed systems. It was originally
studied by Google [2] and inspired by the map and reduce functions commonly used in
parallel list processing (LISP) and functional programming paradigms.

– Component division Each duplicated component can inherit partial or all data
stored in its original component in accordance with user-deÀned partitioning func-
tions, where each function map of each item of data in its original component’s
KVS is stored in either the original component’s KVS or the duplicated compo-
nent’s KVS without any redundancy.
– Component fusion When unifying two components that generated from the same
programs into a single component, the data stored in the KVSs of the two com-
ponents are merged by using user-deÀned reduce functions. These functions are
similar to the reduce functions of MapReduce processing. Each of our reduce func-
tions processes two values of the same keys and then maps the results to the entries
of the keys. Figure 1 shows two examples of reduce functions. The Àrst concate-
8 I. Satoh

nates values in the same keys of the KVSs of the two components, and the second
sums the values in the same keys of their KVSs.

4 Implementation
Our framework consists of two parts: component runtime system and components. The
former is responsible for executing, duplicating, and migrating components. The later
is autonomous programmable entities like software agents. The current implementation
is built on our original mobile agent platform as existing mobile agent platforms are not
optimized for data processing.

4.1 Adaptation for elasticity

When provisioning servers, the framework can divide a component into two compo-
nents whose data can be divided before deploying one of them at the servers. When
deprovisioning servers, the framework can merge components that are running on the
servers into other components.

Dividing component When dividing a component into two, the framework has two
approaches for sharing between the states of the original and clone components.
– Sharing data in heap space Each runtime system makes one or more copies of
components. The runtime system can store the states of each agent in heap space in
addition to the codes of the agent in a bit-stream formed in Java’s JAR Àle format,
which can support digital signatures for authentication. The current system basi-
cally uses the Java object serialization package for marshalling agents. The package
does not support the capturing of stack frames of threads. Instead, when an agent is
duplicated, the runtime system issues events to it to invoke their speciÀed methods,
which should be executed before it is duplicated, and it then suspends their active
threads.
– Sharing data in KVS When dividing a component into two components, the KVS
inside the former is divided into two KVSs in accordance with user-deÀned parti-
tioning functions in addition to built-in functions, and the divided KVSs are main-
tained inside the latter. Partitioning functions are responsible for dividing the inter-
mediate key space and assigning intermediate key-value pairs to the original and
duplicated components. In other words, the partition functions specify the com-
ponents to which an intermediate key-value pair must be copied. KVSs are con-
structed as in-memory storage to exchange data between components. It provides
tree-structured KVSs inside components. In the current implementation, each KVS
in each data processing agent is implemented as a hash table whose keys, given as
pairs of arbitrary string values, and values are byte array data, and it is carried with
its agent between nodes,
where a default partitioning function is provided that uses hashing. This tends to result
in fairly well-balanced partitions. The simplest partitioning functions involve comput-
ing the hash value of the key and then taking the mod of that value using the number of
the original and duplicated components.
Adaptive Scaling Up/Down for Elastic Clouds 9

Merging components The framework provides a mechanism to merge the data stored
in the KVSs of different components instead of the data stored inside their heap spaces.
Like the reduce of MapReduce processing, the framework enables us to deÀne a reduce
function that merges all intermediate values associated with the same intermediate key.
When merging two components, the framework can discard the states of their heap
spaces or keep the state of the heap space of one of them. Instead, the data stored in
the KVSs of different components can be shared. A reduce function is applied to all
values associated with the same intermediate key to generate output key-value pairs.
The framework can merge more than two components at the same computers because
components can migrate to the computers that execute co-components that the former
wants to merge to.

5 Evaluation

We outline our current implementation. A prototype implementation of this framework

was constructed with Java Developer Kit (JDK) version 1.7 or later. The implementation
enabled graphical user interfaces to operate the mobile agents. Although the current
implementation was not constructed for performance, we evaluated the performance of
our framework with CoreOS, which is a lightweight operating system based on Linux
with JDK version 1.8 with Docker, which is software-based environment that automates
the deployment of applications inside software containers by providing an additional
layer of abstraction and automation of operating-system-level virtualization on Linux,
on Amazon EC2. For each dimension of the adaptation process with respect to a speciÀc
resource type, elasticity captures the following core aspects of the adaptation:

– Adaptation speed at provisioning servers The speed of scaling up is deÀned as

the time it takes to switch from provisioning of servers by the underlying system,
e.g., cloud computing environment.
– Adaptation speed at deprovisioning servers The speed of scaling down is deÀned
as the time it takes to switch from deprovisioning of servers by the underlying
system, e.g., cloud computing environment.

The speed of scaling up/down does not correspond directly to the technical resource
provisioning/deprovisioning time. Table 1 shows the basic performance. The compo-
nent was simple and consisted of basic callback methods. The cost included that of in-
voking two callback methods. The cost of component migration included that of open-
ing TCP transmission, marshaling the agents, migrating the agents from their source
computers to their destination computers, unmarshalling the components, and verifying
security.

Table 1. Basic operation performance

Latency (ms)
Duplicating component 10
Merging component 8
Migrating component between two servers 32
10 I. Satoh

Figure 2 shows the speed of the number of divided and merged components at pro-
visioning and deprovisioning servers. The experiment provided only one server to run
our target component, which was a simple HTTP server (its size was about 100 KB). It
added one server every ten seconds until there were eight servers and then removed one
server every ten seconds after 80 seconds had passed. The number of components was
measured as the average of the numbers in ten experiments. Although elasticity is al-
ways considered with respect to one or more resource types, the experiment presented in
this paper focuses on computing environments for executing components, e.g., servers.
There are two metrics in an adaptation to elastic resources, scalability and efÀciency,
where scalability is the ability of the system to sustain increasing workloads by making
use of additional resources, and efÀciency expresses the amount of resources consumed
for processing a given amount of work.

8
Number of servers

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Time (sec)
Number of (de)provisioning servers notified from Cloud
Number of components by the proposed framework
Under-provisioning state (the numbers of components and available servers)
Over-provisioning state (the numbers of components and available servers)

Fig. 2. Number of components at (de)provisioning servers

– A is the average time to switch from an underprovisioned state to an optimal or

overprovisioned state and corresponds to the average speed of scaling up or scaling
down.
– U is the
average amount of underprovisioned resources during an underprovisioned
period. U is the accumulated amount of underprovisioned resources and corre-
sponds to the blue areas in Fig. 2.
– D is theaverage amount of overprovisioned resources during an overprovisioned
period. D is the accumulated amount of underprovisioned resources and corre-
sponds to the red areas in Fig. 2.
The precision of scaling up or down is deÀned as the absolute deviation of the current
amount of allocated resources from the actual resource provisioning or deprovisioning.
We deÀne the average precision of scaling up Pu and that of scaling down Pd . The
efÀciency of scaling up or down is deÀned as the absolute deviation of the accumu-
lated amount of underprovisioned or overprovisioned resources from the accumulated
amount of provisioned or deprovisioned resources, speciÀed as EU or ED .
   
U D U D
Pu = Pd = Eu = Ed =
Tu Td Ru Rd
Adaptive Scaling Up/Down for Elastic Clouds 11

where Tu and Td are the total durations of the evaluation periods and Ru and Rd are
the accumulated amounts of provisioned resources when scaling up and scaling down
phases, respectively.2 Table 2 shows the precision and efÀciency of our framework.
Table 2. Basic operation efÀciency

Rate
Pu (Precision of scaling up) 99.2 %
Pd (Precision of scaling down) 99.1 %
Eu (EfÀciency of scaling up) 99.6 %
Ed (EfÀciency of scaling down) 99.4 %

Our component corresponds to an HTTP server, since web applications have very
dynamic workloads generated by variable numbers of users, and they face sudden peaks
in the case of unexpected events. Therefore, dynamic resource allocation is necessary
not only to avoid application performance degradation but also to avoid under-utilized
resources. The experimental results showed that our framework could follow the elasti-
cally provisioning and deprovisioning of resources quickly, and the number of the com-
ponents followed the number of elastic provisioning and deprovisioning of resources
exactly. The framework was scalable because its adaptation speed was independent of
the number of servers.

6 Conclusion

This paper presented a framework for enabling distributed applications to be adapted

to changes in their available resources in distributed systems. It was useful for adapt-
ing applications to elasticity in cloud computing. The key ideas behind the framework
are dynamic deployment of components and dividing and merging components. The
former enabled components to relocate themselves at new servers when provisioning
the servers and at remaining servers when de-provisioning the servers, and the latter
enables the states of components to be divided, and passed to other components, and
merged with other components in accordance with user-deÀned functions. We believe
that our framework is useful because it enables applications to be operated with elastic
capabilities and resources in cloud computing.

References
1. N. Damianou, N. Dulay, E. Lupu, and M. Sloman: The Ponder Policy SpeciÀcation Language,
in Proceedings of Workshop on Policies for Distributed Systems and Networks (POLICY’95),
pp.18–39, Springer-Verlag, 1995.
2. J. Dean and S. Ghemawat: MapReduce: simpliÀed data processing on large clusters, in Pro-
ceedings of the 6th conference on Symposium on Opearting Systems Design and Implemen-
tation (OSDI’04), 2004.
2
Ru and Rd correspond to the amount of provisioned resources according to cloud computing
environments.
12 I. Satoh

3. D. Garlan, S.W. Cheng, A.C.Huang, B. R. Schmerl, P. Steenkiste: Rainbow: Architecture-

Based Self-Adaptation with Reusable Infrastructure, IEEE Computer Vol.37, No.10, pp.46-
54, 2004.
4. B. Hindman, A. Konwinski, M. Zaharia, A. Ghodsi, A. Joseph, R. Katz, S. Shenker, and I.
Stoica. Mesos: a platform for Àne-grained resource sharing in the data center In Proceedings
of USENIX Symposium on Networked Systems Design and Implementation (NSDI), 2011.
5. C. Inzinger, at al., Decisions, Models, and Monitoring–A Lifecycle Model for the Evolution
of Service-Based Systems, In Proceedings of Enterprise Distributed Object Computing Con-
ference (EDOC), pp.185-194, IEEE Computer Society, 2013.
6. M. A. Jaeger, H. Parzyjegla, G. Muhl, K. Herrmann: Self-organizing broker topologies
for publish/subscribe systems, in Proceedings of ACM symposium on Applied Computing
(SAC’2007), pp.543-550, ACM, 2007.
7. G. Jung, et. al.: A Cost-Sensitive Adaptation Engine for Server Consolidation of Multitier
Applications, In Proceedings of Middleware’2009, LNCS, Vol.5896, pp.163183, Springer,
2009.
8. E. Lupu and M. Sloman: ConÁicts in Policy-Based Distributed Systems Management, IEEE
Transaction on Software Engineering, Vol.25, No.6, pp.852-869, 1999.
9. L. Lymberopoulos, E. Lupu, M. Sloman: An Adaptive Policy Based Management Frame-
work for Differentiated Services Networks, in Proceedings of 3rd International Workshop on
Policies for Distributed Systems and Networks (POLICY 2002), pp.147-158, IEEE Computer
Society, 2002.
10. P. Mell, T. Grance: The NIST DeÀnition of Cloud Computing, Technical report of U.S. Na-
tional Institute of Standards and Technology (NIST), Special Publication 800-145, 2011.
11. A. Verma, L. Pedrosa, M. Korupolu, D. Oppenheimer, E. Tune, and J. Wilkes: Large-scale
cluster management at Google with Borg, EuroSys15, ACM 2015.
12. U. Sharma, P. Shenoy, S. Sahu, A. Shaikh: A cost-aware elasticity provisioning system
for the cloud In Proceedings of International Conference on Distributed Computing Systems
(ICDCS’2011), pp.559570, IEEE Computer Society, 2011.
13. G. Tamura et. al.,: Towards Practical Runtime VeriÀcation and Validation of Self-Adaptive
Software Systems, Proceedings of Self-Adaptive Systems, LNCS 7475, pp. 108132, 2013.
14. V. K. Vavilapalli, el. al.,: Apache Hadoop YARN: Yet Another Resource Negotiator, In Pro-
ceedings of Symposium on Cloud Computing (SoCC’2013), ACM, 2013.
15. World Wide Web Consortium (W3C): Composite Capability/Preference ProÀles (CC/PP),
http://www.w3.org/TR/NOTE-CCPP, 1999.
 A Dynamic Model to enhance the Distributed
Discovery of services in P2P Overlay Networks

Adel Boukhadra, Karima Benatchba, and Amar Balla

Ecole national Supérieur d’Informatique, BP 68M, Oued-Smar, 16309, Alger, Algérie

a [email protected], k [email protected], a [email protected]

Abstract. In Service Computing (SC), online Semantic Web services

(SWs) is evolving over time and the increasing number of SWs with
the same function on the Internet, a great amount of candidate services
emerge. So, efficiency and effectiveness has become a stern challenge for
distributed discovery to tackle uniformed behavior evolution of service
and maintain high efficiency for large-scale computing. The distributed
discovery of SWs according to their functionality increases the capability
of an application to fulfill their own goals. In this paper, we describe an
efficient and an effective approach for improving the performance and
effectiveness of distributed discovery of SWs in P2P systems. As most
Web services lack a rich semantic description, we extend the distributed
discovery process by exploiting collaborative ranking to estimate the sim-
ilarity of a SWs being used by existing hybrid matching technique of
OWL-S (Ontology Web Language for Services) process models in order
to reduce costs and execution time. We mapped our distributed discov-
ery of OWL-S process models by developing a real application based
on Gamma Distribution; a technique used to decrease the bandwidth
consumption and to enhance the scalability of P2P systems. The partic-
ularity of the Gamma Distribution is then integrated for disseminating
request about the P2P networks to perform quality based ranking so
that the best SWs can be recommended first. The experimental result
indicates that our approach is efficient and able to reduce considerably
the execution time and the number of message overhead, while preserv-
ing high levels of the distributed discovery of SWs on large-size P2P
networks.

Keywords: SWs, Distributed Discovery, P2P Computing, Gamma Dis-

tribution, Matching of Ontology, OWL-S process model.

1 Introduction

Service-oriented computing (SOC) has emerged as an effective means of devel-

oping distributed applications and an accurate assessment the full potential of
reputation which is essential for discovering between alternative Web services.
Discovery of Web service is a great interest and is a fundamental area of research
in distributed computing [6].

© Springer International Publishing AG 2017 13

C. Badica et al. (eds.), Intelligent Distributed Computing X,
Studies in Computational Intelligence 678, DOI 10.1007/978-3-319-48829-5_2
14 A. Boukhadra et al.

Actually, with the increasing number of SWs with the same function, in an
open and dynamic environment, such as Internet, a great amount of candidate
services emerge. Moreover, the number of registries that offer available Web
services is also increasing significantly. It is not always easy to find services
that matching users queries. Find a service from the candidate Web service set,
according to the requirements of end-users has become a key hindrance and a
strenuous task even for an experienced user. The Web service discovery satisfying
the query is led by an optimization process, aiming to achieve the best SWs in
the end [1] [4] [5] [7].
This could be very challenging to end-users given the huge number of avail-
able Web services online, who have to decide where to fulfill their requests: on
local directory or to a replace this structure by several registries properly orga-
nized to support the dynamic, flexible and more efficient request propagation in a
highly-dynamic distributed computing. Due to low accuracy, poor performance,
high operational and maintenance cost, and sometimes for low availability of
the functionality of SWs, the distributed discovery of SWs on the P2P systems
provides a unique opportunity to address the above challenges [1] [7]. In recent
years, the rise of P2P networks is attested by the increasing amount of interest in
both commercial and academic areas to develop systems for data sharing simple
and effective with many advantages such as decentralization, self-organization,
autonomy, etc. At the same time, the SWs community has been slowly evolving
toward a higher degree of distribution [4] [5] [6].
In this paper, we propose an efficient and an effective approach that addresses
some aspects related to problems the time complexity of collaboration in the pro-
cess of automatic discovery for SWs in P2P computing. For this purpose, our
approach is based on P2P computing that proved to be scalable, more fault tol-
erant by eliminating the single point of failure, efficient by reducing the overhead
of centralized update of the service discovery and robust solutions for distributed
discovery of SWs. Specifically, our contributions in this work are summarized as
follows. We exploit hybrid matching technique of OWL-S process models to de-
velop functional features of Web services and the desired specification given by
the user. Both the Web services and request can be then represented as OWL-S
process model. We incorporate Gamma Distribution based collaborative ranking
to identify additional functionally relevant Web services, in order to efficiently
and effectively discover appropriate Web services distributed among all peers in
a large-size P2P network. The idea of using Gamma Distribution is to manage
large and continuously growing spaces of Web services with reasonable resolu-
tion times. The top-k most possible Web services are discovered, which are all
considered as functionally relevant to the new Web services.
The remainder of this work is organized as follows: Section 2 describes the
related work. In Section 3 we focus on the proposed distributed approach for dis-
covering SWs in the unstructured P2P computing. An experimental evaluation
is presented in Section 4, and, finally, in Section 5, we present our conclusion
and highlight our future work directions.
A Dynamic Model to enhance the Distributed Discovery … 15

2 Related Work
In this section, we discuss several representative related work and differentiate
them with our work.
In [2], authors propose the P2P-based Semantic Driven Service Discovery
(P2P-SDSD) framework to enable cooperation and communication based on a se-
mantic overlay that organizes semantically the P2P-integrated knowledge space
and emerges from local interactions between peers. The semantic overlay can be
seen as a continuously evolving conceptual map across collaborative peers that
provide similar services and constitute synergic service centers in a given domain.
The semantic overlay enables effective similarity based service search and opti-
mization strategies are defined for request propagation over the unstructured
P2P network keeping low the generated network overload. Each collaborative
peer in the unstructured P2P network has a local knowledge infrastructure con-
stituted by: (i) UDDI Registry; (ii) Peer Ontological Knowledge, that provides
a conceptualization of abstract service operations and Input/output parame-
ters through a given domain ontology; a conceptualization of service categories
through a Service Category Taxonomy (SCT).
In [7], present a technique to improve discovery in unstructured P2P ser-
vice networks, based on a probabilistic forwarding algorithm driven by network
knowledge, such as network density, and traces of already discovered service
compositions (CONs). The technique aims at reducing the composition time
and the messages exchanged during composition, relying on two considerations:
if the network is dense, forwarding can be limited to a small number of neigh-
bors; if the network is semi structured in CONs, forwarding can be directed to
the super peers that may own the desired information. The approach improves
the discovery and composition process by using distributed bidirectional search.
The benefit is twofold: first, it is possible to have concurrent searches in a P2P
service network in both goal directions (from pre- to post- and from post to pre-
conditions), reducing the response time when solutions are present; second, when
no complete solution for a goal is present, gaps in partial found solutions can be
identified. This way, it is possible to have feedbacks about users’ most required
unavailable business operations, allowing providers to discover new business op-
portunities.
In [11], authors present a distributed approach to SWs publication and dis-
covery by leveraging structured P2P network. In this work, the computers con-
cerned constitute a P2P network to maintain the sharable domain and service
ontologies to facilitate SWs discovery. When a requestor submits a semantic
query for desired services, the P2P network can effectively obtain semantically
qualified services. The main contributions of this work can be summarized as
follows: this approach introduces a semantic-based service matching rule. In or-
der to achieve the optimal match between a query, it proposes a concept of
Ordered-Concept-Tree (OCT) to semantically sort the relevant concepts for ser-
vice matching. In addition, to freely share and make full use of the semantic
concepts defined in ontologies for OCT construction, it also proposes a method
to publish ontologies to structured P2P network.
16 A. Boukhadra et al.

3 The Proposed Approach

In this section, we describe in detail the proposed approach that uses dynamic
topology adaptation to improve the efficiency and the effectiveness of distributed
discovery of SWs.

Fig. 1. Unstructured P2P Approach for Discovering SWs.

We propose further model which defines the topology of the P2P computing
for supporting communication and collaboration between different Web service
by the means of an effective use of the resources of P2P network and the man-
agement of scalability in an integrated and practical way. Such our model can
support scalability, maximize search recall, reduce the query traffic and must
also be able to achieve an acceptable performance. There are mainly three dif-
ferent architectures for P2P systems: hybrid, pure and unstructured; but the
unstructured P2P systems are most commonly used in today’s Internet [5] [6]
[11].
In unstructured P2P network, we distinguish between two types of peers,
neighbor and ultra-peer. Each peer is linked to a set of other peers in the P2P
network via bi-directional connections, that is, each peer has a limited number of
neighbors and it can choose the appropriate other neighbors it wishes to connect
when necessary. When a peer receives a search request, it manages the discovery
process locally and/or it will forward the query from other neighbors in the P2P
network using the dynamic querying algorithm.
If a peer frequently returns good results of a large number of queries from
one of its neighbors, it is likely to have common collaboration with the request-
ing peer to discover the appropriate Web services. It will mark this neighbor
(the requesting peer) as an ultra-peer of the requesting peer (see Figure 1). To
illustrate our model, consider the sample network depicted in Figure 1, in which
each peer has a several neighbors and ultra-peers. Peer 11 has Peer 2, Peer 5,
Peer 15 and Peer 17 as neighbors; and Peer 11 as ultra-peer.
Exploring the Variety of Random
Documents with Different Content
The Project Gutenberg eBook of The Literary
History of the Adelphi and Its Neighbourhood
 This ebook is for the use of anyone anywhere in the United
States and most other parts of the world at no cost and with
almost no restrictions whatsoever. You may copy it, give it away
or re-use it under the terms of the Project Gutenberg License
included with this ebook or online at www.gutenberg.org. If you
are not located in the United States, you will have to check the
laws of the country where you are located before using this
eBook.

Title: The Literary History of the Adelphi and Its Neighbourhood

Author: Austin Brereton

Release date: June 22, 2018 [eBook #57372]

Language: English

Credits: Produced by Clarity, Graeme Mackreth and the Online

Distributed Proofreading Team at http://www.pgdp.net
(This
file was produced from images generously made
available
by The Internet Archive/American Libraries.)

*** START OF THE PROJECT GUTENBERG EBOOK THE LITERARY

HISTORY OF THE ADELPHI AND ITS NEIGHBOURHOOD ***
 THE LITERARY HISTORY
OF THE ADELPHI AND
ITS NEIGHBOURHOOD

The Terrace. York Buildings. 1796.

From a water colour by J. Richards R.A.

THE LITERARY HISTORY

OF THE ADELPHI
AND ITS NEIGHBOURHOOD
 By AUSTIN BRERETON

WITH A NEW INTRODUCTION

NEW YORK:

DUFFIELD & COMPANY

36-38 WEST 37th STREET

1909

First Edition 1907

Second Edition 1908

(All rights reserved.)

Note
This book is intended for the general reader, as well as for the
antiquarian and the lover of London. To this end, the history of the
Adelphi and its immediate neighbourhood to the west and on the
south side of the Strand has been related in—as far as possible—
narrative form. At the same time, it need hardly be said, every care
has been taken to present the multitude of details correctly and as a
truthful picture of one of the most interesting parts of the great
metropolis. I should be ungrateful if I did not take this opportunity of
again—as in the case of my chronicle of the Lyceum and Henry Irving
—thanking Mr. E. Gardner for so courteously placing at my disposal
his unique and invaluable collection of London records and
engravings. The majority of the illustrations were kindly lent by him;
others were copied from prints in the British Museum. I have also to
thank the officials of St. Martin's Library for their ready help in
enabling me to consult, at my leisure, some scarce books connected
with the literature of historical London.
A.B.

INTRODUCTION
"The Literary History of the Adelphi" has journeyed from one side of
the neighbourhood to the other, from west to east. That is to say, its
publication has been acquired by Mr. Fisher Unwin, hence the
removal of the book from York Buildings to Adelphi—originally called
"Royal," and still so marked on the old plans—Terrace. This
peregrination gives me the opportunity of supplementing the original
work with some interesting particulars which have just come into my
possession. Who would think that within a short distance of the
Strand, if not actually within the proverbial stone's throw, there are
"cottages," and cottages, too, with trees and flowers and lawns, and
a mighty river, for prospect? Yet such is the case, although it is no
wonder that the rate collector who is new to this part of London has
much ado to find "Adelphi cottages." They belong to that mysterious
region which lies underneath the Strand level of the Adelphi and is
vaguely known as the "arches." If the reader will glance at the
illustration which faces page 32—"The Buildings called the Adelphi"—
he will see, at the top of the arches and under the terrace, some
fifteen semi-circular recesses. These are really capacious rooms, and
from the windows thereof the view of the Embankment Gardens and
the Thames is considerable compensation for the tediousness and
deviousness of the approach. The "cottages" were originally attached
to the houses on the terrace above, and, until recent years, they
were inhabited. Now, however, the majority of them are let
separately and are used as stores or workshops. One of them,
however, is still occupied as a dwelling-place, and, whatever else it
may be, this habitation is certainly unique.
Underneath the "Adelphi cottages," and extending below the houses
of the terrace, and John, Robert, and Adam Streets, are the famous
arches, which few people, either Londoners, who know nothing of
their own city, or Americans, who are versed in the lore of our
ancient streets, have ever visited. Truth to tell, the expedition to the
Adelphi arches is not to be undertaken with too light a heart. The
gloomy recesses do not conduce to joy, and, although the foot-pad
has scant opportunity for indulging in his nefarious practices, he
would be a venturesome person, a stranger to these parts, who
would wander alone in this underground world after the sun, which
never enters these passages, had ceased to illumine the earth above.
This very darkness and dismalness has its advantages at times. When
Messrs. Coutts, for instance, moved from their old premises in the
Strand, there was much speculation as to the manner in which they
transferred their immense stock of securities, deeds, and other
valuables from one side of the road to the other. There was great talk
at the time of armies of detectives and the use of the early hours of
Sundays, and other vague suggestions were allowed to be
promulgated. It was assumed that the transference would take place
from one side of the road to the other, and it was thought that there
might be some audacious attempts at robbery. In reality, the matter
was quite simple and there was not the slightest danger of any attack
upon the priceless possessions. Far removed from the noisy Strand—
in regard to atmosphere and surroundings—there is an arch, dark
indeed, and shut off from the outer world by huge gates, which are
some distance away. Here, many feet below the surface of the
streets, is a secret entrance to the premises of the old bank. And
here, in absolute security, never dreamt of by the enterprising thief,
the carts were loaded with their treasures.
The actual removal of these valuables was effected with great ease.
The carts wended their innocent way through the dreary arches, in
front of the "cottages," and passed out by a "right of way"
underneath the Hotel Cecil, towards Blackfriars. Thus, the would-be
thief was deluded of his prey. This "right of way" marks the bottom of
Ivy Lane, which is still in existence. It runs from the Strand and
denotes the boundary of the Duchy of Lancaster and the City of
Westminster. Formerly, it was an open thoroughfare, but there is
now, at the Strand entrance as well as at the bottom, a gate. At the
river end, there was, in olden times, a bridge, or pier, called Ivy
Bridge. But I think that there must have been, not only a bridge in
the Strand, but that there was a stream which ran hence into the
Thames. John Stow, in his "Survey of London," first published in
1598, speaks of "Ivy Bridge, in the High Street, which had a way
under it leading to the Thames, the like as sometime had the Strand
Bridge." Now, the Strand Bridge was over the stream of St. Clement's
Well, and Strand Lane, like Ivy Lane, ran down to the river, and, like
it, there was a pier at the end. I am the more certain that there must
have been a river of sorts at the junction of Ivy Lane and the Strand,
because to this day, as I found in the course of a recent investigation,
a stream trickles under John Street and renders useless a large cellar.
Nothing can stop it. It percolates now, just as it has done ever since
the excavations made by the Brothers Adam in 1768. It is drained
away, but it is just sufficient to create a damp atmosphere which is
detrimental to the storing of wine.
Hundreds of thousands of bottles of wine—chiefly port, claret, and
burgundy—are in bins here, and a most admirable place for the
purpose it is. The underground Adelphi is absolutely dry—save for the
one spot mentioned—and the temperature does not vary five degrees
in the course of a year. Here, also, are many hundreds of cases of
champagne, and here the jaded Londoner—if he be sufficiently
favoured—might come and feast his eyes on some few dozens of
bottles of "white port"—a wine which is not in fashion in these
degenerate days, but which, I rejoiced to learn, is still sent hence to
a certain royal household. Strange as it may seem, there is a strong
air of royalty about these dimly-lit vaults. What between the secret
entrance to the old premises of the great bankers—Messrs. Coutts
are the bankers for his Majesty and for the Queen[1]—and the "white
port" which gives its benefit to illustrious persons of royal lineage,
there is a distinct feeling that one is moving on an exalted plane
when, paradoxical as it may seem, we are in this subterranean place.
The distinctly regal air which pervades these caves of silence may
have given rise to a certain statement that hereabouts—half a dozen
yards from the royal stock of "white port"—Lady Jane Grey was cast
into a dungeon deep and carried thence to the dreaded Tower, there
to be beheaded. But the "Nine Days' Queen" knew only her gardens
and her flowers when she lived in Durham House—the predecessor of
the Adelphi. Here, in May, 1553, the Duke of Northumberland married
his son, Lord Guildford Dudley, to Lady Jane, in pursuance of his
design for altering the succession from the Tudor to the Dudley
family. The unfortunate girl of scarce seventeen summers certainly
left Durham House for the Tower—but it was with great pomp and
circumstance, in order to be proclaimed Queen. Her execution
followed hard upon, but she knew not imprisonment in what is now
the Adelphi. On the other hand, the haunt of a wretched woman is
still to be seen in this gloomy spot. "Jenny's Holes" figure on the plan
to this day, and are not likely to be obliterated therefrom. Into one or
other of these places—recesses by the main arches—the outcast
came to sleep and, finally, to die; some say, indeed, that she was
murdered here. "Jenny" has no history, but the vague tradition of her
misery still haunts these "dark arches." Nor is the story at all
improbable. The "dark arches" are forbidding enough now, and, even
in the day-time, the sparse gas jets only serve to make darkness
visible. So recently as the early seventies, when Mr. George
Drummond came into the property, cows were kept in the
underground passages of the Adelphi.
Adelphi Terrace, Adam Street to the east, and John Street, which is
parallel with the terrace and the Strand, and in between, still retain
much of their old-world appearance. But at the western side of the
Adelphi changes are afoot. There is a new building, facing the river,
but stunted and barred from its proper height by that bugbear of the
modern builder, "ancient lights." Then, again, the Caledonian Hotel,
in Robert Street, has taken to itself a new storey, and has been
transmogrified into modern flats with—oh, shade of Adam!—bath-
rooms. The searcher after the picturesque in London architecture
might do worse than descend from the Strand, past the Tivoli. He will
then be on the site of one of the gateways of Old Durham House,
and, turning to the right, he will see a bridge of beautiful design. It
was built, in order to connect the Strand and Adelphi premises of the
bank, by Thomas Coutts, who procured a special Act of Parliament
for the purpose.
The entire Adelphi estate occupies a little over three acres and a
quarter, divided as follows:—
Superficial
feet.
Houses (only) 78,400
Roadways,
terrace, and 45,400
areas
Foreground 19,200
————
143,000
The names of two more noted inhabitants of the Adelphi have to be
included in this "History." The learned Vicesimus Knox (1752-1821),
who is best known to fame as the compiler of "Elegant Extracts"
(1789), lived at No. 1, Adam Street. The first floor of the same house
was the place of retirement, for a score of years, of George Blamire,
barrister-at-law, "of very eccentric habits, but sound mind." John
Timbs, in his "Curiosities of London," states that "no person was
allowed to enter his chamber, his meals and all communications being
left by his housekeeper at the door of his ante-room. He was found
dead in an arm-chair, in which he had been accustomed to sleep for
twenty years. He died of exhaustion, from low fever and neglect; at
which time his rooms were filled with furniture, books, plate,
paintings, and other valuable property." The eccentric habits are
evident; but the "sound mind" is a little doubtful.
Finally, I may state that I have followed the fortunes of my book,
and, after a brief excursion into the noisy part of the world on the
other side of Charing Cross, have returned to the quiet and
comparative solitude of the Adelphi, where tubes do not trouble and
motor buses do not annoy. "Sir," said Dr. Johnson, "when a man is
tired of London, he is tired of life." And I think that there is no part of
London of which a man can be in less apprehension of tiring than the
Adelphi. It is of London, yet away from it; in the heart of the world,
yet secluded. To know it is to love it.
Austin Brereton.
September, 1908.

FOOTNOTES:
[1] See page 212.

Contents
CHAPTER I PAGES

Anthony Bek, Bishop of Durham—The 1-20

Papal Legate and the Oxford Clergy—
Henry III. and the Earl of Leicester—
Prince Henry—The Author of
Philobiblon—Edward III.—Thomas
Hatfield—Henry VIII.—Cuthbert
Tunstall—Cranmer at Durham House
—Anne Boleyn—Henry VIII. and Anne
of Cleves feast at Durham House—
Dudley, Duke of Northumberland—
Lady Jane Grey—Queen Mary—Queen
Elizabeth—Philip Sidney—Sir Walter
 Raleigh—Elizabeth Throgmorton—
Glanville v. Courtney—Thomas
Egerton—Fire at Durham House—
Raleigh and his Pipe
CHAPTER II
The New Exchange—The Earl of Salisbury
proprietor—Opened by James I.—
Popular Allusions—The First Edition of
Othello published Here—Samuel
Pepys a Frequent Visitor—Henry
Herringman—Otway —Etherege—
Wycherley—Dryden—Addison—
Durham House Decay—Acquired by
the Earl of Pembroke—Various Public
Offices in Durham Yard—Charles II.
helps to extinguish a Fire Here—
Archbishop Le Tellier—Godfrey Kneller
—David Garrick, wine merchant—Dr
Johnson—Voltaire—Murder in the
21-
New Exchange
48
CHAPTER III
The Romantic Story of the White Milliner, 49-
74
otherwise the Beautiful Frances
Jennings, Duchess of Tyrconnel—Her
Youthful Escapade—Her Connection
with the New Exchange the subject of
a Play by Douglas Jerrold—Its Failure
and the Author's Disappointment
—"Nan" Clarges, afterwards Duchess
of Albemarle, sells Wash-balls in the
New Exchange—Her Burial in
Westminster Abbey—Sir William Read,
the Quack, cures "Wry Necks" in
Durham Yard—Demolition of the New
 Exchange—A Noted Book-shop—
Ambassadors reside Here
CHAPTER IV
Enter the Brothers Adam—Their
Marvellous Transformation of the
Ruins of Durham House and Yard into
the Present Adelphi—The Magnitude
of the Project—Opposition of the City
—Defeated by Special Act of
Parliament—The Adelphi Buildings
only completed by Aid of a Lottery—
The Adams explain their Position—
Robert Adam: His History—His Death
—James Adam—Some Poor Wit,
includin Walpole's, at the Expense of
75-
the Architects
99
CHAPTER V
The Society of Arts—Its Foundation—Its
Removal to the Adelphi in 1774—
James Barry and his Famous Paintings
—Visited in John Street by Burke and
Johnson—The Latter's opinion of his
Genius—Description of his Pictures for
the Society—The Work of the Society
—"Spot" Ward, the Inventor of
"Friar's Balsam"—Johnson speaks in
the Great Room—Forsaken by his
100-
"Flowers of Oratory"
122
CHAPTER VI
David Garrick—His Residence in the 123-
144
Adelphi—Founds the Drury Lane Fund
—His Last Appearance on the Stage—
 Honoured by Parliament—The
Friendship of Mr and Mrs Garrick for
Hannah More—Their Correspondence
—Garrick helps the Production of
Percy—Presents his Buckles to
Hannah More—The Production of
Percy—Garrick's Prologue gives
Offence—Garrick brings Hannah
More's Dinner from the Adelphi to the
"Turk's Head"—The Literary Club—His
Last Illness and Death
CHAPTER VII
Garrick's Funeral from the Adelphi—
Johnson's Opinion of Garrick: "A
Liberal Man"—His Death "Eclipsed the
Gaiety of Nations"—Topham Beauclerk
and Johnson—Mrs Garrick's famous
Dinner Party—Johnson and other
Celebrities Present—Described by
Hannah More and Boswell—Johnson's
Morning Visit to Adelphi Terrace—
Hannah More's Life Here—Another
Dinner Party—Death of Mrs Garrick—
Shakespeare's Gloves sent to Mrs
Siddons from the Adelphi—Goldsmith
writes from a Sponging-House to
Garrick in the Adelphi—Becket, the 145-
Bookseller 171
CHAPTER VIII
The celebrated Quack, Dr Graham—His 172-
191
Temple of Health in the Adelphi—
Satirised by Colman and Bannister
—"Vestina, the Rosy Goddess of
Health"—Emma Lyon, Lady Hamilton
 —Osborn's Hotel—The King and
Queen of the Sandwich Islands—Their
Death in the Adelphi—Isaac d'Israeli—
The Earl of Beaconsfield—Thomas
Hill, the Original of Paul Pry—Thomas
Hood and Charles Dickens—David
Copperfield and Pickwick—Ivy Lane—
The Fox-under-the-Hill—The Adelphi
"Dark Arches"
CHAPTER IX
The First Bankers—Middleton & Campbell,
predecessors of Coutts & Co., "at The
Three Crowns in the Strand"—Patrick
and John Coutts—Patrick and Thomas
Coutts in London—Death of James
Coutts—Enter Thomas Coutts—Letter
by Him—His Stern Character—Married
to Harriot Mellon—Susan Starkie and
"The Three Graces"—Sir Francis
Burdett—Angela Georgina Burdett—
The Duchess of St Albans—Anecdotes
of Thomas Coutts—His Personal
Appearance—Interior of the Bank—
The Chinese Wall-Paper—The Adelphi
Chapel—Illustrious Customers of
Messrs Coutts—Partners in the Firm—
The Wills of Thomas Coutts and the
Duchess of St Albans—The Savage 192-
Club—Thomas Hardy—E.L. Blanchard 217
CHAPTER X
York House—Francis Bacon—The Great 218-
234
Seal taken from Him—Lord Keeper
Egerton—The Duke of Buckingham,
King James' "Steenie"—Magnificence
 of his Entertainments—Contemporary
Descriptions—Bishop Goodman's
Praise—The Second Duke—Dryden's
Revenge—The "Superstitious Pictures"
of York House—Buckingham's
Marriage—Spanish, Russian, and
French Ambassadors Here—Visits by
Pepys and Evelyn—Duke of
Buckingham sells York House—His
Curious Condition of Sale—The Duke's
Litany
CHAPTER XI
The York Water-Gate—Inigo Jones'
Beautiful Work—Built for the Duke of
Buckingham—The Proposal for its
Removal—Satires on the Subject—The
Gate Neglected—Its Restoration—The
Water Tower—The West-end supplied
with Water from Here—The Steam
Engine—Samuel Pepys resides in
Buckingham Street—William Etty and
Clarkson Stanfield—Peter the Great
Lodges Here—His Love of Strong
Drink—The Witty Earl of Dorset—
David Hume and Jean Jacques
Rousseau—Moore writes to his
Publisher Here—The Father of Modern
Geology—A Great Actor dies Here—
The Original of Smollett's Hugh Strap
—David Copperfield's Chambers—
Evelyn lives in Villiers Street—Sir
Richard Steele—Zara acted Here—Mrs
Cibber—Misstatement by "Anthony 235-
Pasquin" 248
 CHAPTER XII
The Strand in 1353—St Mary Rounceval—
Northampton House—Earl of Surrey,
the Poet—Suffolk House—Suckling's
Ballade upon a Wedding—Algernon
Percy, Earl of Northumberland—The
Restoration planned at
Northumberland House—Lady
Elizabeth Percy—Her Romantic
Marriages—Murder of "Tom of Ten
Thousand"—The "Proud" Duke of
Somerset—Edwin and Angelina—
Goldsmith at Northumberland House
—Fire Here—Dr Percy's Library saved
—The Famous Lion—Demolition of the
House—The Duke's Lament—
Northumberland Avenue—Craven
Street—Benjamin Franklin—Sir Joshua
Reynolds—Heinrich Heine—The
Author of Rejecte Addresses—J.S.
249-
Clarke
272
APPENDIX
273-
Samuel Pepys and the Adelphi
284
284-
Hannah More and Garrick's Funeral
285
286-
INDEX
294

List of Illustrations
The Terrace, York Buildings, Adelphi Frontispiece
 The Adelphi (Durham Yard and the New Exchange) and Charing
Cross in 1755

Durham House. Salisbury House. Worcester House

The New Exchange, Strand

"The Buildings called the Adelphi," 1777

The Society of Arts, John Street, Adelphi

Garrick's House, 5 Adelphi Terrace

Adam Street, Adelphi

Adelphi Terrace in Garrick's Time

The Thames, from the Water Works, York Buildings, Adelphi

York Stairs and the Water Tower

The Society of Arts distributing its Awards

The Strand Entrance to Durham Yard

Ivy Lane, Strand (the boundary of the Duchy of Lancaster and the
City of Westminster)

Entrance to the Adelphi Arches

The Fox-under-the-Hill

York House, York Stairs, and Durham House

York Stairs and Water Works

Pepys' Library, Buckingham Street, Adelphi

St Mary Rounceval, the Original Site of Northumberland House

Suffolk (subsequently Northumberland) House

The Strand Front of Northumberland House in 1752

Northumberland House, from the Gardens

The Ball-room, Northumberland House

The Drawing-room, Northumberland House

Charing Cross, before the building of Northumberland Avenue

The Lion, Northumberland House

 AN ADAM DOOR.

The Literary History of the Adelphi and its

Neighbourhood

CHAPTER I
Anthony Bek, Bishop of Durham—The Papal Legate and the
Oxford Clergy—Henry III. and the Earl of Leicester—Prince
Henry—The Author of Philobiblon—Edward III.—Thomas
Hatfield—Henry VIII.—Cuthbert Tunstall—Cranmer at
Durham House—Anne Boleyn—Henry VIII. and Anne of
Cleves feast at Durham House—Dudley, Duke of
Northumberland—Lady Jane Grey—Queen Mary—Queen
Elizabeth—Philip Sidney—Sir Walter Raleigh—Elizabeth
Throgmorton—Glanville v. Courtney—Thomas Egerton—Fire
at Durham House—Raleigh and his Pipe.

It is my pleasant duty to relate in these pages the romantic story of

kings and queens, of prelates and princes, of book-writers and book-
sellers, of artists, architects, and actors, and of other players on life's
fitful stage who, for six centuries and a half, have contributed to one
of the most interesting chapters in the history of London. Within that
small space which has been known as the Adelphi since 1772, a
district so confined that it is contained within five hundred square
yards, came, in its earlier years, several bishops and other clerical
dignitaries, then that prince who was afterwards the fifth King Henry
of England, anon, amid much pomp and pageantry, King Henry VIII.
Queen Mary and Queen Elizabeth were familiar with it, and here
lived, for twenty years, Sir Walter Raleigh, who inhabited one of the
towers which is seen in Hollar's engraving of Durham House. Lady
Jane Grey went hence to the Tower and thence to the scaffold.
Dryden alluded to it in one of his plays. Voltaire drank wine here, and
its memory is hallowed by Dr Johnson, Oliver Goldsmith, Sir Joshua
Reynolds, and a host of other celebrities. Here David Garrick began
his career, and here, curiously enough, he ended it, the funeral
procession of the "poor player" reaching from the Adelphi to
Westminster Abbey, those who followed him to the grave numbering
many men of rank and genius, including Johnson, and a large
concourse of the general public who grieved for the loss of the great
actor.
The history of the world-famous banking firm of Coutts & Co. is
indelibly associated with the Adelphi. Dickens, when a boy, prowled
about its dark arches—until lately, one of the most degraded spots in
London—and last, though not least, the brothers Adam, to whom
London owes several architectural triumphs, in addition to the
Adelphi, claim our attention. It is said that at a public dinner, at the
beginning of the last century, a worthy alderman whose knowledge of
Greek was very vague, was much struck by the toast, in reference to
two royal brothers—George IV. and the Duke of York—of "the
Adelphi."[2] When it came to the alderman's turn to speak, he said
that, as they were on the subject of streets, he would "beg leave to
propose 'Finsbury Square.'" In somewhat similar manner, before we
get to the Adelphi, we must go back to its origin, and this takes us to
the thirteenth century.
Durham House, which, with its grounds, formerly occupied the entire
site of the Adelphi, was the town residence of Anthony Bek
(otherwise Anthony de Beck or Bec), Bishop of Durham in the reign
of Edward I. So it is affirmed by Pennant, and there is no reason to
doubt the assertion. Some mistakes have arisen on this point, in
consequence, as it appears to me, of there having been two men of
the same name, both of whom were bishops. Their ancestor, Walter
Bek, came to England with William the Conqueror, and from his three
sons sprang three great Lincolnshire families: Bek of Eresby, Bek of
Luceby, and Bek of Botheby. Now, Bishop Antony Bek the second
(1279-1343), son of Walter Bek of Luceby, constable of Lincoln
Castle, was at one time Bishop of Lincoln, and, in 1337, Bishop of
Norwich. But Antony Bek, son of Walter Bek, baron of Eresby, was
appointed to the see of Durham in 1283. He was intimately
associated with Edward I., being one of his chief advisers during the
negotiations regarding Baliol, and of great assistance to him in his
Scottish expeditions of 1296 and 1298. Owing to a dispute with the
prior of the convent of Durham, he was deprived of certain of his
rights by the king (but regained them on application to the Pope). As
this, however, occurred in the year 1300, it may safely be assumed
that Antony Bek had occupied Durham House before that event.
But there was a Durham House even earlier than this of Antony
Bek's, if we are to credit an account given by Thomas Fuller. Here, in
1238, the papal legate, Otho, was staying, and hither he summoned
the English bishops in order to debate as to what "further steps
should be taken respecting the churches and schools of Oxford,
which he had laid under interdict on account of the scholars having,
when the legate was staying at Oseney, killed his brother and clerk of
the kitchen in an affray,"[3] the legate himself being obliged to fly
from the city. At the intercession of the bishops, the legate assented
to pardon the university on condition of the clergy and scholars
making their "solemn submission" to him. As a result, the offenders
"went from St Paul's in London to Durham House in the Strand, no
short Italian, but an English long, mile, all on foot; the bishops of
England, for the more state of the business, accompanying them, as
partly accessory to their fault, for pleading on their behalf. When they
came to the Bishop of Carlisle's house, the scholars went the rest of
the way barefoot, sine capis et mantulis, which some understand,
'without capes or cloaks.' And thus the great legate at last was really
reconciled to them."[4]
Some of these old chronicles are not always to be relied upon in the
matter of dates: "This howse called Durham, or Dunelme Howse, was
buylded in the time of Henry 3, by one Antonye Becke, B. of Durham.
It is a howse of 300 years antiquitie; the hall whereof is statelie and
high, supported with lofty marble pillars. It standeth on the Thamise
veriye pleasantlie." So wrote one historian in 1593. But Henry III.
died in 1272, eleven years before Bek was made Bishop of Durham.
That there was a Durham House of sorts before Bek's time is pretty
certain, although it was not the one that is attributed to that bishop.
The story has often been told of Henry III., in 1258, being caught in
a thunderstorm on his way down the Thames on his barge. At that
time, the Earl of Leicester was the head of the barons who were
opposed to the king, and it is said that he was then in occupation of
Durham House (we have already seen that the papal legate was
installed there twenty years earlier). Be this as it may, the king
sought shelter from the storm, and, as the royal barge approached
the shore, the Earl of Leicester went forth and endeavoured to allay
any fears that the king might have felt, saying, "Your Majesty need
not be afraid, for the tempest is nearly over." But the king, being
moved to wrath, fiercely exclaimed, "Above measure, I dread thunder
and lightning, but, by the head of God, I am more in terror of thee
than of all the thunder and lightning in the world." Though this story
may be doubted, one early royal memory of Durham House is that of
Prince Henry (Henry V.), who, in 1411, "lay at the bysshoppes inne of
Darham for the seid day of his comming to towne unto the Moneday
nest after the feste of Septem fratum."[5]
That most correct of London historians, John Stow, sets down the
fourteenth century as the date of Durham House. "On the south side
of which street" (meaning the Strand, which had no name in Stow's
time), he says, "in the liberties of Westminster (beginning at Ivy
Bridge), first is Durham House, built by Thomas Hatfield, Bishop of
Durham, who was made bishop of that see in 1345, and sat bishop
there thirty-six years." But we have already seen from Fuller, whose
Church History of Britain—from which the quotation in regard to the
papal legate, Otho, is taken—was written in 1655, fifty years after
Stow's death, that there was a Durham House in 1238.
And this brings me to a curious point. Thomas Pennant, whose
Account of London affords much entertaining reading, has an
amusing disquisition on the word "palace." He writes: "That the word
is only applicable to the habitations of princes, or princely persons,
and that it is with all the impropriety of vanity bestowed on the
houses of those who have luckily acquired money enough to pile on
one another a greater quantity of stones and bricks than their
neighbours. How many imaginary Parks have been formed within
precincts where deer were never seen! and how many houses
misnamed Halls which never had attached to them the privilege of a
manor!" Leigh Hunt took the "lively Pennant," as he dubs him, to task
on this point: "Unless the words palazzo and piazza are traceable to
the same root, palatium (as perhaps they are), place does not of
necessity mean palace; and palace certainly does not mean
exclusively the habitation of princes and princely persons (that is to
say, supposing princeliness to exclude riches), for in Italy, whence it
comes, any large mansion may be called a palace; and many old
palaces there were built by merchants."[6] But the disquisition does
not really alter the fact that the proper name, that is to say, the
original one, should be Durham House; we have the excellent
authority of Stow and Fuller on this head. The residences in London
of the bishops were almost invariably called "House"—certainly not
"palace." Thus, Worcester House, which is now marked by the Savoy,
originally belonged to the see of Carlisle, and is "the Bishop of
Carlisle's House" which is alluded to in the extract from Fuller. York
House, which stood to the west of Durham House, was originally the
town inn or residence of the Bishop of Norwich, and, subsequently, in
Queen Mary's reign, of Heath, Archbishop of York. In the Aggas map
of London in 1563, which is the frontispiece to Pennant's "Account,"
Duresme Place and York Place are given, but that the name in its
earlier years was Durham House there is no doubt. The London
County Council has lately (1906) perpetuated the name by changing
Durham Street to Durham House Street.
THE ADELPHI (DURHAM YARD AND THE NEW EXCHANGE) AND CHARING CROSS,
IN 1755.
One of the earliest of the literary inhabitants of Durham House was
the learned Richard de Bury (1281-1345), son of Sir Richard
Aungerville. He was tutor to Edward III., when Prince of Wales, and,
subsequently, was of the king's household. He was Dean of Wells and
Bishop of Durham in 1333, lord chancellor from September 1334 to
July 1335, and lord high treasurer in 1337. He was employed by the
king in Paris and in Hainault in 1336, and, in 1337 and 1342, in
Scotland. It is pleasant to think that he wrote his Philobiblon during
his residence by the Thames. At any rate, we may be sure that so
learned and so useful a man, one who had the confidence of the king
for so long, was visited here by Edward III.
Another name of note associated with Durham House is that of
Thomas Hatfield, already alluded to by Stow as having built that
structure. He probably added to it, or he may have rebuilt it. He was
a great prelate, and, in addition to the bishopric of Durham, which he
held from 1345 until his death in 1381, he was made keeper of the
Privy Seal in 1343, and, in 1346 and 1355, he accompanied Edward
III. to France. In Durham, he built part of the south side of the
cathedral choir and the hall of the castle, hence, possibly, the credit
given to him by Stow of building the Thames-side Durham House. His
learned Survey of Durham was edited by the Rev. William Greenwell
in 1856.
It is a far cry from the joyous days of Prince Henry to the turbulent
times of Henry VIII., but the old chronicles do not contain any
mention of Durham House during that lengthy period. In the reign of
the latter king, the then Bishop of Durham "conveyed the house to
the King in fee"; in other words, the noble Henry appropriated the
property to his own uses. He had the saving grace, however, to give
to the see of Durham, in exchange, some houses in Cold Harbour
(now marked by Upper Thames Street), and elsewhere. The exact
date of the transfer is unknown. The history of this bishop, who was
made to surrender Durham House to King Henry, is curious. Cuthbert
Tunstall, or Tonstall, was Master of the Rolls, and bishop successively
of London and Durham. Extolled by Erasmus, and the friend of Sir
Thomas More, he was learned in Greek, Hebrew, mathematics, and
civil law. Harrow-on-the-Hill had him for rector in 1511, he was
prebendary of Lincoln in 1514, archdeacon of Chester in the year
following, ambassador to the Prince of Castile at Brussels, 1515-
1516, Master of the Rolls in 1516, prebendary of York in 1519, and
ambassador to Charles V. in 1519, and again in 1525. He was Bishop
of London from 1522-1530, keeper of the privy seal in 1523, and
Bishop of Durham in 1530. It must have been after the latter year
that he transferred Durham House to Henry VIII. Accused of inciting
to rebellion, 1550, he was deprived of his bishopric of Durham by
Edward VI., in 1552. Queen Mary, however, restored him immediately
on her accession, and he remained in possession of Durham House—
which Mary had also restored to the see—until, in the year of his
death, 1559, he was again deprived by Queen Elizabeth, to whom he
had refused the oath of supremacy.
A very interesting chapter in the history of Durham House came into
existence, thanks to its acquisition by Henry VIII., who granted it to
the Earl of Wiltshire (1477-1539), Thomas Boleyn, father of Queen
Anne Boleyn. It is not impossible that the Earl of Wiltshire was in
occupation of Durham House during the childhood of his daughter: at
any rate, it is certain that Anne's daughter, the Princess Elizabeth,
afterwards Queen, resided here.
Through Henry VIII. we get a glimpse of Cranmer at Durham House,
for that worthy wrote to the Earl of Wiltshire bidding him "let Doctor
Cranmer have entertainment in your house at Durham Place for a
time, to the intent he may bee there quiet to accomplish my request,
and let him lack neither bookes, ne anything requisite for his studies."
[7] Cranmer attended the Earl of Wiltshire as ambassador to Charles
V. in 1530, and it is probable that he lodged in Durham House in
1533, for in that year he returned to England, gave formal sentence
of the invalidity of the king's marriage with Catharine of Aragon, and
pronounced King Henry's marriage with Anne Boleyn to be lawful. So
that it is easy to imagine that the king's "request" occupied Cranmer's
thoughts at Durham House, and that Henry came here in order to
confer with him.
That Henry VIII. was familiar with Durham House there is no room
for doubt, for, as the pious chronicler, Stow, quaintly puts it, "in the
year of Christ 1540," that being the thirty-second year of Henry's
reign, "on May-day, a great and triumphant jousting was holden at
Westminster, which had been formerly proclaimed in France,
Flanders, Scotland, and Spain, for all comers that would undertake
the challengers of England; which were, Sir John Dudley, Sir Thomas
Seymour, Sir Thomas Ponings, and Sir George Carew, knights, and
Anthony Kingston and Richard Cromwell, esquires; all which came
into the lists that day richly apparelled, and their horses trapped all in
white velvet. There came against them the said day forty-six
defendants or undertakers—viz., the Earl of Surrey, foremost, Lord
William Howard, Lord Clinton, and Lord Cromwell, son and heir to
Thomas Cromwell, Earl of Essex, and chamberlain of England, with
other; and that day, after the jousts performed, the challengers rode
unto this Durham House, where they kept open household, and
feasted the King and Queen, with her ladies, and all the court. The
second day Anthony Kingston and Richard Cromwell were made
knights there. The third day of May the said challengers did tourney
on horseback with swords, and against them came forty-nine
defendants—Sir John Dudley and the Earl of Surrey running first,
which at the first course lost their gauntlets; and that day Sir Richard
Cromwell overthrew Master Palmer and his horse in the field, to the
great honour of the challengers. The fifth of May the challengers
fought on foot at the barriers, and against them came fifty
defendants, which fought valiantly; but Sir Richard Cromwell
overthrew that day at the barriers Master Culpepper in the field; and
the sixth day the challengers brake up their household. In this time of
their housekeeping they had not only feasted the king, queen, ladies,
and all the court, as it is afore showed, but also they cheered all the
knights and burgesses of the common house in the parliament, and
entertained the Mayor of London, with the aldermen and their wives,
at a dinner, etc. The king gave to every of the said challengers and
their heirs for ever, in reward of their valiant activity, one hundred
marks and a house to dwell in, of yearly revenue, out of the lands
pertaining to the hospital of St John of Jerusalem, which he had
confiscated."
From the merry-makings of "bluff King Hal" we turn to the more
sober employment of Durham House. Here, in 1550, were lodged the
French ambassador to Edward VI., Mons de Chastillon, and his
colleagues, the house being "furnished with hangings of the kings for
the nonce." In this year, also, Edward VI. granted Durham House for
life, or until she was otherwise advanced, to the Lady Elizabeth,
afterwards Queen Elizabeth; but, "in some way, it passed from the
Princess to Dudley, Duke of Northumberland, and was the principal
London house when Edward VI. died." I do not think that it is very
difficult to account for the transition. During the short reign of
Edward VI., we find it stated in Pennant that "the mint was
established in this house, under the management of Sir William
Sharrington, and the influence of the aspiring Thomas Seymour, lord
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

textbookfull.com