Experience Seamless Full Ebook Downloads for Every Genre at ebookmeta.

com

Implementing MLOps in the Enterprise: A

Production-First Approach 1st Edition Yaron Haviv

https://ebookmeta.com/product/implementing-mlops-in-the-
enterprise-a-production-first-approach-1st-edition-yaron-
haviv/

OR CLICK BUTTON

DOWNLOAD NOW

Explore and download more ebook at https://ebookmeta.com

 Recommended digital products (PDF, EPUB, MOBI) that
you can download immediately if you are interested.

Implementing MLOps (Early Release) Noah Gift

https://ebookmeta.com/product/implementing-mlops-early-release-noah-
gift/

ebookmeta.com

Mastering Azure API Management: A Practical Approach to

Designing and Implementing an API-Centric Enterprise
Architecture 1st Edition Sven Malvik
https://ebookmeta.com/product/mastering-azure-api-management-a-
practical-approach-to-designing-and-implementing-an-api-centric-
enterprise-architecture-1st-edition-sven-malvik/
ebookmeta.com

Managing AI in the Enterprise: Succeeding with AI Projects

and MLOps to Build Sustainable AI Organizations Klaus
Haller
https://ebookmeta.com/product/managing-ai-in-the-enterprise-
succeeding-with-ai-projects-and-mlops-to-build-sustainable-ai-
organizations-klaus-haller/
ebookmeta.com

Photos from Beijing 1989 1st Edition Wesley Gibbs

https://ebookmeta.com/product/photos-from-beijing-1989-1st-edition-
wesley-gibbs/

ebookmeta.com
The Master and Margarita Mikhail Bulgakov

https://ebookmeta.com/product/the-master-and-margarita-mikhail-
bulgakov/

ebookmeta.com

McGraw Hill Education SAT Subject Test Math Level 1 4th Ed

4th Edition John J. Diehl

https://ebookmeta.com/product/mcgraw-hill-education-sat-subject-test-
math-level-1-4th-ed-4th-edition-john-j-diehl/

ebookmeta.com

Plunkett s Computers Hardware Software Industry Almanac

2022 Computers Hardware Software Industry Market Research
Statistics Trends and Leading Companies 2022 ed. Edition
Jack W. Plunkett
https://ebookmeta.com/product/plunkett-s-computers-hardware-software-
industry-almanac-2022-computers-hardware-software-industry-market-
research-statistics-trends-and-leading-companies-2022-ed-edition-jack-
w-plunkett/
ebookmeta.com

React Up Running Building Web Applications Second edition

Stoyan Stefanov

https://ebookmeta.com/product/react-up-running-building-web-
applications-second-edition-stoyan-stefanov/

ebookmeta.com

The Highest Good in Aristotle and Kant First Edition.

Edition Joachim Aufderheide (Editor)

https://ebookmeta.com/product/the-highest-good-in-aristotle-and-kant-
first-edition-edition-joachim-aufderheide-editor/

ebookmeta.com
C C Course for Beginners 5e Callie Bond

https://ebookmeta.com/product/c-c-course-for-beginners-5e-callie-bond/

ebookmeta.com
Implementing
MLOps in the
Enterprise
A Production-First
Approach

Yaron Haviv
& Noah Gift
Implementing MLOps in the Enterprise
This practical guide will help your organization bring data
science to life for different real-world MLOps scenarios. Senior “The authors excel in
data scientists, MLOps engineers, and ML engineers will learn
presenting complex
how to tackle challenges that prevent many businesses from
concepts in a clear
moving ML models to production and scaling their AI initiatives.
and relatable manner.
Authors Yaron Haviv and Noah Gift take a production-first Their emphasis on the
approach. Rather than beginning with the ML model, you’ll learn importance of ROI, risk
how to design a continuous operational pipeline—while making
management, and
sure that various components and practices can map into it.
strategic technology
By automating as many components as possible, and making
the process fast and repeatable, your pipeline can scale to match
adoption provides
your organization’s needs. practical guidance for
organizations looking to
This book will show you how to generate rapid business value
leverage ML effectively.”
while answering dynamic MLOps requirements. You’ll learn the
—Dhanasekar Sundararaman
foundations of the MLOps process, including its technological
Researcher, Microsoft
and business value, and discover how to:
• Build and structure effective MLOps pipelines
Yaron Haviv is a serial entrepreneur
• Efficiently scale MLOps across your organization with deep technological experience in
• Explore common MLOps use cases data, cloud, AI, and networking. Yaron
is the cofounder and CTO of Iguazio,
• Build MLOps pipelines for hybrid deployments, real-time which was acquired by McKinsey and
predictions, and composite AI Company in 2023. He is an author,
• Prepare for and adapt to the future of MLOps keynote speaker, and contributor to
various AI associations, publications,
• Use pretrained models like Hugging Face and OpenAI to and communities.
complement your MLOps strategy
Noah Gift is the founder of Pragmatic
AI Labs. He lectures in the data science
programs at universities including
Northwestern, Duke, UC Berkeley,
UNC Charlotte, and the University
of Tennessee.

DATA Twitter: @oreillymedia

linkedin.com/company/oreilly-media
youtube.com/oreillymedia
US $79.99 CAN $99.99
ISBN: 978-1-098-13658-1
 Implementing MLOps
in the Enterprise
A Production-First Approach

Yaron Haviv and Noah Gift

Beijing Boston Farnham Sebastopol Tokyo

Implementing MLOps in the Enterprise
by Yaron Haviv and Noah Gift
Copyright © 2024 Yaron Haviv and Noah Gift. All rights reserved.
Printed in the United States of America.
Published by O’Reilly Media, Inc., 1005 Gravenstein Highway North, Sebastopol, CA 95472.
O’Reilly books may be purchased for educational, business, or sales promotional use. Online editions are
also available for most titles (http://oreilly.com). For more information, contact our corporate/institutional
sales department: 800-998-9938 or [email protected].

Acquisition Editor: Nicole Butterfield Indexer: WordCo Indexing Services, Inc.

Development Editor: Corbin Collins Interior Designer: David Futato
Production Editor: Beth Kelly Cover Designer: Karen Montgomery
Copyeditor: Piper Editorial Consulting, LLC Illustrator: Kate Dullea
Proofreader: Heather Walley

December 2023: First Edition

Revision History for the First Edition

2023-11-30: First Release

See http://oreilly.com/catalog/errata.csp?isbn=9781098136581 for release details.

The O’Reilly logo is a registered trademark of O’Reilly Media, Inc. Implementing MLOps in the Enterprise,
the cover image, and related trade dress are trademarks of O’Reilly Media, Inc.
The views expressed in this work are those of the authors and do not represent the publisher’s views.
While the publisher and the authors have used good faith efforts to ensure that the information and
instructions contained in this work are accurate, the publisher and the authors disclaim all responsibility
for errors or omissions, including without limitation responsibility for damages resulting from the use
of or reliance on this work. Use of the information and instructions contained in this work is at your
own risk. If any code samples or other technology this work contains or describes is subject to open
source licenses or the intellectual property rights of others, it is your responsibility to ensure that your use
thereof complies with such licenses and/or rights.

978-1-098-13658-1
[LSI]
 Table of Contents

Preface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix

1. MLOps: What Is It and Why Do We Need It?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

What Is MLOps? 2
MLOps in the Enterprise 2
Understanding ROI in Enterprise Solutions 3
Understanding Risk and Uncertainty in the Enterprise 5
MLOps Versus DevOps 6
What Isn’t MLOps? 8
Mainstream Definitions of MLOps 8
What Is ML Engineering? 9
MLOps and Business Incentives 10
MLOps in the Cloud 10
Key Cloud Development Environments 13
The Key Players in Cloud Computing 17
MLOps On-Premises 21
MLOps in Hybrid Environments 22
Enterprise MLOps Strategy 22
Conclusion 23
Critical Thinking Discussion Questions 24
Exercises 24

2. The Stages of MLOps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

Getting Started 25
Choose Your Algorithm 26
Design Your Pipelines 28

iii
 Data Collection and Preparation 29
Data Storage and Ingestion 30
Data Exploration and Preparation 33
Data Labeling 35
Feature Stores 36
Model Development and Training 38
Writing and Maintaining Production ML Code 39
Tracking and Comparing Experiment Results 42
Distributed Training and Hyperparameter Optimization 44
Building and Testing Models for Production 45
Deployment (and Online ML Services) 48
From Model Endpoints to Application Pipelines 49
Online Data Preparation 51
Continuous Model and Data Monitoring 52
Monitoring Data and Concept Drift 54
Monitoring Model Performance and Accuracy 57
The Strategy of Pretrained Models 58
Building an End-to-End Hugging Face Application 59
Flow Automation (CI/CD for ML) 61
Conclusion 64
Critical Thinking Discussion Questions 65
Exercises 65

3. Getting Started with Your First MLOps Project. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

Identifying the Business Use Case and Goals 67
Finding the AI Use Case 69
Defining Goals and Evaluating the ROI 72
How to Build a Successful ML Project 74
Approving and Prototyping the Project 75
Scaling and Productizing Projects 76
Project Structure and Lifecycle 78
ML Project Example from A to Z 80
Exploratory Data Analysis 80
Data and Model Pipeline Development 82
Application Pipeline Development 84
Scaling and Productizing the Project 86
CI/CD and Continuous Operations 88
Conclusion 90
Critical Thinking Discussion Questions 90
Exercises 90

iv | Table of Contents
4. Working with Data and Feature Stores. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Data Versioning and Lineage 92
How It Works 93
Common ML Data Versioning Tools 95
Data Preparation and Analysis at Scale 105
Structured and Unstructured Data Transformations 106
Distributed Data Processing Architectures 107
Interactive Data Processing 108
Batch Data Processing 110
Stream Processing 114
Stream Processing Frameworks 115
Feature Stores 117
Feature Store Architecture and Usage 118
Ingestion and Transformation Service 119
Feature Storage 120
Feature Retrieval (for Training and Serving) 121
Feature Stores Solutions and Usage Example 122
Using Feast Feature Store 123
Using MLRun Feature Store 126
Conclusion 130
Critical Thinking Discussion Questions 131
Exercises 131

5. Developing Models for Production. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133

AutoML 133
Running, Tracking, and Comparing ML Jobs 136
Experiment Tracking 137
Saving Essential Metadata with the Model Artifacts 139
Comparing ML Jobs: An Example with MLflow 140
Hyperparameter Tuning 142
Auto-Logging 144
MLOps Automation: AutoMLOps 146
Example: Running and Tracking ML Jobs Using Azure Databricks 147
Handling Training at Scale 151
Building and Running Multi-Stage Workflows 152
Managing Computation Resources Efficiently 153
Conclusion 158
Critical Thinking Discussion Questions 158
Exercises 159

Table of Contents | v
6. Deployment of Models and AI Applications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
Model Registry and Management 161
Solution Examples 163
SageMaker Example 163
MLflow Example 165
MLRun Example 166
Model Serving 168
Amazon SageMaker 170
Seldon Core 171
MLRun Serving 173
Advanced Serving and Application Pipelines 176
Implementing Scalable Application Pipelines 177
Model Routing and Ensembles 187
Model Optimization and ONNX 189
Data and Model Monitoring 190
Integrated Model Monitoring Solutions 192
Standalone Model Monitoring Solutions 197
Model Retraining 200
When to Retrain Your Models 201
Strategies for Data Retraining 202
Model Retraining in the MLOps Pipeline 203
Deployment Strategies 203
Measuring the Business Impact 206
Conclusion 206
Critical Thinking Discussion Questions 207
Exercises 207

7. Building a Production Grade MLOps Project from A to Z. . . . . . . . . . . . . . . . . . . . . . . . . 209

Exploratory Data Analysis 211
Interactive Data Preparation 220
Preparing the Credit Transaction Dataset 220
Preparing the User Events (Activities) Dataset 223
Extracting Labels and Training a Model 223
Data Ingestion and Preparation Using a Feature Store 224
Building the Credit Transactions Data Pipeline (Feature Set) 225
Building the User Events Data Pipeline (FeatureSet) 228
Building the Target Labels Data Pipeline (FeatureSet) 229
Ingesting Data into the Feature Store 229
Model Training and Validation Pipeline 231
Creating and Evaluating a Feature Vector 232

vi | Table of Contents
 Building and Running an Automated Training and Validation Pipeline 234
Real-Time Application Pipeline 238
Defining a Custom Model Serving Class 238
Building an Application Pipeline with Enrichment and Ensemble 238
Testing the Application Pipeline Locally 240
Deploying and Testing the Real-Time Application Pipeline 241
Model Monitoring 242
CI/CD and Continuous Operations 243
Conclusion 246
Critical Thinking Discussion Questions 246
Exercises 246

8. Building Scalable Deep Learning and Large Language Model Projects. . . . . . . . . . . . . 247
Distributed Deep Learning 248
Horovod 249
Ray 250
Data Gathering, Labeling, and Monitoring in DL 251
Data Labeling Pitfalls to Avoid 252
Data Labeling Best Practices 253
Data Labeling Solutions 254
Using Foundation Models as Labelers 256
Monitoring DL Models with Unstructured Data 257
Build Versus Buy Deep Learning Models 258
Foundation Models, Generative AI, LLMs 259
Risks and Challenges with Generative AI 262
MLOps Pipelines for Efficiently Using and Customizing LLMs 267
Application Example: Fine-Tuning an LLM Model 269
Conclusion 281
Critical Thinking Discussion Questions 281
Exercises 282

9. Solutions for Advanced Data Types. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283

ML Problem Framing with Time Series 284
Navigating Time Series Analysis with AWS 286
Diving into Time Series with DeepAR+ 290
Time Series with the GCP BigQuery and SQL 292
Build Versus Buy for MLOps NLP Problems 296
Build Versus Buy: The Hugging Face Approach 296
Exploring Natural Language Processing with AWS 297
Exploring NLP with OpenAI 303
Video Analysis, Image Classification, and Generative AI 305

Table of Contents | vii

 Image Classification Techniques with CreateML 307
Composite AI 308
Getting Started with Serverless for Composite AI 309
Use Cases of Composite AI with Serverless 312
Conclusion 313
Critical Thinking Discussion Questions 313
Exercises 314

10. Implementing MLOps Using Rust. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315

The Case for Rust for MLOps 316
Leveling Up with Rust, GitHub Copilot, and Codespaces 317
In the Beginning Was the Command Line 321
Getting Started with Rust for MLOps 323
Using PyTorch and Hugging Face with Rust 326
Using Rust to Build Tools for MLOps 330
Building Containerized Rust Command-Line Tools 330
GPU PyTorch Workflows 332
Using TensorFlow Rust 335
Doing k-means Clustering with Rust 336
Final Notes on Rust 337
Ruff Linter 337
rust-new-project-template 337
Conclusion 339
Critical Thinking Discussion Questions 340
Exercises 340

A. Job Interview Questions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341

B. Enterprise MLOps Interviews. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349

Index. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353

viii | Table of Contents

 Preface

As MLOps veterans, we have often seen the following scenario play out across
enterprises building their data science practices.
Traditionally, when enterprises built their data science practice, they would start by
building a model in the lab, with a small team, often working on their laptops and
with a small, manually extracted dataset. They developed the model in operational
isolation, and the results were incorporated manually into applications. Then, once
the model was complete and predicting with accuracy, the true struggle of trying to
bring it to production, to generate real business value, began.
At this point, the enterprise faced challenges such as ingestion of production data,
large scale training, serving in real-time, and monitoring/management of the models
in production. These hurdles would often take months to overcome, presenting a
huge cost in resources and lost time.
The AI pipeline is siloed, with teams working in isolation and with many different
tools and frameworks that don’t necessarily play well with each other. This results
in a huge waste of resources and businesses not being able to capitalize on their
investment in data science. According to Gartner, as many as 85% of data science
projects fall short of expectations.
In this book, we propose a mindset shift, one that addresses these existing challenges
that prevent bringing models to production. We recommend a production-first
approach: starting out not with the model but rather by designing a continuous
operational pipeline, and then making sure the various components and practices
map into it. By automating as many components as possible and making the process
fast and repeatable, the pipeline can scale along with the organization’s needs and
provide rapid business value while answering dynamic and enterprise MLOps needs.
Today, more businesses understand the vast potential of AI models to positively
impact the business across many new use cases. And with generative AI opening
up new opportunities for business innovation across industries, it seems that AI

ix
adoption and usage are set to skyrocket in the coming years. This book explores how
to bring data science to life for these real-world MLOps scenarios.

Who This Book Is For

This book is for practitioners in charge of building, managing, maintaining, and
operationalizing the data science process end to end: the heads of data science, heads
of ML engineering, senior data scientists, MLOps engineers, and machine learning
engineers.
These practitioners are familiar with the nooks and crannies (as well as the challenges
and obstacles) of the data science pipeline, and they have the initial technological
know-how, for example, in Python, pandas, sklearn, and others.
This book can also be valuable for other technology leaders like CIOs, CTOs, and
CDOs who want to efficiently scale the use of AI across their organization, create AI
applications for multiple business use cases, and bridge organizational and technolog‐
ical silos that prevent them from doing so today.
The book is meant to be read in three ways. First, in one go, as a strategic guide that
opens horizons to new MLOps ideas. Second, when making any strategic changes to
the pipeline that require consultation and assistance. For example, when introducing
real-time data into the pipeline, scaling the existing pipeline to a new data source/
business use case, automating the MLOps pipeline, implementing a Feature Store, or
introducing a new tool into the pipeline. Finally, the book can be referred to daily
when running and implementing MLOps. For example, for identifying and fixing a
bottleneck in the pipeline, pipeline monitoring, and managing inference.

Navigating This Book

This book is built according to the phases of the MLOps pipeline, guiding you
through your first steps with MLOps up to the most advanced use cases:

• Chapters 1–3 show how organizations should approach MLOps, how data sci‐
ence teams can get started, and what to prepare for your first MLOps project.
• Chapters 4–7 explain the components of a resilient and scalable MLOps pipeline
and how to build a machine learning pipeline that scales across the organization.
• Chapter 8 covers deep learning pipelines and also dives into GenAI and LLMs.
• Chapters 9 and 10 show how to adapt pipelines for specific verticals and use
cases, like hybrid deployments, real-time predictions, composite AI, and so on.

Throughout the book, you will find real code examples to interactively try out for
yourself.

x | Preface
After reading this book, you will be a few steps closer to being able to:

• Build an MLOps pipeline.

• Build a deep learning pipeline.
• Build application-specific solutions (for example, for NLP).
• Build use-case specific solutions, (for example, for fraud prediction).

Conventions Used in This Book

The following typographical conventions are used in this book:
Italic
Indicates new terms, URLs, email addresses, filenames, and file extensions.
Constant width
Used for program listings, as well as within paragraphs to refer to program
elements such as variable or function names, databases, data types, environment
variables, statements, and keywords.
Constant width bold
Shows commands or other text that should be typed literally by the user.
Constant width italic
Shows text that should be replaced with user-supplied values or by values deter‐
mined by context.

This element signifies a tip or suggestion.

This element signifies a general note.

This element indicates a warning or caution.

Preface | xi
Using Code Examples
Supplemental material (code examples, exercises, and so on) is available for down‐
load at https://github.com/mlrun/demo-fraud and https://github.com/mlrun/demo-llm-
tuning.
If you have a technical question or a problem using the code examples, please send
email to [email protected].
This book is here to help you get your job done. In general, if example code is
offered with this book, you may use it in your programs and documentation. You
do not need to contact us for permission unless you’re reproducing a significant
portion of the code. For example, writing a program that uses several chunks of code
from this book does not require permission. Selling or distributing examples from
O’Reilly books does require permission. Answering a question by citing this book
and quoting example code does not require permission. Incorporating a significant
amount of example code from this book into your product’s documentation does
require permission.
We appreciate, but generally do not require, attribution. An attribution usually
includes the title, author, publisher, and ISBN. For example: “Implementing MLOps in
the Enterprise by Yaron Haviv and Noah Gift (O’Reilly). Copyright 2024 Yaron Haviv
and Noah Gift, 978-1-098-13658-1.”
If you feel your use of code examples falls outside fair use or the permission given
above, feel free to contact us at [email protected].

O’Reilly Online Learning

For more than 40 years, O’Reilly Media has provided technol‐
ogy and business training, knowledge, and insight to help
companies succeed.

Our unique network of experts and innovators share their knowledge and expertise
through books, articles, and our online learning platform. O’Reilly’s online learning
platform gives you on-demand access to live training courses, in-depth learning
paths, interactive coding environments, and a vast collection of text and video from
O’Reilly and 200+ other publishers. For more information, visit https://oreilly.com.

xii | Preface
How to Contact Us
Please address comments and questions concerning this book to the publisher:

O’Reilly Media, Inc.

1005 Gravenstein Highway North
Sebastopol, CA 95472
800-889-8969 (in the United States or Canada)
707-829-7019 (international or local)
707-829-0104 (fax)
[email protected]
https://www.oreilly.com/about/contact.html

We have a web page for this book, where we list errata, examples, and any additional
information. You can access this page at https://oreil.ly/mlops-in-the-enterprise.
Email [email protected] to comment or ask technical questions about this
book.
For news and information about our books and courses, visit https://oreilly.com.
Find us on LinkedIn: https://linkedin.com/company/oreilly-media.
Follow us on Twitter: https://twitter.com/oreillymedia.
Watch us on YouTube: https://youtube.com/oreillymedia.

Acknowledgments
We’d like to thank the people behind the scenes who assisted, guided, and supported
us throughout this book’s journey. Without them, this book wouldn’t have been
brought to life.
Thank you to the dedicated team at O’Reilly, who provided feedback and guidance,
drove the writing process of this book, and helped polish the content. We’d especially
like to thank Corbin Collins for being our partner throughout the process, paying
close attention to all the details and helping us meet deadlines, and to Nicole Butter‐
field, for her unwavering support and valuable input.
We’re deeply appreciative of our tech reviewers, Dhanasekar Sundararaman, Tigran
Harutyunyan, Nivas Durairaj, and Noga Cohen for their expertise and wisdom.

Preface | xiii
Another Random Scribd Document
with Unrelated Content
 AGED 38.'"

"I don't like that one bit," said Stephen; "it has got too
many hard words in it."

"Well, here's another."

"'IN MEMORY
OF
JOHN POWELL.
DIED IN 1781.
ALSO MARY, RELICT OF
THE ABOVE, WHO DIED
JANUARY 20, 1827,
AGED 87.
ALSO TWO GRANDCHILDREN,
WHO DIED YOUNG.'"

"That's much nicer," said Stephen. "I like those two

grandchildren who died young. I wonder how old they were;
do you think they were as old as you and me, Audrey?"

"I don't know," said Audrey; "it doesn't say, and it

doesn't tell if they were girls or boys."

"Never mind," said Stephen, "we can guess. I think one

was a girl and one was a boy. And are their bodies really
down under here, Audrey?"

"Yes, what there is of them," said Audrey; "Aunt

Cordelia says they turn to dust."
 "Oh," said little Stephen, in an awestruck voice, "I wish
we could see the dust of the two grandchildren who died
young! I'll have this grave, Audrey, and take care of them.
Is there any one else inside it?"

"Yes, there's John Powell, died in 1781; also Mary, relict

of the above," read Audrey.

"What does relict mean?" asked Stephen.

"Aunt Cordelia has a relict," said Audrey, "and she

keeps it in a box."

"Is it a woman?" asked Stephen.

"No, it's a bit of grey hair; she cut it off her mother's
head when she was dead, and she says it's a relict. I don't
know what she means, but she keeps it locked up ever so
safe."

"I hope John Powell didn't lock Mary up," said Stephen.

"She must have got out if he did," said Audrey, "for she
lived a long, long, long time after him. He died in 1781, and
she didn't die not until 1827; let me count up, it's quite a
long sum. Why, it's forty-six years, Stephen!"

"Oh dear," said Stephen, "that is a long time! Let's tell

Granny Robin about it, and I'll ask her if she would have
that one if she was me."

Granny Robin quite approved of their plan, and of

Stephen's choice of the two grandchildren who died young.
She told them that relict meant the wife left behind, and
tears came into the old woman's sightless eyes, as she sat
at her knitting and thought of the poor widow left behind for
forty-six years. She pictured her living on and on, year after
year, coming doubtless often to that grave to look at the
place where her John lay, but still kept waiting for forty-six
years for the glad day when she should see him again.

Granny Robin thought it must have seemed a longs

dreary time to poor Mary. And then, maybe, those two
grandchildren were a cheer and comfort to her. Yet they
were taken, they died young, but old Mary still lived on. Till
at last, on that winter's day, January 20, 1827, the call, so
long waited for, came, and she and her John were together
again. Then, too, the old grandmother saw once more the
faces of the two grandchildren who died young.

So Granny Robin mused as she sat at her work; and she

wondered whether the waiting-time seemed as long to old
Mary, as she looked back to it from the brightness and the
joy of the Home above, or did it seem short as a troubled
dream seems when we wake from sleep?

"Our light affliction, which is but for a moment."

So long, when we are passing through it; but for a

moment, as we look back to it from God's eternity.

CHAPTER V
The Collection
 STEPHEN had now quite settled upon the grave which
he was to make his especial care, but he promised not to
begin his work until Audrey had chosen hers. She was very
undecided for a long time, but at length she chose one,
sacred to the memory of another John.

"It will be nice for us each to have a John," she said.

"'BENEATH IS DEPOSITED
ALL THAT WAS MORTAL OF
JOHN HUTTON,
WHO DIED THE 12TH OF APRIL, 1793,
AGED 47.'"

"'Go home, dear wife, and shed no tear,

I must ly here till Christ appear;
And at His coming hope to have
A joyful rising from the grave.'"

"How do you spell lie, Granny Robin?" said Audrey,

when she had finished reading it to her.

"L-i-e," said the old woman.

"Well, it's l-y here," said the child.

"That's the old-fashioned way," said Granny Robin.

"Well, now, we'll set to work," said Audrey; "we must

wash them first, Stephen. Do you think your father would
give us some water in a basin? I daren't ask Aunt Cordelia;
she would say I should dirty my pinafore."

"If Stephen's father will give him a basin, I will give you
one, Audrey," said Granny Robin.

"And I'll get you both an old sponge," said Mr. Robin,
who was smoking his pipe in the window.

What a scrubbing went on after that! Stephen's father,

who was always pleased to do anything his poor little boy
asked him, brought out soap and two scrubbing brushes,
and the children worked away diligently for more than an
hour.

At the end of it, they were far from satisfied with their
work.

"The two grandchildren who died young won't come

clean, Granny Robin," said little Stephen mournfully.

"They're quite as nice as my John is," said Audrey.

"Anyhow," she added more hopefully, "they're a deal
cleaner than they were before. Now what's the next thing to
be done?"

"We must cut the long grass behind them," said

Stephen, "and then we must dig up the grave in front of the
stone. I'll get father's big scissors and my little spade."

Father's big scissors cut the grass down very

successfully, but Stephen's little spade refused to go into
the hard ground. It had been trodden underfoot for many
years, and it lay hard and dry and stony over the heads of
the two grandchildren who died young.
 But at this point old Mr. Robin came to the rescue. He
brought a large spade out of his house and dug the grave
over for little Stephen, and then, after he had rested a little,
he did the same for Audrey's John, as she called him.

AT THIS POINT OLD MR. ROBIN CAME TO THE RESCUE.

"Wouldn't the wife be pleased if she saw we were doing

it?" she said.

"What wife?" asked Stephen.

"This wife it says about in the hymn—"

"'Go home, dear wife, and shed no tear.'"

 "I wonder if she did shed any," said Stephen.

"I expect she did," said Audrey; "I wonder what has
become of her. Do you think she will ever come to see how
nice we have made her John's grave, Granny Robin?"

"When did John die?" asked the old woman.

"In 1793," said Audrey.

"1793—a hundred years ago!" said Granny Robin. "Why,

Audrey, the wife must have been dead long since!"

"And she never sheds any more tears now," said

Stephen, "because she's in heaven."

"I hope so," said Granny Robin.

"Does everybody go to heaven when they die?" asked

the child.

"No, my dear boy, not every one."

"Shall I go there when I die, Granny Robin? I do hope I

shall," said little Stephen.

"I hope so too, my little man. The Lord wants to have

you there," she said.

"What is it like, Granny Robin?" asked Stephen.

"We know very little about it, Stephen," said the old
woman, "but we can't help thinking about it, and dreaming
about it; and I always think of it as a beautiful garden,
where the King walks with His friends. I may be wrong,
Stephie, but that's what I always see in my mind when I
think of it."
 "The two grandchildren who died young will like being in
the garden," said Stephen. "Do you think they're glad they
died young, Granny Robin?"

"I think they are, Stephie," she said; "they did not have
to tread far on life's rough ways; their little feet reached the
garden long, long years ago."

"And there will be soft grass for them to walk on there,"

said Stephen, "Maybe I'll see them when I get there. Do
you think I'll know them, Granny Robin?"

"I think you will, Stephie; I feel almost sure you will,"
she said.

"If I see any very dear little children playing under the
trees of the garden," said little Stephen, "I might ask them,
'Are you the two grandchildren who died young?' And then
they could tell me, couldn't they?"

"God bless you, my dear little lad!" was all the answer
Granny Robin gave him.

The next day was Saturday, which was market-day in

the old city. It was Audrey's holiday, and the happiest day in
the week to Stephen and to herself. Aunt Cordelia was
always busy cleaning from morning till night, and sent
Audrey into the churchyard, that she might be out of the
way of her sweeping-brush and dust-pan.

On this particular Saturday, Audrey and Stephen were

whispering together under the lilac tree for a very long
time; and about ten minutes afterwards, Mr. Robin, who
was smoking his pipe in the window, saw a sight which
made him laugh so much, that for a long time he could not
tell Granny Robin at what he was laughing.
 As he looked across the churchyard, he saw Audrey and
Stephen coming towards the window arm in arm. Stephen
was dressed in the tall hat which his father wore when he
went to chapel on Sunday night, and in an old greatcoat,
which was fastened round his neck, and dragged like a long
tail behind him, whilst the sleeves were turned up so far
that there was far more lining than cloth to be seen. Audrey
had a red shawl thrown over her head, and her pinafore
was tied round her waist like an apron. Each child carried a
tin, on which old Mr. Robin distinctly read the words
"Colman's Mustard."

As soon as they came up to the window both children

made a low bow, but neither of them spoke.

"Well, what do you want?" said Mr. Robin, as gravely as

he could. "Are you going round begging this fine spring
morning?"

"Please, sir, we're making a collection," said Audrey.

"Yes, it's a collection," echoed little Stephen.

"What's it for, my little dears?" said Granny Robin, as

she laid down her knitting, and began to put her hand into
her pocket.

"Mine's for the TWO GRANDCHILDREN WHO DIED

YOUNG," said little Stephen.

"And mine's for ALL THAT WAS MORTAL OF JOHN

HUTTON," said Audrey.

"Oh, I see," said the old woman; "you want to go and

get some roots in the market for your graves—is that it?"
 That is it, and Granny Robin's hand must go in the
pocket again. It goes in empty, but it comes out well filled.
Three pennies for the grandchildren go into Stephen's tin,
and three more for John Hutton go rattling to the bottom of
Audrey's.

Now it is Mr. Robin's turn, and his pocket seems to be

full of pennies too; and the tins make such a noise when
they are shaken that Granny Robin pretends to stop her
ears, that she may not hear the din.

Then the two children go on to the next window, where

Stephen's father sits busy with his work. But the boot is laid
down, that the collection may have due attention, and it is
silver this time which goes into the tins, two quiet silver
threepences, which make no noise, but which the two
children admire greatly as they slip in amongst the copper.

"Now for Aunt Cordelia," says Audrey. "You must go

first, Stephen; she won't say 'No' to you."

Aunt Cordelia makes a dive at Audrey's pinafore, the

bottom of which she declares is collecting all the dust in the
churchyard, but she is not angry when she hears why they
have come. And when Stephen pleads for something for his
two grandchildren, she goes to her till and brings out
several pence for each tin, and willingly gives Audrey leave
to go that afternoon to the market with Mr. Robin to make
her purchases.

CHAPTER VI
 Angels' Visits

WHAT an important little person Audrey was, as she set

out to do her marketing that afternoon! Stephen was not
able to go, for the crowd in the market-place was so great
on Saturday afternoon, that his father was afraid he might
get hurt. So Audrey and the old man were to do the
business between them; Audrey carried the money, and Mr.
Robin brought a basket for the flowers.

The market was an open one, and was held in a wide

street in the centre of the city. There were stalls for all
manner of articles in that market—toys, and kettles, and
tins, and slippers, and caps, and all sorts of other things;
but the flower-stalls were by far the prettiest, Audrey
thought, and these were placed by themselves, all down
one side of that long street. The little girl went from stall to
stall, admiring all the flowers, and wondering which Stephen
would like best.

It was well that Mr. Robin was there to help her to

decide, or Stephen's patience would have been exhausted
long before she reached home. He was sitting at the
window looking out for them the whole time they were
away. And oh, what excitement there was when the basket
was unpacked, and the contents spread out on Granny
Robin's round table!

Then, when all had been duly admired, they were

divided into two heaps. One heap was for Stephen's
grandchildren, as he called them, and the other was for
Audrey's John. There were a yellow and a purple pansy, a
red and a white daisy, a yellow musk, a sweet william, a
primrose, a violet, a lily of the valley, and two or three
beautiful roots of forget-me-nots in each heap.

Then the children went out in great glee to plant their

flowers. But what was their surprise to find that, whilst
Audrey had been in the market, Stephen's father had been
very busy bringing bucketfuls of earth from the garden of a
friend of his who lived not far away, and making the two
graves as tidy and neat as the daintiest flower-garden. It
was easy to plant the roots after this; and oh, how
delighted the children were, as they saw the graves growing
more and more pretty every moment!

And that evening there was a grand procession. Every

one was invited to see and to admire their work. Mr. Robin
walked first, with Granny Robin on his arm. The old woman
had insisted on climbing out of her window to visit the
graves. If she could not see them, she could feel them, she
said, and she could smell the flowers.

As for Stephen's father, he had done very little shoe-

mending all day, for every few minutes he had come
hopping out of his window to see how they were getting on.
Yet, although he had helped to choose the place of each
plant, the little cobbler still came behind Mr. and Mrs. Robin
in the procession which was about to visit the graves, and
when he arrived there, he seemed as much surprised and
interested as if he had never seen them before.

The difficulty was to get Aunt Cordelia there. Not that

she was unwilling to come, for she was anxious to see
Granny Robin, of whom the children talked so much; but
the trouble was this, she could not make up her mind to
climb out of the window. It took Audrey and Stephen nearly
an hour to coax her to make the attempt. She even wanted
to go down the street to the house of the deaf old woman,
that she might get the key of the churchyard gate.

It was only when Audrey told her that if she did so, it
would spoil everything—for the old woman would be sure to
come with her, and would perhaps be angry with them for
doing it without her leave—it was only then that Aunt
Cordelia consented to try the undignified descent.

But it was a terribly serious business. A stool was

placed outside the window, and Mr. Robin and the cobbler
came forward to give her a hand, whilst she gathered her
petticoats round her, and at length, slowly and gracefully,
managed to alight on the churchyard grass. Then the
procession began, and the children's work was duly admired
by the whole party. They all had some remark to make
about it, and these remarks were very different from each
other.

Aunt Cordelia, who was in a very good temper, and who

was much gratified by the politeness of Mr. and Mrs. Robin,
said, "Well, I declare, it's a very pretty garden, and a deal
better play than climbing all day long over those black,
filthy old stones, Audrey. You won't dirty half as many
pinafores!"

Stephen's father was full of his boy's delight. It would

be a pleasure for Stephen every day, he said, and he would
buy him a nice little red can with which to water the
flowers.

Mr. Robin said it was a real treat to see a bit of flower-

bed again; it reminded him of his garden in the country,
and was like a bit of home to him.

But Granny Robin, as she knelt on the grass to smell

the flowers, repeated softly to herself the words of a verse,
which Audrey and Stephen thought very beautiful.

"'Saint after saint on earth

Has lived, and loved, and died,
And as they left us one by one,
We laid them side by side;
We laid them down to sleep,
But not in hope forlorn;
We laid them but to ripen there
Till the last glorious morn.
Come, then, Lord Jesus, come!'"

"Do you think they know what we've been doing?" said
little Stephen.

"Who, my dear child?"

"The two grandchildren who died young."

"I don't know," said Granny Robin; "I can't say,

Stephen."

"Perhaps the angels will tell them when they go back to-
night," said Stephen. "They are sure to notice it when they
come to look at the graves, and I think the little children
will be glad when they hear."

And that night, long after Stephen's father thought he

was fast asleep, the little boy stood at his bedroom window
in the moonlight, looking for the angels. The calm, quiet
light was streaming through the trees and down upon the
desolate graves. It made even the saddest of them look
beautiful, little Stephen thought, and he fancied that the
moonbeams must be the reflection of the brightness of the
angels' wings.

His own grave, as he loved to call it, was lying full in the
pure, silvery light. He could see the flowers he had planted
distinctly, and he could even distinguish some of the words
on the old tombstone. He loved to fancy to himself that the
angels were glad to see it looking so beautiful, that they
were pleased with what he had done, and that they were
lingering round it with bright and happy faces. Some of the
other graves were lying in shadow, but the angels, so he
thought, had gathered round the one upon which he had
bestowed so much care, and were unwilling to leave it
behind.

It was not until clouds came drifting across the sky, and
one of them was driven over the face of the moon, and the
whole churchyard was left in darkness again; it was not
until every ray of moonlight had disappeared, that little
Stephen crept back to bed. The angels were gone, he said,
as he laid his head on the pillow; they had flown away to
the King's Garden, and perhaps, even then, they were
telling the two grandchildren who died young that the
flowers were blooming on their grave, and that it was no
longer forsaken and desolate.

CHAPTER VII
The Mysterious Light
 "GRANNY ROBIN—" said Stephen, when he and Audrey
were leaning on her window-seat on the bright Sunday
afternoon which followed that busy Saturday, "Granny
Robin, do you think I shall die young?"

"I can't tell, my dear child," said the old woman, as she
stroked Stephen's little thin hand; "only the dear Lord
above knows that."

"I think I should like to die young," said the child.

"If you go to heaven," said Audrey, "it won't be a good

thing if you don't."

"Shall I go to heaven, Granny Robin?" asked the little

boy.

"If you have come to Jesus you will, Stephen," she said.

"I would like to come to Jesus, Granny Robin," said

Audrey; "but how can I come to Him? If He was in the city,
and I knew which house He was in, I would go to Him—
wouldn't you, Stephie?"

"Yes," said the child; "we would take hold of hands and
go together."

"You have not far to go," said Granny Robin, as she laid
down her knitting and put her arms round the two children.

"Tell us just what we must do," said Audrey, "and we'll

do it, Stephen and me, both of us, Granny Robin."

"You have a river here, haven't you, somewhere in the

city?" asked the old woman.
 "Yes," said Audrey; "and it's beautiful in summer-time,
it's covered with boats."

"And barges," said Stephen.

"Yes, but the barges are ugly," said Audrey. "But the
boats are lovely, and sometimes they have a sail up, and
then I like them best of all. One day me and Stephen went
down on the new walk by the river, and we sat on one of
the seats and watched them."

"Have you ever been over the river?" said Granny

Robin.

"Yes, heaps of times," said Audrey.

"Did you swim across?" asked the old woman.

"Oh no, Granny, I should be drowned if I fell in! There's

bridges, you know—great big bridges. There's the pay
bridge, where you pay a halfpenny to go across, and there's
two more; but Aunt Cordelia always goes over the pay
bridge."

"Are you afraid of falling in when you're on the bridge,

Audrey?"

"Oh no; it's so strong, Granny Robin. Hundreds of folks

go over it every day."

"Audrey," said the old woman, "do you know, I once

lived in a country which had a very dismal name? It was
called the Kingdom of Darkness."

"I shouldn't like to go there," said Stephen. "I don't like

to be in the dark."
 "I was born there," said Granny Robin, "and I did not
notice how dark and gloomy it was. I was used to it, and it
did not strike me as being very dismal, not for a long, long
time."

"How did you find out it was dark, Granny Robin?"

"I looked across the river to the other side."

"And what did you see there? asked Stephen.

"I saw another kingdom, Stephie, which was full of

light; glorious sunshine was streaming on it; there was not
a dark corner to be seen in it."

"Had that country a name?" said Audrey.

"Yes, it was named the Kingdom of Light."

"Didn't you want to go there, Granny Robin?"

"Yes, I was not happy in the Kingdom of Darkness any

longer," she said. "It looked black as night to me, and I
wanted—oh, so much!—to get to the other side."

"Why didn't you go, Granny Robin?"

"I couldn't find the way, Stephie. I tried to get across,

but the water was too deep, and I had to turn back."

"Did you ever get over?" asked Audrey.

But just then Aunt Cordelia's voice was heard calling

loudly, "Tea, Audrey—tea," and the little girl had to run
home without hearing the old woman's answer.

The evening was dark and gloomy. Clouds came driving

up and covered the blue sky, and the wind blew mournfully
amongst the forlorn trees in the churchyard. Granny Robin's
window was closed, and Stephen whispered to Audrey,
when she came out, that he heard strange voices inside,
and that his father said two people had come to see Mr.
Robin from the village where he used to live.

Audrey and Stephen wandered about the churchyard

together, but it was very dismal that Sunday evening; even
the flowers on the two graves did not look fresh and
beautiful, as they had done the night before.

After a time, they climbed on the square tomb and

peeped into the church, but it seemed more gloomy there
than it did outside. Even the swallow had settled down on
his nest, as if he felt too depressed to venture to fly into the
churchyard.

"It looks like the Kingdom of Darkness," said Stephen in

an awestruck voice. "I wonder if Granny Robin ever got
across; don't you, Audrey? Shall we go in now?"

"No, let's stay outside," said the little girl. "Your father's
at chapel, and Aunt Cordelia's at church, and it's much
darker in than out."

So they wandered about for another half-hour, and then

even Audrey owned it would be better to go in.

"Let's have one more look in the church," said Stephen;

"I want to see if the swallow has gone to sleep."

They climbed on the stone, but they could see nothing.

The old church was quite dark now, and Stephen tried in
vain to see the swallow in its nest. They could only
distinguish the outline of the chancel window, and Audrey
thought she could see the pulpit with its heavy top, but she
could not be sure even of that.
 "Let's go," said little Stephen, shivering; "it looks more
like the Kingdom of Darkness than before."

"LOOK, STEPHEN!" SHE CRIED. "WHAT'S THAT?"

He was climbing quickly down from the gravestone

when Audrey called him.

"Look, Stephen," she cried. "What's that?"

The child got on his feet again and pulled himself up to

the window. When he had looked into the church a moment
before all had been dark, but now a bright light streamed
across the chancel. They could see the old, crooked stone
pillars standing out clearly against it; they could distinguish
the communion table, and the wooden rail in front of it;
they could see the high pews and the uneven stone floor;
they could even make out the swallow's nest in the arch
nearest to the pulpit.
 "What can it be?" said Stephen, trembling from head to
foot.

"Some one must be inside," said Audrey. "Let's watch."

"It must be old Maria," said Stephen.

"No, it can't be Maria; we should have seen her come,"

said Audrey. "Why, we've been sitting looking at that little
stone path leading from the gate for nearly an hour. I am
quite sure it isn't Maria!"

"Who can it be?" said Stephen. "Oh, Audrey, let's fetch

father?"

"No, wait a bit," said Audrey; "your father won't be in.

Let's watch it; perhaps we shall see some one."

But although they watched for a long time, no one

appeared in the church, nor were they able to discover from
whence the light came.

It was not a steady light; it flickered up and down, and

the shadows on the roof flickered with it. Nothing else
moved in the old church; all else was still as death. But
Stephen's heart was beating faster and faster as the
minutes went on; and Audrey, feeling how much he
trembled, was just going to yield to his wish to go home,
when quite suddenly, as suddenly as it had appeared a few
minutes before, the light went out, and the old church was
once more left in darkness.
 CHAPTER VIII
Children of Light

IT was a little hard on the children, when they went

home to tell their tale of the strange light in the church, to
find that no one would believe them. Aunt Cordelia was
inclined to be angry, and said:

"Whoever heard of a light in the old church? They

shouldn't make up such stories!"

Stephen's father only laughed, and told him he must

not think of it again; little boys fancied strange things
sometimes. Even Granny Robin seemed to imagine that
their being out late, and their having felt rather frightened,
had something to do with it.

But Stephen and Audrey knew better, and, without

saying a word to any one, they would creep out night after
night, and would climb upon the fiat tomb that they might
look for the sight.

But day after day of that week went by, and they never
saw it again.

"Perhaps it only comes on Sunday," said Stephen.

"We must wait and see," said Audrey. "I do hope it will
come, and then we will fetch them to see it, and they will all
believe it."

Meanwhile the two children had another talk with

Granny Robin.
 "Did you ever get into the Kingdom of Light?" little
Stephen asked her.

"Yes, my child," said the old woman, "I did."

"And how long did you stay there?" said Audrey.

"Weren't you sorry to come away, Granny Robin?"

"I never came away," said the old woman; "I'm there
now!"

"There now?" repeated Stephen; and he looked round

the room, which was fast growing dark, and was full of the
shadows of the trees outside. "Did you say you were there
now, Granny Robin? I don't think it's very bright now; and
then you're blind, you know, and couldn't see it if it was."

"It's always bright in the Kingdom of Light, and I can

see that sunshine, Stephen," she said.

"I think Granny Robin means her soul can see," said
Audrey.

"Quite right, dear child," said the old woman; "it is my

soul, and not my poor old body, that is in the Kingdom of
Light."

"I wish I was in the Kingdom of Light," said Audrey.

"You must cross over, Audrey," said Granny Robin.

"How can we cross?" asked little Stephen.

"Just as I did," she said. "You can't walk over or swim

over; there is only one way."

"What way is that?" asked Audrey.

 "The bridge," said Granny Robin quietly; "you have
several bridges over your river—I have only one over mine."

"Is it the pay bridge?" asked Stephen.

"No, it's the free bridge," said Granny Robin, smiling

—"'without money and without price.'"

"What is its name, Granny Robin?"

"Christ Jesus," said the old woman reverently. "Did not

He say—"

"'I am the way; no man cometh unto the Father but

by Me?'"

"Tell me how you got across," asked Audrey.

"I came to Jesus," said the old woman. "I said to

myself, 'My Lord is not far away in heaven; He is close
beside me in this room, standing by my side, waiting for me
to come to Him. I cannot take myself across into the
Kingdom of Light, but He can take me. He died on the cross
that He might be able to take me. I will trust myself to Him.
Hundreds of others have crossed by this bridge, and have
crossed safely; I will cross also. I will come to Christ to be
saved by Him alone.' And, children, I did cross, and here I
am," said the old woman, smiling again.

"I want to go into the Kingdom of Light," said little

Stephen, laying his small, thin hand on Granny Robin's.
 "Trust yourself to Him this night, Stephen," she said;
"you and Audrey, both of you. Tell Him you want to come to
Him, just now, that He may be your bridge, and take you
safe over from the Kingdom of Satan to the Kingdom of the
dear Son, the blessed Kingdom of Light."

And with a child holding each hand, the old woman

knelt down and prayed—

"O Lord, Thou art close beside us, and we come to

Thee. We want, each one of us, to belong to the Kingdom of
Light. Little Audrey and Stephen are coming to Thee now;
Thou art the way to God. We trust ourselves to Thee to be
saved. Thou hast taken many others out of the darkness
and into the light. Lord, we believe Thou wilt take us, just
now, just as we are. As we kneel here, hand in hand, we
trust ourselves to Thee. Lord, take us now, we do beseech
Thee. Amen."

And then, as they sat beside her, and as she went on

with her knitting, Granny Robin sang softly—

"Out of my bondage, sorrow, and night,

Jesus, I come—Jesus, I come!
Into Thy freedom, gladness, and light,
Jesus, I come to Thee.
Out of my sickness into Thy health,
Out of my want and into Thy wealth,
Out of my sin and into Thyself,
Jesus, I come to Thee!"

"Now, Audrey and Stephen," said Granny Robin, "if you

have come into the Kingdom of the dear Son—the Kingdom
of Light—you must remember what you are."

"What are we, Granny Robin?"

"Children of Light," said the old woman, smiling. "Isn't

that a beautiful name?"

"It sounds like an angel's name," said little Stephen.

"And the Children of Light must never do the works of

darkness," said Granny Robin. "If you are tempted to be
cross, or disobedient, or untruthful, you must say to
yourself, 'I am a Child of Light; I have crossed over the
bridge, and all those deeds of darkness must be left behind
in the Kingdom of Darkness.' Will you remember, Audrey
and Stephen?"

Just then Aunt Cordelia's voice was heard calling for

Audrey to come to her at once. It was an impatient, angry
voice, and Audrey said, as she got up reluctantly to go—

"What a bother! There's Aunt Cordelia calling, and she's

as cross as two sticks!"

"Children of Light, remember, Audrey," said Granny

Robin softly, as she jumped out of the window.

It was Friday, and Aunt Cordelia's baking-day. And if

there was one day in the week when Aunt Cordelia was
more cross than usual, that day was baking-day. Standing
over the large oven and the scorching fire, baking cakes
and pies and buns for her shop, with the perspiration
streaming down her face, it was no wonder that Aunt
Cordelia's temper was tried.

"Come along, Audrey, you lazy child!" she cried, as she

took a tray of cakes out of the oven. "Here am I, slaving
away this hot day, and you doing nothing but waste your
time. There's that shop bell been tinkling as if it was mad
this last hour; and how in the world am I to get my baking
done, if I'm running backwards and forwards every minute!"

Audrey was just going to say that she had been at

school all day, and it was very hard if she couldn't have a bit
of play when she came home; but she remembered Granny
Robin's words.

"I am a Child of Light," she said to herself; and was

quiet.

And when Aunt Cordelia called after her, as she was

going into the shop, "Audrey, put a clean pinafore on!
Audrey, I never saw such a dirty girl as you are! You're not
fit to be seen!"

Audrey went quietly upstairs without a word, changed

her pinafore, and came down with a bright and pleasant
face to take her place behind the counter.

It was wonderful how happy she felt, even though she

knew it was the time when she and Stephen watered the
graves, and he would be waiting for her outside. And at tea-
time Aunt Cordelia put one of her best cakes on her plate,
telling her that she had been a good child, and she might go
and play after tea.

That evening Stephen and she talked a great deal about

the Children of Light; and Stephen said he wondered if the
angel's, when they came to look at the graves to-night,
would come and look at them as they lay in their little beds.

"Because you know, Audrey, I think if we are Children of

Light, we must be their little brothers and sisters—don't you
think so, too?"
 When Sunday evening came, the children were very
anxious to watch the old church. They sat for a long time
keeping a strict look out on the gate, for they fancied that
old Maria must have some special business in the church on
Sunday, and that they might see her come in.

But though they watched carefully for more than an

hour, no one came to turn the rusty lock; and at last, when
darkness came on, the children crept somewhat tremblingly
to peep into the church.

Yes—there was the light again, flickering as before in

the chancel; and yet they could see no sign of any one in
the church.

"Now," said Audrey, "we'll make them believe us. You

fetch Mr. Robin, Stephen, and I'll fetch Aunt Cordelia; she
hasn't gone to church to-night, and we'll show it to them."

Aunt Cordelia refused to come; she was wearing her

Sunday dress, and would spoil it, scrambling on those dirty
stones, she said.

But Mr. Robin put on his cap, and came with Stephen as
quickly as he could.

"Now you will see, Mr. Robin," Audrey whispered to him,

as he climbed on the flat stone and looked in at the window.
"What do you think it can be?"

"I can't see anything," said Mr. Robin; "it's all dark
inside."

The light had entirely vanished—not the faintest

glimmer was to be seen; and poor Audrey and Stephen
were as far as ever from convincing those at home that it
had ever appeared, except in their imaginations.
 "It all comes of wandering about amongst those
graves," said Aunt Cordelia. "It isn't good for children. But
what can we do with them? We can't turn them out into the
streets to play."

CHAPTER IX
Under the Yew Tree

IT was wonderful what a difference that talk with

Granny Robin had made in Audrey's and Stephen's lives.
They never forgot what she had said to them, or that they
were the Children of Light. And when Audrey's birthday
came, Granny Robin gave her a beautiful text to hang up in
her bedroom; and the words of the text were these—

"WALK AS CHILDREN OF LIGHT."

"How ought Children of Light to walk, Granny Robin?"

asked Audrey. "What does it mean?"

"It means, behave as Children of Light," said the old

woman; "do nothing and say nothing which the Children of
Light ought not to do or say."

"I will try, Granny Robin," said Audrey.

 And she did try, from that day forward. When Aunt
Cordelia was in one of her difficult, fault-finding moods,
Audrey would say to herself:

"I am a Child of Light," and would keep back the angry,

fretful answer.

When she was tempted to be idle at school, or to join in

the disturbance which most of the children were making,
she would say again:

"I am a Child of Light," and the thought would make her

work with all her might.

When she lay awake at night, and felt lonely in the little
top room where she slept, she would creep out of bed and
look into the churchyard, and above the old tower to the
stars shining overhead, and would say to herself:

"I am a Child of Light."

And then she never felt afraid; for were not the angels
watching over all the Children of Light, their little brothers
and sisters in the Kingdom?

As for Stephen, he was happier than he had ever been

before; it was his one thought night and day.

He was always asking his father, "Father, are you one of

the Children of Light? Have you crossed over the bridge?"

And when he received no answer to his question, he

would throw his arms round his father's neck, and would
say, "Oh, I do hope you are in the Kingdom of Light, father
—the Kingdom of the dear Son!"