Download Full Version ebook - Visit ebooknice.

com

(Ebook) Systems Architecture Modeling with the

Arcadia Method: A Practical Guide to Capella by
Pascal Roques ISBN 9781785481680, 1785481681

https://ebooknice.com/product/systems-architecture-modeling-
with-the-arcadia-method-a-practical-guide-to-capella-6986160

Click the button below to download

DOWLOAD EBOOK

Discover More Ebook - Explore Now at ebooknice.com

 Instant digital products (PDF, ePub, MOBI) ready for you
Download now and discover formats that fit your needs...

Start reading on any device today!

(Ebook) UML in Practice: Art of Modeling Software Systems

Demonstrated through Worked Examples and Solutions by
Pascal Roques ISBN 9780470092798, 9780470848319,
0470092793, 0470848316
https://ebooknice.com/product/uml-in-practice-art-of-modeling-
software-systems-demonstrated-through-worked-examples-and-
solutions-1988790
ebooknice.com

(Ebook) Object-Role Modeling Fundamentals: A Practical

Guide to Data Modeling with ORM by Terry Halpin ISBN
9781634620741, 1634620747
https://ebooknice.com/product/object-role-modeling-fundamentals-a-
practical-guide-to-data-modeling-with-orm-5526556

ebooknice.com

(Ebook) Bayesian Analysis with Python: A practical guide

to probabilistic modeling by Osvaldo Martin ISBN
9781805127161, 1805127160
https://ebooknice.com/product/bayesian-analysis-with-python-a-
practical-guide-to-probabilistic-modeling-56265722

ebooknice.com

(Ebook) A Practical Guide to Verilog-A: Mastering the

Modeling Language for Analog Devices, Circuits, and
Systems by Slobodan Mijalkovi■ ISBN 9781484263501,
1484263502
https://ebooknice.com/product/a-practical-guide-to-verilog-a-
mastering-the-modeling-language-for-analog-devices-circuits-and-
systems-46150156
ebooknice.com
(Ebook) Simulation and Modeling of Systems of Systems by
Pascal Cantot; Dominique Luzeaux ISBN 9781118616659,
1118616650
https://ebooknice.com/product/simulation-and-modeling-of-systems-of-
systems-51258118

ebooknice.com

(Ebook) A Practical Psychoanalytic Guide to Reflexive

Research: The Reverie Research Method by Joshua Holmes
ISBN 9781138607040, 1138607045
https://ebooknice.com/product/a-practical-psychoanalytic-guide-to-
reflexive-research-the-reverie-research-method-7415766

ebooknice.com

(Ebook) Practical Building Energy Modeling with

OpenStudio: A Guide for Students and Professionals by Team
Educohack ISBN 0000000000000
https://ebooknice.com/product/practical-building-energy-modeling-with-
openstudio-a-guide-for-students-and-professionals-55602044

ebooknice.com

(Ebook) Systems Engineering: Holistic Life Cycle

Architecture Modeling and Design with Real-World
Applications by Sandra Furterer ISBN 9780367532802,
0367532808
https://ebooknice.com/product/systems-engineering-holistic-life-cycle-
architecture-modeling-and-design-with-real-world-applications-35169026

ebooknice.com

(Ebook) Modeling of Combustion Systems: A Practical

Approach by Joseph Colannino ISBN 0849333652

https://ebooknice.com/product/modeling-of-combustion-systems-a-
practical-approach-2111452

ebooknice.com
Systems Architecture Modeling
with the Arcadia Method

A Practical Guide to Capella

Pascal Roques
Implementation of Model Based System Engineering Set

coordinated by
Pascal Roques
Table of Contents

Cover

Title page

Copyright

Foreword

Preface
Aims of this book

Structure of the book

Acknowledgments

1: Reminders for the Arcadia Method

Abstract

1.1 Novelties, strengths and principles

1.2 Architecture levels and associated concepts

1.3 Main types of Arcadia diagrams

2: Capella: A System Modeling Solution

 Abstract

2.1 Radius considered and stakes involved

2.2 Principles of the tool

3: Complete Example of Modeling with Capella: Operational

Analysis
Abstract

3.1 Presentation of the case study and project creation

3.2 Operational Analysis

4: Complete Example of Modeling with Capella: System Analysis

Abstract

4.1 Main concepts and diagrams

4.2 Going from the Operational level to the System level

4.3 System Capabilities

4.4 Functional Analysis at the System level

4.5 Functional Chains at the System level

4.6 Allocation of Functions to the System or to Actors

4.7 System-level Scenarios

4.8 Modes and States at the System level

4.9 Data modeling at the System level

5: Complete Example of Modeling with Capella: Logical Architecture

Abstract

5.1 Main concepts and diagrams

 5.2 Moving from the System level to the Logical level

5.3 Logical Components

5.4 Allocation of the Logical Functions

5.5 System to Subsystem Transition

5.6 Scenarios on the Logical level

5.7 Logical subcomponents

6: Complete Example of Modeling with Capella: Physical

Architecture
Abstract

6.1 Main concepts and diagrams

6.2 Moving from the Logical level to the Physical level

6.3 Physical Components

6.4 Allocating the Functions to the Physical Components

6.5 Functional Chains on the Physical level

6.6 Return to the Physical Components and the structural links

6.7 Integrating Specialty Viewpoints

6.8 Replicable and Replica Elements

7: Complete Example of Modeling with Capella: EPBS

Abstract

7.1 Main concepts and diagrams

7.2 Moving from the Physical level to the EPBS level

7.3 Configuration Item

 7.4 Traceability between Configuration Items and Physical Components

Conclusion: Capella’s Key Strengths

Summary of key strengths

The Capella ecosystem

Bibliography

Index
Copyright
First published 2018 in Great Britain and the United States by ISTE
Press Ltd and Elsevier Ltd

Apart from any fair dealing for the purposes of research or private
study, or criticism or review, as permitted under the Copyright,
Designs and Patents Act 1988, this publication may only be
reproduced, stored or transmitted, in any form or by any means,
with the prior permission in writing of the publishers, or in the case
of reprographic reproduction in accordance with the terms and
licenses issued by the CLA. Enquiries concerning reproduction
outside these terms should be sent to the publishers at the
undermentioned address:

ISTE Press Ltd

27-37 St George’s Road
London SW19 4EU
UK

www.iste.co.uk

Elsevier Ltd
The Boulevard, Langford Lane
Kidlington, Oxford, OX5 1GB
UK

www.elsevier.com

Notices
Knowledge and best practice in this field are constantly changing.
As new research and experience broaden our understanding,
changes in research methods, professional practices, or medical
treatment may become necessary.

Practitioners and researchers must always rely on their own

experience and knowledge in evaluating and using any
information, methods, compounds, or experiments described
herein. In using such information or methods they should be
mindful of their own safety and the safety of others, including
parties for whom they have a professional responsibility.

To the fullest extent of the law, neither the Publisher nor the
authors, contributors, or editors, assume any liability for any injury
and/or damage to persons or property as a matter of products
liability, negligence or otherwise, or from any use or operation of
any methods, products, instructions, or ideas contained in the
material herein.

For information on all our publications visit our website at

http://store.elsevier.com/

© ISTE Press Ltd 2018

The rights of Pascal Roques to be identified as the author of this
work have been asserted by him in accordance with the Copyright,
Designs and Patents Act 1988.

British Library Cataloguing-in-Publication Data

A CIP record for this book is available from the British Library
Library of Congress Cataloging in Publication Data
A catalog record for this book is available from the Library of
Congress
ISBN 978-1-78548-168-0

Printed and bound in the UK and US

Foreword
The Arcadia method appeared during the year 2007 as part of the
Thales Airborne Systems: this structured engineering method, which
aimed to define and validate the architectural design of complex
systems, was immediately followed by the Melody Advance tool,
which aided its implementation.
The method has since demonstrated its benefits in all of Thales’
areas of excellence (Defense, Space, Aeronautics, Land transport,
Security, etc.) providing structure to the collaborative work of those
involved, who are often numerous during the definition phase of the
system. The Arcadia method and the Melody Advance tool have
progressively been deployed within the Group, and later externally.
After going public, Melody Advance entered the world of Open
Source software, under the name Capella.
Pascal has followed this deployment since the early stages of the
project, and his collaboration with Thales University, which started
10 years ago, continues to this day, namely in the form of training
that he provides on the Arcadia method and the Melody
Advance/Capella tool.
Capella provides specific solutions to the problems of System
Engineers and System Architects. It allows for the construction and
maintenance of global and coherent systemic visions that integrate
operational, system, logical and physical views, as well as the
associated viewpoints of all stakeholders.
This book will allow you to experience Capella, and there is
nothing better for this than a practical case… As Albert Einstein said:

“Knowledge is experience, everything else is just information”.

 The example that illustrates all of the aspects involved in the
implementation of Capella is EOLE, a balloon probe system whose
main goal is to provide meteorological data to different types of
users. It is representative through:
– its multiple stakeholders and user types;
– its various types of component (hardware, software, etc.);
– its types of functional and non-functional requirements.
It is this same case study that highlights the training provided by
Thales University, which puts trainees very quickly into concrete
situations.
Since 2008, Pascal has trained more than 1,200 Thales employees in
the different sites in France and Europe (more than 100 sessions). It
is this passion and his qualities as a teacher that will guide you
throughout this logical and structured application of Capella.
So jump right in and follow Pascal in the fascinating experience of
modeling.
Odile Mornas, Thales Global Services, INCOSE ESEP Certified
Preface
Pascal Roques October 2017

Aims of this book

The Arcadia modeling method was designed by Thales for its own
needs. Since 2011 it has been applied to a growing number of
projects over a variety of domains (avionic, rail systems, defense
systems in all environments, satellite systems and ground stations,
communication systems, etc.) and in many different countries.
This method is supported in its modeling aspect by a dedicated tool
that responds to the constraints present during full-scale
deployment in an operating context. This tool, called Capella
(Melody Advance internally in Thales), is currently available free of
charge for the system engineering community as an Open Source
solution (www.polarsys.org/capella/).
My goal in this work is the introduction of this new environment
for modeling in system engineering. Based on my experience as a
modeling consultant in a number of domains, my pedagogical
experience as an UML and SysML trainer for over 15 years, and as a
provider of training with Melody Advance within the Thales group
(more than 120 training sessions in France and in Europe, more than
1,500 trainees), I hope to show the advantages and assets of the
Capella tool based on the Arcadia method.
This book is first of all aimed at system engineering professionals,
those who are in charge of complex systems, both material and
software, whether in the domains of aeronautics, space, energy,
transport, defense, automotive, etc.
Structure of the book
Chapter 1 constitutes a reminder of the Arcadia method, described
in detail in another book from the same series: “Model-based System
and Architecture Engineering with the Arcadia Method” [VOI 18].
Chapter 2 presents the stakes and principles involved with the
Capella tool, which implements the Arcadia method. First, we shall
establish the perimeter that is targeted by the tool, as well as its
origin. Next we shall describe the principles behind the human–
machine interface, as well as the varying nature of the diagrams
involved.
Chapter 3–7 demonstrate concrete use of the Capella tool (and
therefore also the Arcadia method) in a realistic case study from
Operational Analysis to EPBS. The difficulty in system engineering is
often finding a case study that is representative enough, but not too
complex or too specific to a single technical domain. In the context of
this book, we have reused and adapted an example that we used
over a hundred times during Capella training sessions carried out
first at Thales, and later outside of Thales.
The conclusion provides a summary of the important points
regarding the Capella tool and presents its “under construction”
ecosystem mainly based on the collaborative project Clarity
(www.clarity-se.org/) and the future Capella IC (Industry
Consortium).

Acknowledgments
This work would doubtless not have seen the light of day without
the support of the Clarity project, as part of which I was asked to
write a book on the Capella tool, with the objective of spreading
awareness about it. A big thanks first of all to Daniel Exertier (Thales
Corporate) for involving me in this adventure, and to all the others
from Thales with whom I have worked over the course of these
many years as a trainer with the internal tool Melody Advance.
While there are too many to name all of them here, I would like to
thank in particular Jean-Luc Voirin (Thales Airborne Systems) and
Stéphane Bonnet (Thales Corporate); Philippe Lugagne, Stéphanie
Cheutin and Laetitia Saoud (Thales Alenia Space in Toulouse); as
well as Patricia Pancher and Odile Mornas (Thales University).
Thank you to my technical proofreaders for their astute
comments:
– Olivier Casse (expert in embedded system modeling languages and tools, veteran of I-
Logix/Telelogic and Atego/Artisan);
– Jérôme Montigny (Capella evangelist with Continental Automotive in Toulouse);
– Benoit Viaud (Clarity project member with Artal in Toulouse).
1

Reminders for the Arcadia

Method

Abstract
System engineers have been making use of modeling techniques for a long
time. Structured analysis and design technique (SADT) and structured
analysis for real time (SA/RT) are some of the best known of these, and date
back to the 1980s. There are many other approaches based on Petri nets or
finite state machines. However, these techniques are also limited by their
range and expressivity, as well as by the difficulty in integrating them with
other formalisms and with requirements.

Keywords
Arcadia; Computer software configuration item (CSCI); End Product Breakdown
Structure (EPBS); Functional Chain; Functional Exchanges; Logical
Components; Structured analysis and design technique (SADT); SysML
language

1.1 Novelties, strengths and principles

1.1.1 History
System engineers have been making use of modeling techniques for
a long time. Structured analysis and design technique (SADT) and
structured analysis for real time (SA/RT) are some of the best known
of these, and date back to the 1980s. There are many other
approaches based on Petri nets or finite state machines. However,
these techniques are also limited by their range and expressivity, as
well as by the difficulty in integrating them with other formalisms
and with requirements.
The rise of UML [ROQ 04] in the world of software and the
industrial effort toward developing the tools that accompany it have
naturally led to its use being considered in system engineering.
However, due to a design process that was strongly influenced by its
intended use in object programming, the language was, at least in
the early versions, not particularly adapted to modeling complex
systems, and was therefore not well suited to system engineering.
An interesting attempt was the publication of a UML variant for
system engineering in 2006–2007. This new language, called SysML
[CAS 18], was strongly inspired by version 2 of UML, but added the
possibility of representing system requirements, non-software
elements (mechanical, hydraulic, sensors, etc.), physical equations,
continuous flows (matter, energy, etc.) and allocations.
Unfortunately, in practice it has been shown that the filiation of the
SysML language to UML often leads to difficulty in terms of
comprehension and use for system engineers who are not also
computer scientists.
This is the reason that led Thales to define the Arcadia method
[VOI 16, VOI 17], along with its underlying formalism, for its own
needs. It has been applied since 2011 in a growing number of
projects across a great variety of domains (avionics, railway systems,
defense systems in all fields, air traffic control, command control,
area surveillance, complex sensor systems, satellite systems and
ground stations, communications systems, etc.), and in many
countries (France, Germany, United Kingdom, Italy, Australia,
Canada, etc.).
The modeling aspect of the method is supported by a dedicated
tool that responds to the constraints involved with full-scale
application in an operational context. This tool, called Capella
(Melody Advance internally at Thales), is currently freely available
for the engineering community as an Open Source application.
1.1.2 Founding principles
Today’s complex systems are limited by a number of requirements
or constraints, often concurrently, and sometimes contradictorily:
functional requirements (services expected by the users), and
nonfunctional requirements (security, operating safety, mass,
scalability, cost, etc.). The initial engineering phases of such systems
are critical as they condition the aptitude of the architecture used to
answer the needs of the clients, as well as the proper distribution of
the requirements toward the components, arising from the
architecture used. In order to properly handle delays and costs, it is
vital to be able to verify the adequacy of the solution with regard to
needs from the system design phase, and to minimize the risk of
coming across limitations of the solution – thus jeopardizing the
architecture – at a more or less advanced stage of development, or
even during integration or qualification of the system.
Current practice in systems engineering in the last few years has
made use (and rightly so) of a formalization of needs and
expectations expressed by the client, in the form of textual
requirements, which are then traced (manually) during realization to
justify them in relation to the client needs. The limitations of this
approach arise mainly from the fact that non-formalized, textual
requirements make it harder to verify their coherence and their
completeness. Moreover, they are confined to the expression of need
and are therefore poorly adapted to describing the solution and to
mastering its complexity, or to structuring the engineering. This is
one of the reasons that led Thales to the development and
deployment of an innovative approach called Arcadia.
Arcadia is a structured engineering method aimed at defining and
validating the architecture of complex systems. It favors
collaborative work between all stakeholders – of which there are
often many – involved in the engineering (or definition) phase of the
system. It allows for iterations to be carried out from the definition
phase that will help converge the architecture toward congruence
with all of the needs identified.
 Textual requirements are still present and used as a main
contribution toward expressing need at the start of the engineering
process. As such, Arcadia takes its place as a major support for
engineering and its control, relying on a formalization of the analysis
of need, whether operational, functional, or non-functional
(functions expected of the system, functional chains, etc.), and on the
definition/justification of the architecture based on this functional
analysis.
The general principles of Arcadia are the following:
– all of the engineering stakeholders share the same methodology,
the same information, the same description of the need and the
product in the form of a shared model;
– each specialized type of engineering (for example security,
performance, cost and mass) is formalized as a “viewpoint” in
relation to the requirements from which the proposed
architecture is then verified;
– the rules for the anticipated verification of the architecture are
established in order to verify the architecture as soon as possible;
– co-engineering between the different levels of engineering is
supported by the joint elaboration of models, and the models of
the different levels and specialties are deducted/validated/linked
one to the other.
To summarize, Arcadia possesses innovative characteristics that
are yet to be properly demonstrated in its domain:
– it covers all of the structuring activities of engineering, from
capturing the operational needs of the client to
integration/verification/validation (IVV);
– it takes into account the multiple levels of engineering, as well as
the efficient collaboration between them (system, subsystem,
software, hardware, etc.);
– it integrates coengineering with specialty engineering types
(security, safety, performance, interfaces, logistics, etc.) and of
IVV;
– it is based on the use of models that are not only descriptive, but
also able to validate the definition and the properties of the
 architecture, and that constitute the main support for
coengineering between the teams involved;
– it has successfully passed the test of applicability in real-size
projects and in constrained operational situations, as it is
currently being used in several dozen large projects in various
divisions (and countries) of Thales.

FIGURE 1.1 Coengineering with Arcadia

1.2 Architecture levels and associated

concepts
1.2.1 Overview

Note
The vocabulary that we shall explain in the following sections is
the one used in version 1.1 of the Capella tool.
 The different working levels of Arcadia are the following:
– OPERATIONAL ANALYSIS: “What the users of the system need
to accomplish”:
- analysis of the issues of operational users by identifying the
actors that must interact with the system, their activities and
their interactions with each other.
– ANALYSIS OF THE SYSTEM NEEDS: “What the system has to
accomplish for the users”:
- external functional analysis as a response to identify the system
functions needed by its users (e.g. “calculate the optimal path”
and “detect a threat”), limited by the non-functional properties
asked for.
– LOGICAL ARCHITECTURE: “How the system will work to fulfill
expectations”:
- internal functional system analysis: which are the subfunctions
that must be carried out and put together to establish the
“user” functions identified during the previous stage;
- identification of the logical components that carry out these
internal subfunctions, by integrating the non-functional
constraints that we choose to deal with at this level.
– PHYSICAL ARCHITECTURE: “How the system will be
developed and built”:
- the goal of this level is the same as that of the logical
architecture, except that it defines the final architecture of the
system as it must be created;
- it adds the functions required by the implementation and by the
technical choices, and highlights behavioral components (e.g.
software components) that carry out these functions. These
behavioral components are then implemented using
implementation components (e.g. processor board), which
provide the necessary material resources.
– EPBS (End Product Breakdown Structure) AND INTEGRATION
CONTRACTS: “What is expected from the provider of each
component”:
- this step deduces from the physical architecture the conditions
that each component must fulfill to satisfy the architecture
 p y
design constraints and limitations, established in the previous
phases.

FIGURE 1.2 The main engineering levels of Arcadia

It must be noted that the method does not always have to be top-
down in nature, but can also perfectly be bottom-up, for example if
we start with an existing system that is to be worked on. The
question relates more to architectural levels than to phases or steps.
Moreover, not all architectural levels are mandatory for all
projects. Operational Analysis, Logical Architecture and EPBS are
considered to be optional, depending on the complexity of the
system under study and the goals of the model.

1.2.2 Operational Analysis

The highest level of the Arcadia method is Operational Analysis
(“what the users of the future system need to accomplish”). The goal
here is to focus on the identification of the needs and objectives of
future users of the system in order to guarantee the adequacy of the
system faced with these operational needs.

Note
At this level, the system is not (yet) recognized as a modeling
element. It will only be recognized as such from the System
Analysis level onward.

This level can be treated as a model of the jobs of future users:

what are their activities, what roles must they fulfill and under
which operational scenarios?
The main concepts proposed by Arcadia at this level are as
follows:
– Operational Capability: capability of an organization to provide a
high level service leading to an operational objective being
reached (for example Provide weather forecasts, etc.);
– Operational Entity: entity belonging to the real world
(organization, existing system, etc.) whose role is to interact with
the system being studied or with its users (for example Crew,
Ship, etc.);
– Operational Actor: particular case of a (human) non-
decomposable operational entity (for example Pilot, etc.);
– Operational Activity: process step carried out in order to reach a
precise objective by an operational entity, which might need to
use the future system in order to do so (for example Detect a
threat, Collect meteorological data, etc.);
– Operational Interaction: exchange of information or of
unidirectional matter between operational activities (for example
meteorological data, etc.);
– Operational Process: series of activities and of interactions that
contribute toward an operational capability.
– Operational Scenario: scenario that describes the behavior of
entities and and/or operational activities in the context of an
 operational capability. It is commonly represented as a sequence
diagram, with the vertical axis representing time.
In Figure 1.3, first of all we can see the structural elements (gray
rectangles), i.e. the Entities and the Actors. An Operational Actor is a
human and non-decomposable Operational Entity. An Entity can
contain other Entities or Actors, like Entity 1, which contains the
Actors 2 and 4. Next, we can see the Activities (orange rectangles),
within the Actors or Entities: this is the allocation relation. One or
several Activities can be associated with a same Entity or Actor. This
is the case with Activity 2 and Activity 4, which are associated with
Actor 3. Between the Activities we can see the Interactions, or
operational exchanges (orange arrows). A succession of Activities
and Interactions constitutes an Operational Process. This is the case
for the Operational Process 1 in the figure, which is made up of the
sequences of Interactions 1–3. This Operational Process is
represented both by a blue square on the diagram, and also by the
bold and blue Interactions involved, as well as the source and target
Activities (here the same Activity 1).
 FIGURE 1.3 Diagram showing the main concepts behind
Operational Analysis. For a color version of the figure, see
www.iste.co.uk/roques/arcadia.zip

Other more advanced concepts are also available if needed:

Operational Role, Communication Mean, Mode and State, Exchange
Item, Class, etc. We shall discuss some of these throughout the case
study presented in Chapter 3.

1.2.3 System Analysis

System Analysis involves the identification of the Capabilities and
Functions of the system that will satisfy the operational needs
(“what the system must accomplish for the users”).
This involves carrying out an external functional analysis of the
system under study in order to identify the System Functions
required by the users (e.g. “calculate the optimal itinerary”, “detect a
threat”) from the response, under the constraint of the non-
functional properties ordered.

Note
The System is identified as a modeling element at this level. It is a
“black box” containing no other structural elements, only
allocated Functions.

The main concepts proposed by Arcadia at this level are as

follows:
– System: organized group of elements that function as a unit (black
box) and respond to the needs of the users. The System owns
Component Ports that allow it to interact with the external
Actors;
– Actor: any element that is external to the System (human or
nonhuman) that interacts with it. (for example Pilot, Test
operator, etc.);
– System Capability: capability of the System to provide a high-
level service allowing it to carry out an operational objective (for
example provide meteorological data, etc.);
– Function: behavior or service provided by the System or by an
Actor (for example detect a threat, measure altitude, etc.). A
Function owns Function Ports that allow it to communicate with
the other Functions. A Function can be split into subfunctions;
– Functional Exchange: unidirectional exchange of information or
of matter between two Functions, linking two Function Ports;
– Component Exchange: connection between the System and one of
its external Actors, allowing circulation of Functional Exchanges;
– Scenario: dynamic occurrence describing how the System and its
Actors interact in the context of a System Capability. It is
commonly represented in the form of a sequence diagram, with
the vertical axis representing time;
– Functional Chain: element of the model that enables a specific
path to be designated among all possible paths (using certain
Functions and Functional Exchanges). This is particularly useful
for assigning constraints (latency, criticality, etc.), as well as
organizing tests.
In Figure 1.4, we can first of all see the structural elements (blue
rectangles), i.e. the System and the Actors. An Actor is an entity that
is external to the System (human or not). Next we can see the
Functions (green rectangles), which are inside the System or the
Actors: this is the allocation relation. One or several Functions can be
allocated to the same structural element. This is the case for
Functions 2 and 4, which are allocated to Actor 3, as well as for
Functions 1 and 3, which are allocated to the System. The Functional
Exchanges (green arrows) are represented between the Functions,
linking a Function Port of the output of a Function (green square) to
a Function Port of the input of another Function (orange square).
One or more Functional Exchanges can be allocated to the same
Component Exchange (blue line). This is the case for Functional
Exchanges 1 and 2, which are both allocated to the Component
Exchange 1, as shown by the dotted line linking the Function Ports
to the Component Ports. A Component Exchange has to link the
System to one of its Actors, via Component Ports (white squares),
which can be uni- or bidirectional. In the case of Figure 1.4, the Ports
are bidirectional, since the Functional Exchanges are of opposite
directions.

FIGURE 1.4 Diagram showing the main concepts behind

System Analysis. For a color version of the figure, see
www.iste.co.uk/roques/arcadia.zip

Other more advanced concepts are also available if needed:

Mission, Mode and State, Exchange Item, Class, Interface, etc. We
will discuss several of these in the context of the case study in
Chapter 4.

1.2.4 Logical Architecture

The level of Logical Architecture aims to identify Logical
Components inside the System (“how the system will work to fulfill
expectations”), their relations and their content, independently of
any considerations of technology or implementation.
Next an internal functional analysis of the system must be carried
out: the subfunctions required to carry out the System Functions
chosen during the previous phase must be identified; next, a split
into Logical Components to which these internal subfunctions will
be allocated must be determined, all the while integrating the
nonfunctional constraints that have been chosen for processing at
this level.
The main concepts proposed by Arcadia at this level are as
follows:
– Logical Component: structural element within the System, with
structural Ports to interact with the other Logical Components
and the external Actors. A Logical Component can have one or
more Logical Functions. It can also be subdivided into Logical
subcomponents;
– Logical Actor: any element that is external to the System (human
or non-human) and that interacts with it (for example Pilot,
Maintenance operator, etc.).
– Logical Function: behavior or service provided by a Logical
Component or by a Logical Actor. A Logical Function has
Function Ports that allow it to communicate with the other
Logical Functions. A Logical Function can be subdivided into
Logical subfunctions;
– Functional Exchange: a unidirectional exchange of information or
matter between two Logical Functions, linking two Function
Ports;
– Component Exchange: connection between the Logical
Components and/or the Logical Actors, allowing circulation of
 the Functional Exchanges;
– Logical Scenario: dynamic occurrence describing the interactions
between Logical Components and Logical Actors in the context of
a Capability. It is commonly represented as a sequence diagram,
with the vertical axis representing the time axis;
– Functional Chain: element of the model that enables a specific
path to be designated among all possible paths (using certain
Functions and Functional Exchanges). This is particularly useful
for assigning constraints (latency, criticality, etc.), as well as
organizing tests;
In Figure 1.5, we can first of all see the structural elements (blue
rectangles), i.e. the Logical Components (contained in overarching
box that represents the System at the Logical level) and the Actors.
Next, we can see the Functions (green rectangles), within the Logical
Components or the Actors: this is the allocation relation. The
Functional Exchanges are represented between the Functions,
always linking a Function Port from a Function output to a Function
Port from the input of another Function. A Component Exchange
either links the Logical System to one of its Actors, or a Logical
Component directly to an external Actor, or two Logical
Components via Component Ports (uni- or bidirectional). In the case
of Figure 1.5, the Ports of Component Exchange 2 that link the two
Logical Components inside the System are unidirectional, since only
one Functional Exchange is allocated to it. On the other hand,
Component Exchange 1, which still links Actor 3 to the System, is
now delegated to the two Logical Components via unidirectional
Component Ports, each belonging to a different Logical Component.
This mechanism allows us to finely specify the responsibilities of
each Logical Component by attaching it to the responsibilities of the
System level.
 FIGURE 1.5 Diagram showing the main concepts behind
Logical Architecture. For a color version of the figure, see
www.iste.co.uk/roques/arcadia.zip

Other more advanced concepts are also available if needed:

Mission, Mode and State, Exchange Item, Class, Interface, etc. We
shall discuss some of these during the case study presented in
Chapter 5.

1.2.5 Physical Architecture

The objective of this level is the same as for Logical Architecture,
except that it defines the final architecture of the system, and how it
must be carried out (“how the system will be built”).
It adds the Functions required for implementation, as well as the
technical choices, and highlights two types of Physical Component:
– Behavior Physical Component: Physical Component tasked with
Physical Functions and therefore carrying out part of the
behavior of the System (for example software component, data
server, etc.);
– Node (or Implementation) Physical Component: Physical
Component that provides the material resources needed for one
 or several Behavior Components (for example processor, router,
OS, etc.).
At this level, the main concepts proposed by Arcadia are similar to
those of the Logical Architecture: Physical Function, Functional
Exchange, Physical Component, Physical Actor, etc. However, there
are some additional concepts, notably:
– Physical Port: non-oriented port that belongs to an
Implementation Component (or Node). The structural port
(Component Port), on the other hand, has to belong to a Behavior
Component;
– Physical Link: non-oriented material connection between
Implementation Components (or Nodes). The Component
Exchange remains a connection between Behavior Components.
A Physical Link allows one or several Component Exchanges to
take place (for example Ethernet cable, USB cable, etc.);
– Physical Path: organized succession of Physical Links enabling a
Component Exchange to go through several Implementation
Components (or Nodes).
In Figure 1.6, we can first of all see the Node Components (yellow
rectangles). Next, we can see the Behavior Components (blue
rectangles) deployed over each Node. Finally, we can see the
Functions (green rectangles) inside the Behavior Components: this is
the allocation relation. The Functional Exchanges are represented
between the Functions, always linking a Function Port from a
Function output to a Function Port from the input of another
Function. A Component Exchange links either a Behavior Physical
Component to an external Actor, or two Behavior Physical
Components, via Component Ports (uni or bidirectional). One or
more Functional Exchanges can be allocated to the same Component
Exchange. Next, the Component Exchanges can themselves pass
through Physical Links (red lines), linking two Physical Ports
(yellow squares) of Node Components. This is the case in Figure 1.6,
where Functional Exchange 5 is allocated to Component Exchange 4,
which passes through the Physical Link 1.
Exploring the Variety of Random
Documents with Different Content
thought I heard some one inside. But after I had pounded on the
door there was not a sound.”
Suddenly her voice broke and, giving us an appealing look, she
asked if we would not go back to the library and break open one of
the windows, so we could get within. There was no doubt she was
afraid something had happened to Warren. When she finished
speaking, there was just one response. It came from Ranville.
“I think we better do as she says, Carter,” was all he said.
At these words the woman ran from the room, returning in a
moment with two flashlights, which she gave us. She half started to
follow us from the piazza, and then, as if thinking better of her
resolution, stopped by the door. As we went down the steps to the
ground, our last sight was the housekeeper, standing in the open
door with the light from the hall streaming out into the night.
It was now dark. As we retraced our steps, the high hedges on
each side of the path caused the walk to appear like a black tunnel.
Above our heads we could catch a glimpse of the stars, and could
hear the faint rustle of the branches of the trees. For some reason no
one spoke, nor for that matter did we hurry.
Climbing the slight hill, we approached the building, which
loomed a dark mass before us. On the veranda we paused for a
second, and then the darkness was split by the sudden ray from
Carter's flashlight. We tried the door again, but it was still locked, and
there came no response to our knock. The window was six or seven
feet above our heads, and to reach it some one would have to do a
little climbing.
As I was the lightest, they proposed to lift me from the floor to the
ledge of the window. If I found it was locked, I was to break the
glass, lift the window, and climb into the room. Ranville gave me his
hand, and I reached the sill. Balancing myself on the narrow ledge, I
tried to peer into the room, but it was a dense black shadow of
gloom. Nothing could be distinguished, and though I waited a
second, the only sound to come to my ears was the wind in the
branches of the near-by trees.
Trying the window, I found it locked. Then Carter reached up to
me the second flashlight, and without turning it on I broke the glass
with the heavy end. The glass fell with a tinkling sound to the floor,
and slipping my hand through the hole, I turned the catch and lifted
the window. As I did this, I dropped the flashlight, which fell with a
thud within the room. Hesitating a second, I dropped into the library
and fumbled on the floor for the flashlight.
I found it without any trouble and, putting on the catch, played the
light hastily around the room. Just what I expected to see I cannot
say; but the brief sweep which I made over the floor and the walls
revealed nothing. The room evidently took in the entire house, and
the walls showed only long lines of books, and a gallery which ran
around the eight sides. In the center was a large desk, the surface
littered with books and papers. But of Warren there was not a trace.
Turning the light to the door, I found the spring lock was on. It
took but a second to fling the door open, and Carter and Ranville
slipped within. The same question was on both faces, and I slowly
shook my head in reply. Carter's first words were for me to find the
switch for the lights. The button was near the door, and, pressing it,
the room in an instant was a blaze of light.
The room was octagonal in shape, with a window placed high on
each of the sides. The wall space was filled with bookcases, and
there must have been many thousand volumes. A gallery at a height
of around twelve feet ran completely around the room. Even this was
filled with books. The furniture was simple. Near the door stood a
safe, and there were a number of stands in various corners. But in
the center of the room was the largest desk I had ever seen—a huge
affair made out of an old-fashioned square piano—with its surface
littered with books and papers. Near it stood a typewriter stand, with
the machine uncovered. And then, suddenly, we saw something else
—something which drove all other thoughts from our minds. Peering
from behind the desk was a foot—a foot which did not move.
We must have seen it at the same moment, for Carter's hand
gripped my arm, and for a second we stood silent. Then without a
word slowly we went across the floor, knowing what we would find.
Though we were sure what was behind the desk, yet it came as a
shock. For there, lying on his back upon the floor—with both arms
outstretched from the body—lay a man. A man to whom the dinner
waiting in the big house would never matter; and it needed but a
glance to know that death had come suddenly—and violently.
 As Ranville's eyes and mine met, they framed the same question.
It was Carter who spoke the two words:
“It's Warren.”
The scientist was a man of about fifty, and perhaps a little over
that age. The face was self-willed, and the lines around the distorted
lips were stern. Though past middle life, his hair was a dense black,
without a sign of gray, and there was not a white hair in the close-
cropped mustache. One could tell by his figure that he had been a
man of the strongest physique. He was dressed in a light summer
suit, without a coat, and upon the white shirt, just over the heart, was
a crimson stain.
Carter dropped on his knees and made a hasty examination. In a
second he turned and pointed with his finger at the dark stain upon
the white shirt. Then as he straightened up we saw something else—
something we had overlooked. It was a sheet of paper. Evidently it
had fallen off the body, though perhaps it had been placed by its
side. A piece of bond paper with but five letters—large letters,
evidently written with a hurried hand, the beginning of an
incompleted word:

—ANANI—

There had been little conversation, for we were too upset by what
we had found. But the piece of paper puzzled us. That Warren had
been stabbed there was no doubt; but what the paper meant we
could not tell. The letters seemed to mean nothing, and we were not
sure that they had anything to do with the crime. For a moment we
puzzled over it, and then my eyes wandered again to the still figure
upon the floor. As I glanced at it, I gave a sudden start and dropped
to my knees for a closer look. And then—then, after one glance, I
gave a startled cry.
For there upon the forehead of the murdered man were two faint
lines—lines now swollen and red. Not very long lines, nor for that
matter very noticeable, but lines which I could not understand. There
upon the forehead of the famous scientist were two faint lines cut
into the skin. A cross—the lines of which had just been made. Cut
faintly, I judged, with a knife. A cross—the lines now red and swollen,
and a crooked cross at that.
 Chapter III.
The Broken Bookcase
At my cry of astonishment Carter and Ranville had turned in
surprise. I simply pointed to the forehead of the murdered man, and
they bent forward for a closer look. I saw a startled expression
sweep over the Englishman's face, and he slowly shook his head. It
was Carter who broke the silence, speaking to no one in particular.
“Do you think that was made by the murderer?”
“There is not the slightest doubt of it,” was Ranville's quick retort.
“That man has not been dead over two hours, and the cut itself is not
any older.”
I cast a hurried glance at the grewsome lines of the red cross and
gave a little shiver as I asked:
“But what under heavens can it mean; why should there be a
mutilation of that kind?”
Carter simply shook his head, and it was the Scotland Yard
Inspector who replied:
“We do not know, of course. I have seen a good many murdered
people in my time, but as a rule the murderer had never marked his
victim. Once in a while you will run into a murder which was
committed by a woman—committed in a fit of frenzy. Sometimes in
such a case they mark up their victims. But of course we know
nothing of this crime. What the motive was we do not know. How he
was killed is rather easy to understand—a long thin knife or dagger.”
The body lay upon the floor near the desk, but about two feet
behind it. The position was such that any one coming into the
building by the front door would have been unable to see it. Save for
the crimson spot upon the shirt and the faint cross upon the
forehead, there were no signs of violence.
 But the position in which the body lay was rather odd. The man
lay flat upon his back, the staring eyes fixed upon the ceiling. But
both arms were outstretched as far as they could reach. It was this
that puzzled me. I knew it was impossible that the man could have
fallen in the position in which he was. Some one must have arranged
the body after the crime—but why?
Behind the desk was the chair in which the scientist must have
sat while at work. Near it, on the left, was another chair, back of the
typewriting stand. And on the other side of the desk, very close to it,
was a third chair. The surface of the desk was covered with papers
and pamphlets. A small heap of manuscript was piled in an orderly
manner in the very center. But of any weapon there was not a sign.
I was just starting to comment upon this when I observed that
Ranville was carefully studying the position of the chairs. In a
moment he went around the desk, studying the place where a chair
stood. Then he turned to us.
“I have an idea I can reconstruct the murder. See the three
chairs? There is no trouble about the two on your side of the desk;
one was where Warren sat when at work, the other was for his
secretary. But this chair on my side of the desk tells us a good deal.”
I cast an inquiring gaze at the chair, a tall antique piece of
furniture, while Ranville continued:
“In a room as large as this you will not as a rule find a chair pulled
up to a desk, across from which a man is working. But if some one
comes in, the natural thing is to bring a chair near the desk, to be as
close to the man you are talking to as you can. Now there are other
chairs in the room across the desk from where Warren sat; but they
are all rather far away. All but that one, and I am pretty sure the
murderer sat there.”
When he mentioned it, I noticed for the first time that there were a
number of other chairs across from us. Some were near the wall,
and one in front of the safe; but the chair he was speaking of stood
but two feet from the desk. Seeing we did not speak, he went on:
“What happened, I think, was this. The murderer sat in this chair
talking to Warren. I have an idea it was some one he knew. Though I
do not know his habits, yet I doubt if Warren would spend much time
while at work with any one he did not know. The papers said he was
rushing his book. Maybe there were some words passed, maybe not.
But then, suddenly, Warren was killed.”
“Why suddenly?” came Carter's dry question.
“Warren seems to have been a man of strong physical
development. There is no evidence of any struggle. In a fight I judge
he could have held his own with any one. So if there was no
struggle, it follows he was killed suddenly. I judge whoever sat in that
chair must have risen—perhaps said he was going—strolled to
Warren's side and suddenly stabbed him.”
I again turned my eyes to the figure upon the floor, and again the
outstretched arms puzzled me.
“But he never fell to the floor in that position,” I said.
“He never did,” was the reply. “The body was arranged in that
position, and the cut on his forehead was made after he was dead.”
“But why?” asked Carter.
“God knows!” was the retort. “But then, Carter, this is not our
show anyway.”
Carter gave a sudden start, saying slowly: “You are right. I will
call up our chief of police. He will get a mighty big shock, for there
has not been a murder in this town in years. And then”—he paused
—“then I'd better call the housekeeper and break the news to her.”
There was a telephone in the building near the door. After several
attempts Carter got the housekeeper and told her that she had better
come to the summer house. Then he held a short conversation with
the police station, after which he returned to our side.
“While we are waiting for the police, we had better look this place
over,” he said.
As I have mentioned, the building was an odd one with eight
sides and only one story. There was a window at each side, placed
rather high, and the space between the windows was filled with
bookcases. All these cases had glass doors, some of which were
open, while others we found locked. The books in the cases were
mostly upon science and anthropology—the library of a professional
scientist. It was not until we reached the further side of the room that
we found anything out of the way. But there we found one of the
bookcases with the glass in the locked door smashed into hundreds
of pieces—pieces which lay upon the rug at our feet.
 Behind the broken glass were seven book shelves with books
packed tightly together. They were mostly bound in a uniform red
morocco, small volumes, not very thick nor very tall. Only in the third
shelf was there a gap, and there several books seemed to be
missing.
And the books themselves turned out to be a rather curious
collection, yet when one remembered Warren's profession, perhaps
they were not so out of place as I first thought. The word “erotic”
describes them best, though several went beyond that. Why a
scientist should wish to have upon his shelf “The Perfumed Garden,”
“The Ananga Ranga,” “Aretino” and others one could understand.
But there were certain other things in the case which seemed out of
place.
Side by side with the classics of the underworld of literature stood
the witty and immoral romances of the eighteenth century of France.
But there were a few modern books, decidedly pornographic in type,
which flanked the more classical ones. An odd collection at the best,
worth a good deal of money, it is true. But the oddest thing to explain
was why some one had broken the glass to get at the contents of the
case.
The Englishman gave a low whistle, and I saw his eyebrows
raise. Reaching in his hand, it came forth with a volume. It stood on
the shelf which had the empty space, the one where, if the broken
gap told the truth, several books were missing. He turned the leaves
slowly, shrugged his shoulders at several of the engravings, and
then without a word handed the book to me. It was the first volume of
De Sade's “Justine,” the first edition with the illustrations. I
remembered once hearing Bartley say that it was the worst book
ever written and very difficult to secure. In turn, without speaking, I
handed the thin volume to Carter just as Ranville expressed what
was in his mind.
“That's not only a pretty rare book, but it is also a rather rotten
one. It looks very much as if some one smashed the glass in this
case to get at the books. What they took I cannot tell, though it might
be the other volumes of that ‘Justine.’ I cannot understand why any
murderer should want the books. Besides, it's the French edition,
and not every chap reads French, you know.”
 We agreed to this, and placed the book back in the case. Then
climbing a narrow winding stairs, we went up to the gallery. It ran
around the entire length of the room—a narrow gallery, built
evidently to give more space for books. The walls were lined with
books, thousands of them, of every kind. But there were no doors or
glass before the cases in the gallery.
Nothing had been disturbed so far as we could see. I glanced
over the rail to the floor below, giving a shudder as my eyes fell upon
the still figure by the desk—the figure with the outstretched arms.
Leaving the gallery, we tried the rear door, finding as we
expected that it was locked. As both doors had a spring lock, it would
have been only necessary for the murderer to close them when he
went out. But why the windows were down, and also locked, puzzled
us. It had been a warm day, and it hardly seemed possible that
Warren had worked in a room without any fresh air. We were
commenting on this when there came a voice from the front door,
and two men stepped into the room.
One was a very short man with a vivid red face, and I could tell
by his blue uniform that it was the chief of police. He was very warm,
as if he had been hurrying, and there was a questioning look in the
glance he gave us. He had a rather kindly face, though it was not an
over-intelligent one, and I decided that he did not fancy the task
before him. The young man with him he introduced as the coroner, a
young man named Hasty.
The chief held a short conversation with Carter and then went
over to the desk. He came to a sudden halt by the body, and I saw
the look of dismay which swept over his face. Even the doctor
seemed shocked, but went about his examination at once. When he
had finished, he rose to answer the eager questions of the chief.
He informed us the man had been dead several hours, and that
he had been stabbed. The blow had evidently reached the heart, and
the scientist must have died at once. The faint cross on the forehead
he could not explain, but he agreed that it had been made after
death.
“But,” came the heavy voice of the chief, “why should any one
wish to kill Warren? There are very few people in the town that know
him. Though this is his birthplace, he has been away so long that he
has hardly any friends here. He never cared to bother much with
people.”
He paused to throw a curious look around the room.
“If he was stabbed, where is the weapon?”
We assured him that we had seen no signs of a weapon, though
we had looked the building over. Carter said he agreed with the chief
regarding Warren's acquaintance in the town. There was no doubt
he was their most distinguished citizen, but he had been away so
many years that few knew him. But why he had been murdered, or
by whom, there was not the slightest kind of a clew.
The police chief listened, his face growing very long as Carter
went on. Like most police chiefs in small places, his work was the
usual small town routine. Confronted with a murder, and one as
mysterious as now before him, he did not know what to do. And as
he gave a glance at the body on the floor, I knew that he was much
perplexed.
As Carter and the chief started a low conversation, Ranville and I
went to the desk. No one had looked at the papers on its surface,
and as we started to glance through them, we found just about what
we had expected. The greater part of them were notes, and as I read
a sentence or two, I could see that they dealt with Warren's two
years' stay in the heart of China. Many of them had crude drawings
of bones and fossils. But the handwriting was rather bad, and I did
not bother to read more than a few lines.
There were a number of books upon the desk, but they were
mostly scientific works of reference. One red-covered volume turned
out to be a popular mystery story, and beside it stood one of the
adventure story magazines. A number of typewritten sheets,
evidently corrected work of his secretary were near the edge of the
desk, the pages filled with corrections in red ink. But there was
nothing of importance, only the natural data of a scientist who was
writing an account of his last expedition.
Just as I was about to turn away from the desk, my eyes fell upon
a piece of paper which was peering out from under the typewritten
manuscript. I pulled it forth to see what it might be. It was part of a
typewritten letter dated the day before, but with no address or
signature. There could not have been a signature, for the lower half
of the letter was missing. The sheet was torn across as if some one
had wished to destroy the signature. It read:

“Tuesday.
“Mr. Henry Warren,
“My dear Professor,
“I will call upon you to-morrow around five o'clock. I feel sure you
can spare me a few moments. If I can only make you see how great
a thing you can do for humanity, I am sure you will take my
viewpoint. The consequences of the step you are taking are so
momentous that unless—”

And there the letter ended, for the rest of the sheet had been torn
off.
It was a curious sort of a letter, and seemed to contain a warning
of some sort. It was written upon a typewriter whose ribbon was far
from clean. Not only did it contain a warning, but it seemed to me
there was a threat in the words. But more important than anything
else was the statement that the writer would call upon Warren. As it
had been written the day before, Warren must have seen the person
only an hour before he died.
Without a word I handed the letter to Ranville and watched his
face as he read. When he came to the end, I saw his eyebrows raise
a little, and he turned to me.
“This looks important, Pelt. Any signs of the missing portion of the
sheet?”
I shook my head, and we both turned to the desk. We went
through every paper, lifting them from each other, and even turning
the pages of the books. But we found nothing. Then we turned our
attention to the wastebasket, turning the contents upon a small rug.
But the basket contained only the discarded notes which had been
thrown aside and a few matches. The missing half of the letter we
did not find.
As we paused, I noticed that the chief and Carter were before the
bookcase—the bookcase with the broken glass. Ranville placed the
letter in his pocket and said: “What do you make of it?”
 I told him what I thought, that it contained both a warning and a
threat, and then said that it looked as if the missing part had been
taken in order to destroy the signature.
“True enough,” came the drawling answer. “But why did they not
take the entire letter? Why destroy half of it and leave the other? If
the whole note had been taken we would never have known
anything about it. To take but half seems a very illogical thing to do.”
Hearing our voices, Carter and the chief came to the desk and
asked what we had found. Ranville handed him the letter, and after
they had both looked at it the chief held it a long while in his hand.
His face was a study, and he slowly shook his head. He might have
spoken if Carter had not said:
“The chief agrees with me that the murder of Professor Warren is
going to make a great deal of comment. He will have the inquest to-
morrow, and hopes before then to have something to go on. As it
stands now all we know is that Warren was murdered, but nothing
else. The—”
There came a commotion at the door, and we turned, only to see
the housekeeper rush into the building. Her face was red as if she
had been running, but why she had taken so long to come to the
library after Carter called up I could not tell. For a second she leaned
against the door as if out of breath, and then gave a quick glance
around the room. In her eyes was terror, and the glance at length
rested upon Carter. With one step in his direction, she gasped in a
trembling voice so low that we could barely hear her:
“Mr. Warren—is—is he dead?”
Carter nodded, and again the woman's eyes swept the room.
This time they went slowly as if seeking for something, and as if
afraid of what she might find. Suddenly she stiffened into attention as
her glance fell upon the foot of the dead man, which could be seen
around the desk. Then slowly, a step at a time, she crossed the floor
to a place beyond the desk. There she stood, silently looking down
at the still figure of her employer. The red had faded from her
cheeks, and her face was a dull white. Slowly she turned, her eyes
asking the question before her lips spoke:
“Was he murdered?” came the quivering voice.
“Yes,” some one said.
 For a moment she did not speak. Again her eyes came back to
the silent figure. For an instant as her lips moved I thought she would
speak, but she gave a shudder and shut them tightly. But the flush
had come back to her face, and when she turned toward us, I could
see the veins in her forehead throb. And then suddenly, in a shrill
voice which rang through the room, she shrieked: “I knew it, I knew
it. It's that secretary. I knew that girl would—”
But the sentence was not completed. As the shrill voice rose
higher and higher, her hands began to beat the air; the voice died
away in her throat as if suddenly cut off. Then with a little gasp she
staggered a second and fell fainting to the floor.
 Chapter IV.
We Discuss the Crime
So unexpected was the woman's action that for a second none of
us stirred. It was the doctor who reached her first. The eyes opened
with a little flutter, and the color came flooding back into her cheeks.
As he placed her in a chair, her hands went out in a confused,
questioning gesture, as if seeking aid. Then when she realized what
had happened, she cast one horrified look in the direction of the
body.
When she was feeling more composed, the chief tried to question
her. But she refused to say a word. Before she fainted, in a voice
which rang with conviction, she had practically accused the secretary
of the murder. Now in a listless tone she refused to say a word,
shutting her thin lips in a determined manner. At last, seeing that she
did not care to speak, and in fact would not, the chief suggested that
the coroner assist her back to the house.
When they had left, he turned with an astonished air to Carter.
“What in the devil did she mean by that crack about the
secretary?”
“I don't know,” was his reply. “She seemed to be a bit angry. Who
is the secretary anyway?”
“Why it's the former stenographer of Judge Williams. She is as
good looking a girl as you will find in many a day. But that
housekeeper is crazy if she thinks that girl killed Warren.”
“Well,” came the drawling voice of Ranville, “I know nothing about
the girl you speak of; but if I were you, I would look her up.”
A few moments later, concluding that we could do no good if we
remained, we left. It was growing late, and the police had much work
to do. Besides we were beginning to feel the need of the dinner we
had not eaten. We told the chief all we knew, showed him the broken
glass in the bookcase, and mentioned what the housekeeper had
said regarding the visit of a Chinaman. Last of all we pointed out the
faint cross on Warren's forehead. This seemed to impress him more
than anything else, and I saw his eyes grow big. Then with Carter's
remark, that we would aid him in any manner he wished, we said
“good night” and went out.
The stars were bright above our heads, but it was dark at that.
The path between the hedges was a dense black line, and the trees
loomed in a somber manner above us. Reaching the lawn before the
house, we saw that the building was a blaze of lights, though we
glimpsed no one. We did not turn to the house, but instead passed
through the iron gateway and out to the road.
No one spoke, and I judged that none of us felt like speaking. As
we went along, I thought of the famous scientist, who only a few
short weeks before had been hailed in every paper of the world.
There had been many wild guesses made as to what he found, more
so after he had said that the question of man's origin was forever
settled. What he had found no one knew, and he refused to say,
simply stating that it would all come out in his book. And then the
whole controversy burst into flame.
This was caused by the theological argument which was raging
over evolution. The controversy had increased after Warren's
statement. Back and forth flew the arguments. The scientists
contented themselves by saying that every intelligent person
believed in evolution, and that if Warren said he had found the final
proof that settled it. His reputation and word was enough for the men
of science. On the other hand, theologians and demagogues who
knew nothing about science cried long and loud that Warren could
not have found any proofs of evolution, for, as they said, evolution
was not a fact.
In all this controversy—one which filled many pages of the
papers—Warren bore no part. As soon as he arrived in America, he
had gone directly to his home and made the announcement that he
would have his manuscript ready as soon as possible. Only one
statement he gave the papers—it was to repeat what he had said
before: that he had found the final proofs. The proofs which settled
for all time the question of man's origin. After that he was silent. And
now he was murdered, and I pictured the papers when they heard of
his death.
And then I began to wonder why he should have been killed. A
man of his decided personality must, of course, have made enemies.
I puzzled over the man from China, who the housekeeper said had
come to the house. I played with this thought for a while, only to
decide that perhaps it was better to stop wondering about the case
until I had more facts to puzzle over. And by the time we came in
sight of Carter's home the only thing I was thinking about was
something to eat.
The tall grandfather's clock was striking eleven as we entered the
living room. With the remark that we must be hungry, Carter went out
into the kitchen saying he would see if the cook had left anything in
the ice box. Ranville and myself dropped into the nearest chairs. I
was too tired to talk, and the experiences of the last few hours had
not been pleasant. But to look at the Scotland Yard Inspector one
would never have guessed that anything had taken place. The fine
face of the Englishman was as peaceful and contented as if he had
just returned from a wedding—instead of a murder. He lay back in
his chair, his eyes half closed, watching the curling smoke of his
cigarette.
Carter's voice hailed us from the kitchen, and we rose and joined
him. Upon the white enameled table was a cold chicken, three
bottles of ale, and some rye bread. We pulled our chairs to the table
and set to work. When the chicken had become but a memory,
Carter rummaged in the ice box and found a pie—a pie of which we
did not leave a crumb.
The lunch over, we went out on the large veranda; the night was
cool, with a slight breeze, and down at the edge of the lawn we could
hear the water lapping on the shore. As Carter handed me a cigar, I
happened to think of Trouble, locked in the garage, and went down
to rescue him. He greeted me with a loud bark, but at my command
followed to the piazza and dropped by the side of my chair. For a
while nothing was said, and in the darkness I watched the glowing
tips of my friends' cigars. It was Carter who broke the silence, saying
to no one in particular:
“Well—what do you think about the murder?”
 Ranville's drawling voice came floating from his chair, and his
tone was serious:
“It looks to me, Carter, as if we had stumbled upon what will
prove one of the most perplexing murder mysteries we have ever
seen. There are some very curious things about this affair; and it's
my idea it's going to prove rather difficult to solve.”
“It will cause a sensation all right,” was the reply. “You know for
weeks Warren's name has been on the front pages of the papers.
First there came the accounts of his trip to China. When he did not
return at the time expected, the papers began to say his expedition
was lost. Then the outlaw war broke in China, and it was thought he
was killed; and when he suddenly made his appearance, he certainly
got a lot of publicity.”
As he paused, I added my bit. I reminded them that his statement
that he had settled the question of evolution had made more
comment than anything else.
“That's right,” replied the Englishman. “Even in London the old
Times gave a good many columns to that feature. But as he refused
to say what it was he had found, the whole affair led to some little
controversy.”
“You have had a good deal of experience in murder cases in your
Scotland Yard work,” I said to the Inspector. “What do you think was
back of Warren's death?”
Ranville was silent a while, replying at last:
“That is the question. It is pretty hard to say from what we found
to-night, just what could be the motive. Men are murdered as a rule
for three reasons—robbery, revenge, or, say, in a sudden passion.
Now it does not look like robbery, for we saw no signs of anything
being taken. That is, unless we figure the murderer broke the glass
of the bookcase and took a book. But that seems hardly reasonable.”
“Still some one did take a book or two from that case,” was my
retort.
“Perhaps. Of course Warren might have broken the glass himself
by accident. Then again, though I do not know much about books, I
do know a bit about that kind of literature. Once in a while we clean
up some book dealers who put it out in London. And I know this.
None of that stuff sells at a very high figure. It's rare, of course,
mostly because it's sold under cover. But a few pounds would buy
anything in that case. It does not seem reasonable to start out by
assuming he was murdered for a book of that class.”
“Well, let's put that out of the question and say revenge,”
suggested Carter.
“That would look more reasonable,” Ranville commented. “A man
of Warren's type would, of course, have made enemies. And the two
odd things about the murder—the position in which we found the
body and the cross on the forehead—seem to suggest revenge. You
cannot tell what he might have done while he was in China. He may
have made enemies there.”
“That suggests the Chinaman who the housekeeper says came
to the house about six,” was my remark.
“Maybe and maybe not,” was Ranville's quick retort. “I admit that
six o'clock is pretty near the time Warren was killed. Also, why a
Chinaman should wish to see him is something which must be
looked into. But I have had a good deal of experience with criminal
Chinese in our Limehouse section of London. They are capable of
the most devilish torture, the weirdest kinds of murder. But I fail to
remember a single case where they ever marked their victim after
death. And no Chinaman, it seems to me, would ever mark his victim
with a cross. Of course, once in a while you run into one who goes
wild, and there is no telling what he might do. But as a rule, though
they will in seeking revenge impose the most cruel tortures on some
of their victims, they do not as a rule mark them after death.”
“Disfigurement after a killing is often the work of a frenzied
woman,” was Carter's shrewd remark.
“That's true, Carter. Women, far more than men, are apt not to be
satisfied with murder alone. When a woman in a sudden passion kills
a man, she often, while the rage is on her, goes further.”
“And that makes one think of what the housekeeper said about
the secretary,” was my comment.
There was a moment's silence, broken by Carter's saying:
“I wonder what the housekeeper meant by that remark. She
certainly shut up like a clam when we tried to question her. There is
something back of it—at least back of the housekeeper's attitude.”
 “Well,” came Ranville's voice, “there is one thing sure; I think I am
right when I say that whoever killed Warren was some one who knew
him. He sat in that chair, the one across from the desk, and I think
perhaps I am right when I add that he might have gone to Warren's
side to say good-by and then plunged the knife into him. But why he
paused to arrange the body on the floor and to make the cross on
his forehead I cannot say, but—”
Just what he might have added I do not know. We were
interrupted by the dog suddenly rising to his feet and starting to
growl. Deep, heavy growls at some object we could not see. Then
came the sound of footsteps on the walk, and a deep voice came
from the lawn:
“Hearing voices, Mr. Carter, I could not resist stopping.”
As the man came up the steps, I pushed the dog behind my
chair, telling him to be quiet. Carter rose and turned on the porch
lamp. For a moment the light, after the dense darkness, blinded me.
I wondered who could be coming to see Carter at this late hour. It
was a very tall and an extremely thin man who accepted the chair
which Carter pulled out. A man with a deep lined face and nervous
shifting eyes. As he came over to the chair, I saw that he was
wearing a clergyman's collar, though he did not look as calm as most
of the clergymen I have seen.
He proved to be Carter's next-door neighbor, and he told us he
was on his way home when he heard the sound of our voices. As he
talked, I could see that he was of a nervous, restless disposition, for
his hands were never still, and he moved his feet in an uneasy
manner. His voice was rather harsh, though the English he used was
perfect.
After the introductions had been acknowledged Carter said:
“Woods is my next-door neighbor. He's been in England,
Ranville.”
The clergyman admitted that he had been in England many
times. He changed the conversation at once by remarking:
“I have just come from down town, and they are all excited over
the murder of Mr. Warren. I did not know Warren very well, but it
seems almost incredible a man of his position should have met with
such a sudden death. Have they any idea who the guilty person is?”
 We all shook our heads, and then Carter went into a brief
description of the finding of the body. The eyes of the minister grew
larger as he went on, and I saw a horrified look sweep across his
face. As I did not know many clergymen, I studied the man before
me with interest. It was easy to see that he had a good education,
and I wondered why he had buried himself in such a small country
town. Long before Carter had finished I had decided that the minister
was as nervous a man as I have ever met. His hands were never
still, and his eyes were as uneasy as his hands.
He said nothing until Carter mentioned the secretary, and then
half rose as he burst forth:
“Why, of all things,” came the rough high-pitched voice, “I know
Mr. Warren's secretary very well. She comes to my church. You must
know her also—Miss Harlan?” and he turned to Carter.
Carter shook his head; then said he knew her by sight and that
was all. He added she was a very fine-looking girl.
“That's very true,” the minister eagerly replied. “She is not only a
very fine-looking girl, but a very fine girl in all ways. It's absurd to
think she knows anything about Mr. Warren's death. I saw her myself
this afternoon.”
The conversation for some reason lagged after this, and after a
while the minister gave a glance at his watch, and then with a
sudden exclamation rose saying it was late. We said “good night,”
and he went down the steps and was lost to sight. After he was out
of hearing Ranville asked:
“How long have you taken up with clergymen, Carter?”
His friend laughed. “Oh, I do not know him so very well. He has
lived here for some time. It seems that about fifty years ago his
grandfather—for some unknown reason—built the church next door.
Woods sort of fell into it. He has a good deal of money they say, but
very few people ever go to his church. In fact he supports it himself.
You see he is about as high church as you can find—all sorts of
rituals and that kind of thing. They don't go very well in a place like
this. Then again, he is always attacking something.”
“Attacking something; what do you mean?” was Ranville's
puzzled question.
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!

ebooknice.com