Professional Documents
Culture Documents
(Download PDF) Electronic Noses and Tongues in Food Science 1St Edition Maria Luz Rodriguez Mendez Online Ebook All Chapter PDF
(Download PDF) Electronic Noses and Tongues in Food Science 1St Edition Maria Luz Rodriguez Mendez Online Ebook All Chapter PDF
https://textbookfull.com/product/food-carotenoids-chemistry-
biology-and-technology-1st-edition-delia-b-rodriguez-amaya/
https://textbookfull.com/product/food-quality-balancing-health-
and-disease-1st-edition-alina-maria-holban/
https://textbookfull.com/product/emerging-nanotechnologies-in-
food-science-1st-edition-busquets/
https://textbookfull.com/product/the-chemistry-of-thermal-food-
processing-procedures-1st-edition-maria-micali/
Electron Spin Resonance in Food Science 1st Edition
Shukla
https://textbookfull.com/product/electron-spin-resonance-in-food-
science-1st-edition-shukla/
https://textbookfull.com/product/chromatography-in-food-science-
and-technology-first-edition-cserhati/
https://textbookfull.com/product/handbook-of-food-science-and-
technology-1-food-alteration-and-food-quality-1st-edition-romain-
jeantet/
https://textbookfull.com/product/essentials-of-food-science-food-
science-text-series-vaclavik/
https://textbookfull.com/product/tongues-of-fire-language-and-
evangelization-in-colonial-mexico-nancy-farriss/
Electronic Noses and
Tongues in Food Science
Series Editor
Victor Preedy
Department of Nutrition and Dietetics,
King’s College London,
United Kingdom
Editor
María Luz Rodríguez Méndez
Group of Sensors
Industrial Engineering School
University of Valladolid, Spain
No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, in-
cluding photocopying, recording, or any information storage and retrieval system, without permission in writing from the
publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our ar-
rangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found
at our website: www.elsevier.com/permissions.
This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may
be noted herein).
Notices
Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understand-
ing, 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 in-
formation, 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.
ISBN: 978-0-12-800243-8
Numbers in paraentheses indicate the pages on which the authors’ Carla Guanais Branchini (301), Department of Chemi-
contrbutions begin. cal Science and Technologies, University of Rome “Tor
A.H. Abdullah (103), Centre of Excellence for Advanced Vergata,” Rome, Italy
Sensor Technology (CEASTech), Universiti Malaysia Jesús Brezmes (49), SIPOMICS, Department of Elec-
Perlis, Perlis, Malaysia tronic Engineering, Universitat Rovira i Virgili, Avda.,
Priscila Alessio (171), Faculty of Science and Technology, Tarragona, Spain
Sao Paulo State University (UNESP), São Paulo, Brazil Susanna Buratti (291), Department of Food, Environmen-
Jayendra Amamcharla (59), Animal Sciences & Industry, tal and Nutritional Science, University of Milan, Milan,
Kansas State University, Manhattan, KS, United States Italy
of America Xavier Cetó (235), Sensors and Biosensors Group, Depart-
Francesca Antonucci (73), Consiglio per la ricerca in ment of Chemistry, Autonomous University of B arcelona,
agricoltura e l’analisi dell’economia agraria (CREA), Barcelona, Spain; Mawson Institute, University of South
Agricultural engineering research unit, Rome, Italy Australia, Mawson Lakes, SA, Australia
Constantin Apetrei (277), Department of Chemistry, Patrycja Ciosek (209), Department of Microbioanalytics,
Physics and Environment, Faculty of Sciences and Envi- Faculty of Chemistry, Warsaw University of Technol-
ronment, “Dunarea de Jos” University of Galati, Galati, ogy, Warsaw, Poland
Romania Salvatore Claps (73), Consiglio per la ricerca in agri-
Sundar Balasubramanian (59), PepsiCo America Bever- coltura e l’analisi dell’economia agraria (CREA),
ages R&D, Barrington, IL, United States of America Research Unit for the Extensive Animal Husbandry,
Bella, Italy
Ulrich Banach (115), Bundesanstalt für Materialforschung
und –prüfung (BAM), Berlin, Germany Carlos José Leopoldo Constantino (171), Faculty of
Science and Technology, Sao Paulo State University
Rajib Bandyopadhyay (125, 245), Department of Instru- (UNESP), São Paulo, Brazil
mentation and Electronics Engineering, Jadavpur Uni-
versity, Kolkata, India Maria Stella Cosio (83), Department of Food, Environ-
mental and Nutritional Sciences, University of Milan,
Runu Banerjee(Roy) (125), Department of Instrumenta- Milano, Italy
tion and Electronics Engineering, Jadavpur University,
Kolkata, India Corrado Costa (73), Consiglio per la ricerca in agricoltura
e l’analisi dell’economia agraria (CREA), Agricultural
José M. Barat (199), Departamento de Tecnología de los engineering research unit, Rome, Italy
Alimentos, Universitat Politècnica de València, Spain
Cristiane Margarete Daikuzono (171), São Carlos Institute
Nadia Bazihizina (73), Production Sciences Department of of Physics, University of São Paulo, São Paulo, Brazil
Agri-Food and the Environment, University of Florence,
Sesto Fiorentino, Italy Osvaldo Novais de Oliveira, Jr (171), São Carlos Insti-
tute of Physics, University of São Paulo, São Paulo,
Simona Benedetti (291), Department of Food, Environ- Brazil
mental and Nutritional Science, University of Milan,
Milan, Italy José A. De Saja (265), Group of Sensors, Industrial Engi-
neering School, University of Valladolid, Spain
Nabarun Bhattacharyya (125, 245), Centre for Develop-
ment of Advanced Computing, Kolkata, India Manel del Valle (235), Sensors and Biosensors Group,
Department of Chemistry, Autonomous University of
Evandro Bona (31), Federal University of Technology— Barcelona, Barcelona, Spain
Paraná (UTFPR), Campo Mourão, Paraná, Brazil
xi
xii Contributors
Corrado Di Natale (301), Department of Electronic Eduard Llobet (49), MINOS-EMaS, Department of Elec-
Engineering, University of Rome “Tor Vergata,” Rome, tronic Engineering, Universitat Rovira i Virgili, Avda,
Italy Tarragona, Spain
Luís G. Dias (179), Agrarian School, Polytechnic Institute of Jesús Lozano (137), GRIDSEN, Institute of Physics
Bragança, Campus Santa Apolónia, Bragança; CQ-VR, and Technology Information, Superior Council for
Chemical Center–Vila Real, University of Trás-os- Scientific Research, Madrid; University of Extremad-
Montes e Alto Douro, Vila Real, Portugal ura, School of Industrial Engineering, Badajoz, Spain
Rui Sérgio dos Santos Ferreira da Silva (31), State Uni- Larisa Lvova (151, 301), Department of Chemical Science
versity of Londrina (UEL), Londrina, Paraná, Brazil and Technologies, University of Rome Tor Vergata,
Mats Eriksson (255), Department of Physics, Chemistry Rome, Italy; Institute of Earth Sciences, Saint-Petersburg
and Biology, Linköping University, Linköping, Sweden State University, Saint-Petersburg, Russia
C. García-Hernández (265), Group of Sensors, Industrial Stefano Mancuso (73), Production Sciences Department
Engineering School, University of Valladolid, Spain of Agri-Food and the Environment, University of Flor-
ence, Sesto Fiorentino, Italy
Eduardo Garcia-Breijo (199), Centro de Reconocimiento
Molecular y Desarrollo Tecnológico (IDM), Universitat Ramón Martínez-Máñez (199), Centro de R econocimiento
Politècnica de València, Valencia, Spain Molecular y Desarrollo Tecnológico (IDM), Universitat
Politècnica de València, Valencia; CIBER de Bioing-
Mahdi Ghasemi-Varnamkhasti (277), Department of eniería, Biomateriales y Nanomedicina (CIBER-BBN),
Mechanical Engineering of Biosystems, Shahrekord
Spain
University, Shahrekord, Iran
M.J. Masnan (103), Institute of Engineering Mathematics,
Arunangshu Ghosh (245), Department of Instrumenta- Pauh Putra Main Campus, Universiti Malaysia Perlis,
tion and Electronics Engineering, Jadavpur University, Perlis, Malaysia
Kolkata, India
C. Medina-Plaza (265), Group of Sensors, Industrial Engi-
Luis Gil-Sánchez (199), Centro de Reconocimiento Mo- neering School, University of Valladolid, Spain
lecular y Desarrollo Tecnológico (IDM), Universitat
Politècnica de València, Valencia, Spain Paolo Menesatti (73), Consiglio per la ricerca in agricol-
tura e l’analisi dell’economia agraria (CREA), Agricul-
M.T.S.R. Gomes (21), CESAM & Department of Chemis- tural engineering research unit, Rome, Italy
try, University of Aveiro, Aveiro, Portugal
Irina Mirela Apetrei (277), Department of Pharmaceutical
Manuel Gutiérrez-Capitán (189), Instituto de Micro- Sciences, Faculty of Medicine and Pharmacy, “Dunarea
electrónica de Barcelona (IMB-CNM), CSIC, Campus de Jos” University of Galati, Romania
UAB, Bellaterra, Spain
Giuseppe Morone (73), Consiglio per la ricerca in
Thomas Hübert (115), Bundesanstalt für Materialforsc- agricoltura e l’analisi dell’economia agraria (CREA),
hung und –prüfung (BAM), Berlin, Germany Research Unit for the Extensive Animal Husbandry,
M. Carmen Horrillo (137), GRIDSEN, Institute of Phys- Bella, Italy
ics and Technology Information, Superior Council for O. Omar (103), Malaysian Agricultural Research and
Scientific Research, Madrid, Spain Development Institute (MARDI), Penang, Malaysia
Guohua Hui (15), School of Information Engineering, Federico Pallottino (73), Consiglio per la ricerca in agri-
Key Laboratory of Forestry Intelligent Monitoring and coltura e l’analisi dell’economia agraria (CREA), Agri-
Information Technology of Zhejiang Province, Zhejiang cultural engineering research unit, Rome, Italy
A & F University, Linan, China
Roberto Paolesse (301), Department of Chemical Science
Cecilia Jiménez-Jorquera (189), Instituto de Microelec- and Technologies, University of Rome “Tor Vergata,”
trónica de Barcelona (IMB-CNM), CSIC, Campus UAB, Rome, Italy
Bellaterra, Spain
António M. Peres (179), Laboratory of Separation and Re-
Dmitry Kirsanov (225), Chemistry Department, St. Peters- action Engineering–Associate Laboratory, LSRE-LCM,
burg State University, St. Petersburg, Russia Agrarian School, Polytechnic Institute of Bragança,
Aleksander Kubiak (93), Department of Food Science and Campus Santa Apolónia, Bragança, Portugal
Department of Process Engineering and Equipment, Uni- Krishna Persaud (3), School of Chemical Engineering
versity of Warmia and Mazury in Olsztyn, Olsztyn, Poland and Analytical Science, The University of Manchester,
Andrey Legin (225), Chemistry Department, St. Peters- Manchester, United Kingdom
burg State University, St. Petersburg, Russia
Contributors xiii
N.A. Rahim (103), Centre of Excellence for Advanced Jin-E Shin (59), PepsiCo America Beverages R&D,
Sensor Technology (CEASTech), Universiti Malaysia Barrington, IL, United States of America
Perlis, Perlis, Malaysia Maria Concetta Strano (73), Consiglio per la ricerca in
Antonio Riul, Jr (171), Institute of Physics “Gleb Wataghin,” agricoltura e l’analisi dell’economia agraria (CREA),
University of Campinas (UNICAMP), SP, Brazil Research center for citrus and Mediterranean crops,
María Luz Rodríguez Méndez (265), Group of Sensors, Acireale, Italy
Industrial Engineering School, University of Valladolid, Yusuke Tahara (161), Graduate School of Information
Spain Science and Electrical Engineering, Kyushu University,
Lígia R. Rodrigues (179), CEB—Centre of Biological Fukuoka, Japan
Engineering, University of Minho, Campus de Gualtar, Cosimo Taiti (73), Production Sciences Department of
Braga, Portugal Agri-Food and the Environment, University of Florence,
María Teresa Rodriguez-Estrada (39), Interdepartmen- Sesto Fiorentino, Italy
tal Centre for Agri-food Industrial Research, Univer- Pradip Tamuly (245), Tea Research Association, Jorhat,
sity of Bologna; Department of Agricultural and Food Assam, India
Sciences, University of Bologna, Bologna, Italy Carlo Tiebe (115), Bundesanstalt für Materialforschung
Santina Romani (39), Interdepartmental Centre for Agri- und –prüfung (BAM), Berlin, Germany
food Industrial Research, University of Bologna; Depart- Kiyoshi Toko (161), Graduate School of Information Sci-
ment of Agricultural and Food Sciences, University of ence and Electrical Engineering, Kyushu University,
Bologna, Bologna, Italy Fukuoka; Research and Development Center for Taste
Andrea Romano (83), Free University of Bolzano, Faculty and Odor Sensing, Kyushu University, F ukuoka,
of Science and Technology, Bolzano, Italy Japan
Alisa Rudnitskaya (225), CESAM and Chemistry Depart- Bipan Tudu (125, 245), Department of Instrumentation and
ment, Aveiro University, Aveiro, Portugal Electronics Engineering, Jadavpur University, K
olkata,
F.S.A. Sa’ad (103), Centre of Excellence for Advanced India
Sensor Technology (CEASTech), Universiti Malaysia Mikhail Yu. Vagin (255), Department of Physics, Chem-
Perlis, Perlis, Malaysia istry and Biology, Linköping University, Linköping,
José Pedro Santos (137), GRIDSEN, Institute of Physics Sweden
and Technology Information, Superior Council for Sci- Ana C.A. Veloso (179), Polytechnic Institute of Coimbra,
entific Research, Madrid, Spain ISEC, DEQB, Coimbra; CEB—Centre of Biological
Matteo Scampicchio (83), Free University of Bolzano, Engineering, University of Minho, Campus de Gualtar,
Faculty of Science and Technology, Bolzano, Italy Braga, Portugal
Lucia Sepe (73), Consiglio per la ricerca in agricoltura e Fredrik Winquist (255), Department of Physics, Chem-
l’analisi dell’economia agraria (CREA), Research Unit istry and Biology, Linköping University, Linköping,
for the Extensive Animal Husbandry, Bella, Italy Sweden
A.Y.M. Shakaff (103), Centre of Excellence for Advanced A. Zakaria (103), Centre of Excellence for Advanced
Sensor Technology (CEASTech), Universiti Malaysia Sensor Technology (CEASTech), Universiti Malaysia
Perlis, Perlis, Malaysia Perlis, Perlis, Malaysia
Preface
Electronic noses and tongues are the products of advanced chemical and physical sciences combined with the intuitive in-
tegration of sensors, microprocessors, advanced informatics, and statistics. They include resistive, optical, electrochemical,
or piezoelectrical platforms, where a variety of sensing materials (including, among many others, metal oxide semiconduc-
tors, conducting polymers, nanoparticles, phthalocyanines, or enzymes) have been immobilized using numerous different
techniques.
Some of these devices are available commercially, whereas others are home-grown prototype devices that require
commercialization. Electronic noses and tongues have been used to characterize components that contribute to sensory
or compositional profiles, from ripening to harvesting and from storage of raw materials to packaging and consumption.
Electronic noses and tongues are thus suitable for high-throughput analysis and quality control or for determining the
nature and extent of spoilage and adulteration. These devices have also been used to ascertain the geographical origins of
food and mixtures. Devices used to analyze one particular food item can theoretically be adapted for other food items or
components. This does not just mean redeploying the sensing devices but also the mode of statistical analysis. This includes
supervised and unsupervised tools such as principal component analysis (PCA), linear discriminant analysis (LDA), partial
least squares (PLS), and artificial neural networks (ANN). In other words, there is cross-transference of chemistry, phys-
ics, concepts, techniques, findings, and approaches from one food to another. However, finding all this information in a
coherent and comprehensive text has been a problem because, until now, no publication has attempted to marshal together
all the relevant information on these important devices in relation to food science. This is addressed in “Electronic Noses
and Tongues in Food Science.”
Its unique feature is the three parts dedicated to the electronic nose, the electronic tongue, and the combined systems
of electronic nose and tongue. Part I covers a description of electronic nose systems and their applications to the analysis
of the volatile composition of different foods and beverages. Part II focuses on the electronic tongue, which has become
increasingly important over recent years because it can analyze complex liquids, such as wines or milk, by direct immersion
in the samples and not restricted to the headspace. Part III covers newer developments combining both the electronic nose
and tongue. Each part presents the main applications in the food industry. Not only classical applications in the fields of
meat, wine, dairy products, or beers are presented but also other lesser-known applications, such as the detection of gliadins
or the assessment of the phenolic content in foods.
This book is designed for food scientists, technologists, and food-industry workers, as well as research scientists. Con-
tributions are from leading national and international experts, including those from world-renowned institutions. Readers
can dip into the book for reference purposes, read any chapter as a standalone treatise, or read it from cover to cover if food
analysis is an integral part of their day-to-day job.
I must conclude by thanking all those who have contributed to this book, each a recognized expert in their field. I also
wish to thank Elsevier Publishing for all the guidance on pulling together such an eclectic book. Finally, many thanks to my
colleagues and PhD students for their help and support!
xv
Chapter 1
the perception of flavor consists of the sensory combina- As such, it attempts to isolate the sensory properties of the
tion and integration of odors, tastes, oral irritation, thermal foods themselves and provides important and useful infor-
sensations, and mouth feel that originate from a particular mation to product developers and food scientists about the
food (Breslin, 2001; Breslin and Spector, 2008). When we sensory characteristics of their products. Sensory evalua-
describe the aroma of foods, we often use the term “taste” tion has been defined as a scientific method used to evoke,
to indicate sensations that usually are quite complex and in- measure, analyze, and interpret those responses to products
clude to a large extent smell sensations. When we introduce as perceived through the senses of sight, smell, touch, taste,
food into our mouth, taste receptors located on the surface and hearing.
of our tongue are stimulated and send signals to the brain.
However, at the same time, especially as a result of mastica-
tion, many volatile components are also released, reaching 1.1.1 Discrimination Tests
the olfactory mucosa through an opening situated on the up- Discrimination tests are used when the sensory specialist
per wall of the palate. This combination of sensations is what wants to determine whether two samples are perceptibil-
makes up “flavor.” The general categorization of flavor can ity different. It is possible for two samples to be chemi-
be the broad sensations perceived when consuming different cally different in formulation, but for human beings not to
foods. From a biological point of view the flavor sensation perceive this difference. Product developers exploit this
can be defined as the perception originated after a substance possibility when they reformulate a product by using dif-
has been taken into the mouth and it is a specific characteristic ferent ingredients while simultaneously not wanting the
of the substance being perceived (Labows and Cagan, 1993; consumer to detect a difference. If the difference between
Lawless, 1991). The multisensory integration that makes up the samples is very large and thus obvious, discrimina-
our perception of flavor is extremely complex and continues tion tests are not useful. Therefore, discrimination testing
to be a fascinating field of study for psychologists and biolo- is most useful when the differences between the samples
gists (Auvray and Spence, 2008; Spence, 2015). are slight. Discrimination tests are usually performed when
The two main sense organs by which flavor is perceived there are only two samples. There are a number of different
are the nose and the mouth. Flavor comes from three differ- tests available, including triangle tests, n-alternative forced
ent sensations: taste, trigeminal (those sensations perceived choice tests, tetrad tests, and polygonal and polyhedral tests
by a human as astringency, pungency, and cooling) sensa- (O’Mahony, 2013).
tions, and aroma. Trigeminal and taste sensations are real-
ized with the presence of food in the mouth. The aroma
sensation occurs when the molecules of the aromatic com- 1.1.2 Descriptive Analysis
pound are detected at the olfactory receptors reaching this Descriptive sensory analysis is a highly sophisticated tech-
site by the nasal or oral passageways. Assessing food quality nique that allows the sensory scientists to obtain complete
is complex, because this translates to assessing appearance, sensory descriptions of products. They help identifying un-
color, organoleptic qualities, and for taste—mouth feel and derlying ingredient and process variables, and/or determine
so forth that results in determining “liking” or “disliking.” which sensory attributes are important to acceptance. Usu-
This assessment is subjective, requiring human noses ally, descriptive techniques produce objective descriptions
and tongues; typically a trained human panel is required. of products in terms of the perceived sensory attributes. De-
The many different possible flavors are due to interactions pending on the specific technique used, the description can
of chemical compounds with taste, trigeminal sensations, be more or less objective, as well as qualitative or quantita-
or aroma receptors. The characteristic taste of a food is nor- tive. Most descriptive methods can be used to define sen-
mally related to a single class of compounds. But, an odor sory instrumental relationships.
is usually elicited by a combination of volatile compounds,
each of which imparts its own smells. Differences in char-
1.1.3 Flavor Profile
acteristics of certain aromas can be equated to the varying
proportions of these volatiles. However, some substances Flavor profiling is a consensus technique. The vocabulary
contain trace amounts of a few volatile compounds that pos- used to describe the product and the product evaluation itself
sess the characteristic essence of the odor. These are called is achieved by reaching agreement among the panel mem-
character-impact compounds. One must also realize that the bers. The flavor profile considers the overall flavor and the
chemicals of a single compound class can induce many di- flavor noted and estimates intensity and amplitude (overall
verse flavors, especially as their concentrations vary. impression) of these descriptors. The technique provides a
Sensory evaluation comprises a set of techniques for tabulation of the perceived flavors, their intensities, their
accurate measurement of human responses to foods and order of perception, their aftertastes, and their amplitude.
minimizes the potentially biasing effects of brand identity If the panelists are trained appropriately, this tabulation is
and other information influences on consumer perception. reproducible. Using standardized techniques of preparation,
Electronic Noses and Tongues in the Food Industry Chapter | 1 5
presentation, and evaluation, a panel consisting of four to combinatorial approach utilizing arrays of broad specificity
six judges are trained to precisely define flavors of a product sensors outlined by Persaud and Dodd (1982) have formed
within a specific food category. The food samples are tast- the basis of many types of “electronic noses” and “electron-
ed and all perceived notes are recorded for aroma, flavor, ic tongues” developed since then. The concepts are based
mouthfeel, and aftertaste. After this exposure, the panelists broadly on our understanding of biological odor and taste
review and refine the descriptors used. transduction mechanisms where large numbers of different
At the completion of the training phase, the panelists odor receptor types or taste receptors respond to the same
have defined a frame of reference for expressing the inten- chemical stimulus but with different affinities. The result
sities of the descriptors used. The samples are then served is a complex pattern of neural responses that the brain as-
to the panelists in the same form that they would be served sociates with previously learned stimuli in order to identify
to the consumer and the intensities of the perceived flavor or discriminate between these chemical stimuli. In order to
notes are rated typically on a defined scale. A consensus understand these concepts, it is worth spending some time
of scores is obtained by discussion and reevaluation of the understanding some of these biological mechanisms.
products by the panelists.
It can be seen that sensory evaluation of foods is very 1.2.1 Biology of Smell and Taste
complex and subject to high variability, and instrumental
means of carrying out some of these evaluations would be The nose is capable of detecting a large repertoire of mol-
most useful to the food industry. The remarkable capabili- ecules—the latest estimates indicate that these may num-
ties of the nose and tongue in detection, recognition, and ber over 1 trillion odors (Bushdid et al., 2014). It resembles
discrimination of complex mixtures of chemicals, together somewhat the concepts inherent in the immune system where
with rapid advances in understanding how these systems it is not possible predict what molecules may be encountered
operate has stimulated the imagination and interest of many at any time. These molecular families include aliphatic and
researchers and commercial organizations for the develop- aromatic molecules with varied carbon backbones and di-
ment of electronic analogues of the biological systems. This verse functional groups, including aldehydes, esters, ketones,
chapter focuses on “electronic noses and tongues” as ap- alcohols, alkenes, carboxylic acids, amines, imines, thiols,
plied to the food industry. The responses of individual odor halides, nitriles, sulfides, and ethers. With rapid advances
sensors or taste sensors combined into an array, by which in molecular biology, we now have a good understanding of
each sensor possesses slightly different response selectivity the transduction processes in the nose and signal-processing
and sensitivity toward the sample odors or tastants, when pathways in the brain that are responsible for detection, dis-
combined by suitable mathematical methods, can provide crimination, and recognition of odors (Firestein, 2001).
information to discriminate between many sample odors About 6 to 10 million olfactory sensory neurons (OSN)
or tastants and when combined together—flavors. Arrays are found in the nasal cavity of mammals. Each of these cells
of gas and odor sensors, made using different technologies is a bipolar neuron where the apical end ends in a knob from
have become known as “electronic noses” and consist of which protrude a number of cilia that lie in a thin layer of
three elements: a sensor array which is exposed to the vola- mucus on the surface of the olfactory epithelium (Fig. 1.1).
tiles; conversion of the sensor signals to a readable format; These cilia contain the sensory transduction mechanisms
and software analysis of the data to produce characteristic constituting receptor proteins and an enzymatic cascade
outputs related to the odor encountered. The output from that transforms the energy of an odorant molecule binding
the sensor array may be interpreted via a variety of meth- to the receptor into a neural signal. There is great degree of
ods—principal component analysis, discriminant function similarity between receptors, but there is a region of hyper-
analysis, cluster analysis, and artificial neural networks—to variability that is associated with the ligand-binding pock-
enable discrimination between samples. A similar approach ets of these proteins. This accounts for the large and diverse
is adopted with “electronic tongues,” by which the sensors range of molecules that can be detected by these receptors
in this case detect chemical species in solution. (Firestein, 2001; Mombaerts, 1996, 1999).
The opening of ion channels in the cell membranes when
an odorant molecule interacts with the receptors causes a
1.2 BIOMIMETIC SYSTEMS change in electrical potential in the cell membrane (Chiu
Attempts to mimic the “chemical senses” using artificial et al., 1997), leading to generation of action potentials that
systems are still in the process of development and have are propagated to the olfactory bulb, a structure in the fore-
been evolving over the last few decades. One of the pio- brain devoted to the processing of olfactory signals. De-
neers in thinking about the concepts, Dravnieks (1968) spite the huge number of olfactory neurons, there are only
envisaged an instrument that would inspect samples of 300–400 different types of receptors in humans. Olfactory
odorous air and report the intensity and quality of an odor neurons expressing a particular receptor type converge to
without the intervention of a human nose. The ideas of a distinct regions in the olfactory bulb called glomeruli. The
6 PART | I The Electronic Nose
FIGURE 1.1 The olfactory epithelium consists of OSNs, which are bipolar neurons, sustentacular cells that are glial cells with microvilli, and
basal cells that are stem cells, from which new OSNs are generated. Each OSN expresses only one of the ∼1000 olfactory receptor genes and the
axons from all cells expressing that particular receptor converge onto one or a few glomeruli in the olfactory bulb. The glomeruli contain the incoming
axons of OSNs and the apical dendrites of mitral cells that are the output layer of cells. Mitral axons leaving the OB project to higher brain structures
including the piriform cortex, hippocampus, and amygdala.
outputs from this region are specialized cells called mitral uncovered (Kinnamon, 2012; Simon et al., 2008; Teeter and
(M/T) cells that propagate a highly reduced number of sig- Brand, 1987). On the tongue are found structures called pa-
nals to higher parts of the brain—many thousand olfactory pillae that contain taste buds (Fig. 1.2). These papillae have
neurons converge onto the dendrites of between 5 and 25 different shapes: fungiform found mainly on the anterior of
mitral cells in each glomerulus. Each ORN type projects into the tongue, foliate along the sides, and circumvallate at the
one or two glomeruli and a single M/T cell receives input back of the tongue. Within each taste bud are found approx-
from just one type of ORN, expressing the same type of re- imately 50–100 cells that contain the transduction elements
ceptor, and sends its axon to the olfactory cortex. Therefore, associated with taste. Taste receptor cells are not neurons,
the first representation of the odor is transduced at the glo- but interrelate with projection neurons that make up the
merular level, to produce a second spatial and very ordered taste nerves. In mammals, the taste buds, depending upon
pattern, representing its molecular features and distributed their location, are innervated by one of several paired crani-
among a much smaller number of output cells (chemotopic al nerves: the chorda tympani (anterior tongue) and greater
coding) (Mombaerts, 1999, 2006; Mori et al., 2006; Mori superficial petrosal (soft palate) branches of the facial nerve
and Sakano, 2011). A combinatorial strategy is adopted by (CN VII), the glossopharyngeal (CN IX) nerve (posterior
the biological system (Gupta et al., 2015; Korsching, 2001; tongue), and the vagus (CN X) nerve (root of tongue and
Manzini et al., 2014). Most odor molecules are recognized esophagus). The maxillary and mandibular branches of the
by more than one receptor and most receptors recognize trigeminal nerve (CN V) mediate touch, pain, temperature,
several odors that may be related by some properties such and other somatosensory sensations throughout the nasal
as the size, shape, and charges associated with functional and oral epithelia—that is, coolness and fizziness. The
groups on the molecule. A particular odor molecule may taste fibers integrate information from taste buds and relay
also consist of a number of these “epitopes” or “determi- this information to the nucleus of the solitary tract of the
nants” that possess some of these features. Thus the recog- brainstem. The primary taste cortex contains neurons that
nition of an odorant molecule depends on which receptors respond best to sweet, sour, bitter, salty, and umami-tasting
are activated and to what extent. agents, as well as neurons responsive to touch and smell.
While taste sensations have been described for many It has direct reciprocal connections with the frontal tempo-
centuries, it is only comparatively recently that the fun- ral, parietal, entorhinal, and orbitofrontal cortexes and other
damental mechanisms underlying taste transduction were cortexes. The entorhinal and orbitofrontal cortex functions
Electronic Noses and Tongues in the Food Industry Chapter | 1 7
FIGURE 1.2 The tongue showing fungiform, foliate, and circumvallate papillae. These contain taste buds containing a number of sensory cells;
membrane-bound receptors—sweet taste and umami are transduced by heterodimers of T1R2 and T1R3 subunits, and T1R1 and T1R3 subunits, respec-
tively, while bitter taste is transduced by T2R receptors (not shown). Sour taste involves activation of a type of transient receptor potential channel (TRP),
while salt taste is transduced via epithelial sodium ion channels in the cell membrane. The signals from taste buds project to the solitary nucleus of the
brainstem, and then is processed by higher regions of the brain.
to interpret tastes and includes memory and cognition. The affinities, so exhibit broad selectivity in terms of ligands
insular cortex is associated with conditioned taste aversions. that can be detected.
The cell membranes of the taste bud cells contain a va-
riety of ion channels that are involved with signal transduc-
tion. Salt taste is transduced via an epithelial sodium ion 1.3 ELECTRONIC NOSES AND TONGUES
channel (ENaC) that is ubiquitously expressed and function-
1.3.1 Electronic Noses
al in the anterior part of the tongue. Sour is the taste of acid,
that is, protons (H+). This taste is complex and is transduced So-called “electronic noses” comprise a vapor sampling
through transient receptor potential (TRP) ion channels, but system, an array of chemical sensors, and a method of sig-
there may at least three possible receptor mechanisms: H+ nal processing that leads to classification of the responses
blocks K+ ion channels, H+ ions go through ENaC channels, of the chemical sensors. The sensors in an electronic nose
or H+ ions go through a proton channel. are desired to have a broad selectivity rather than being
Sweet and umami receptors are heteromeric receptors specific to one type of volatile chemical. The human nose
made up of a combination of different subunits, coded for by can identify many odors that may contain hundreds of indi-
a small gene family—T1R. Sweetness receptors are a com- vidual chemical components and, therefore, the sensors for
bination of two types of receptor subunits (T1R2 + T1R3) an electronic nose should be generalized at the molecular
and umami receptors a combination of another set of recep- level. The desired properties for sensors are high sensitiv-
tor subunits (T1R1 + T1R3) (Doty, 2012). The bitter re- ity, rapid response, good reproducibility, and reversibility
ceptors consist of taste-2 receptors (T2Rs). Fifty to eighty to large numbers of chemicals. It is also better for an elec-
members are expressed in small subset of all taste papillae. tronic nose to be small in size and flexible, able to adapt to
T2Rs are membrane bound proteins that are bitterness re- being exposed to many types of environments, and operate
ceptors (Meyerhof et al., 2011). We still poorly understand at ambient temperatures. The working principle of these in-
taste, and other types of receptors that are under investiga- struments is based on the employment of an array of differ-
tion include those involved in the taste of fats (DiPatriz- ent nonselective chemical sensors. Each sensor in the array
io, 2014; Passilly-Degrace et al., 2014). It is clear however is able to detect a range of different odors, not just one,
that like olfactory receptors, taste receptors are capable of with a sensitivity that is different from the sensitivity shown
interacting with many types of substances with different by the other elements of the array. Such features make the
8 PART | I The Electronic Nose
FIGURE 1.3 Electronic nose concepts: sampling, sensor array, signal processing, pattern recognition.
response of the electronic nose dependent on the whole of different sensors chosen to respond to a wide range of
range of chemical information contained in an odor. This chemical classes and discriminate mixtures of volatile com-
is something similar to what occurs in biological olfaction. pounds that make up an odor. The outputs from individual
Usually, gas sensors are categorized based on their trans- sensors are processed in such a way in order to produce
duction principle and sensing material. Many different types a distinct response pattern reflecting the relative responses
of sensor technologies have been so far deployed (Gutierrez of all the sensors in the array to a given chemical stimu-
and Horrillo, 2014; Vergara and Llobet, 2011; Wilson and lus. Identification and classification of an analyte mixture
Baietto, 2009)—these include metal oxide sensors, conduct- is accomplished through comparison of this electronic fin-
ing polymers, optical sensors, piezoelectric sensors, electro- gerprint of sensor responses against previously learned pat-
chemical sensors, field effect transistors (Feng et al., 2014), terns (Fig. 1.3). Unlike conventional analytical instruments,
and more recently olfactory biosensors containing either ol- there is no need for separation of a complex mixture into
factory receptors (Ault and Broach, 2006; Du et al., 2014) individual components and this holistic approach resem-
or odorant-binding proteins (Di Pietrantonio et al., 2013; bles the process found in biological olfaction. Importantly,
Pelosi et al., 2014; Persaud, 2012) have been utilized. The these devices give information that allow comparison or
interaction with an analyte induces a physical and/or chemi- discrimination of odors without necessarily having to do a
cal change in the chemical sensing layer, which produces a chemical analysis of individual components in the mixture.
signal. The nature of the transduction processes utilized by There has also been a blurring of instrument technologies
these sensors may include electrical measurements, includ- where “electronic nose” concepts are combined with tradi-
ing changes in current, voltage, resistance or impedance, tional analytical instruments; for example, a chromatogram
electrical fields, and oscillation frequency. Other transduc- may be treated as a set of virtual sensors, or direct injection
ers involve measurements of mass changes, temperature of a mixture of substances into a mass spectrometer with-
changes or heat generation. Optical sensors measure the out prior separation will produce a complex spectrum that
modulation of light properties or characteristics such as can be processed as a “fingerprint” to describe that mix-
changes in light absorbance, polarization, fluorescence, ture (Crespo et al., 2012; Moon et al., 2014). Taking this
and other optical properties. These have been extensively to the opposite extreme, modulation of a single sensor can
reviewed in other publications (Di Natale et al., 2005, 2009; produce a virtual sensor array—for example, modulating
Walt and Sternfeld, 2006; Walt, 2010) and will not be de- the operational temperature of a single metal oxide sensor
scribed in detail here. The choice of sensor is often appli- can process a complex response that can be deconvoluted
cation dependent and are based on response and recovery (Amini et al., 2013; Herrero-Carron et al., 2015). On the
times, sensitivities, detection range, operating limitations, other hand, large arrays of sensors consisting of up to 4096
physical size, robustness to being poisoned, power con- sensors have been described and these start to resemble
sumption, and other factors. However, very few types of more realistically the olfactory receptor system (Beccher-
chemical sensor available respond specifically to a single elli et al., 2010; Bernabei et al., 2012).
chemical—they tend to show cross-sensitivity to a variety While a number of companies were formed to exploit
of chemicals, but the range of selectivity can be tailored “electronic nose” technologies over the last few decades,
so that each type of sensor responds to a different range of the majority of these have failed. One reason is “oversell-
chemicals. This inherently poor selectivity can be utilized ing” of what such devices can actually do. In fact they are
to advantage when sensors of different type are combined not analytical instruments and neither are they even close
into an array. The cross-reactive sensor array is composed to the functionality of the biological nose (Boeker, 2014).
Electronic Noses and Tongues in the Food Industry Chapter | 1 9
Another problem has been that of calibration of such instru- consistent results from such devices—especially when mea-
ments, and that of “drift” over time. The latter is associated suring a complex matrix such as food—much care needs to
with limitations of the sensor technologies utilized—chang- be taken and in fact one complaint often heard is the amount
es in responses over time due to aging, poisoning, or envi- of work required for method development for a particular
ronmental factors, together with changes in the chemical application. The techniques that are used are those that have
analytes that are being assessed. For example, a typical food been well developed over the years for gas chromatogra-
product will change organoleptic characteristics over time phy and have been adapted for “electronic nose” detectors
due to aging, decomposition, or other factors, and different (Feng et al., 2014; Pillonel et al., 2002).
batches of a product may have large differences in odor or The simplest technique consists of placing the sample
flavor. Nonetheless, there have been real successes in the into a vial that is then closed and allowing equilibrium to be
use of these technologies in the food industry that shall be reached between the matrix and the vapor phase. This “stat-
described later. ic headspace” of the vial then constitutes the sample that is
The performance of today’s artificial systems is far from introduced into the “electronic nose” system. The sample
that found in biological olfaction. The better performance of temperature, equilibration time, vial size, and sample quan-
the latter is due to the great number of receptor neurons and tity are parameters that have to be optimized for a particular
to the unique architecture of the olfactory pathway, where application and often this can be automated.
three main elements, the olfactory epithelium, the olfactory Instead of a generating a static headspace, the sample
bulb and the olfactory cortex, are completely integrated. In may be placed into a vial and in this case the vial is purged
the artificial olfactory devices, similarly to the brain, the with a continuously flowing gas stream. This in effect strips
signals coming from the sensors are processed in order to volatile components from the matrix of the sample as there
classify odors. However, this process is carried out using is a displacement of the equilibrium between the sample
mathematical tools. An initial sensor signal preprocessing matrix and the headspace. This “dynamic headspace” has
is often performed. Different preprocessing metrics can be been used to good effect in many applications. The flow
used with different aims. For example, the time dependence rate of the gas and the sample temperature needs to be opti-
of the signal can be removed in order to reduce the amount mized for a particular application.
of data that needs to be handled during the phase of pat- Often the concentrations of volatile components from
tern recognition. Linearization of the sensor response can a sample are very low, and the dynamic headspace method
be performed when the intensity of the odors is important, may be combined with a preconcentration method such as
as the sensor response is often nonlinear with increasing “purge and trap” where the vapor is trapped on an adsor-
concentration of analyte. Data reduction techniques, such as bent such as activated carbon, polymers, or inorganic ma-
principal component analysis (PCA), imitating the massive trices. After sampling the dynamic headspace over a period
convergence of the ORNs to the glomeruli, can also be ap- of time, the trap is heated to a high temperature in order to
plied to the whole sensor array response, in order to reduce discharge the trapped volatiles as a bolus into the electronic
the data dimensionality (Pearce, 1997a,b). nose. The flow rate of the gas, the purge rate, and the desorp-
The data resulting from this first processing step, or tion temperature are parameters that need to be adjusted for
coming directly from the sensors without any pretreat- a given application (Chai et al., 2008; Chen et al., 2015; del
ment, are subjected to pattern recognition (PARC) analysis Nogal Sanchez et al., 2014; Feng et al., 2014; Lopez-Feria
to obtain discrimination and/or classification. During this et al., 2008; Pillonel et al., 2002; Plutowska et al., 2011).
phase, the sensor patterns are compared to known odor pat- Solid-phase microextraction (SPME) introduced by
terns, which have been stored in a knowledge base during Pawliszyn et al. (2012) and Pawliszyn (2012) and stir-bar
a previous training stage of the device, and, using any of adsorption (SBE) are convenient methods that are used to
the many classification techniques available, identification preconcentrate a static headspace (Baltussen et al., 2002;
of the stimulus is carried out. Existing sensor arrays imple- David and Sandra, 2007). In the former a silica fiber coated
ment the broad and overlapping sensitivity feature exhib- with an organic polymer is introduced into the headspace
ited by biological receptors, but it has not yet been possible and this sorbs volatile components from the headspace and
to achieve the same dimensionality and redundancy level entraps them. A range of sorbent coatings are available that
available in the olfactory epithelium. Current chemical sen- range from hydrophilic to hydrophobic so some selectivity
sor arrays, either homogeneous or heterogeneous, have few in what is extracted from the headspace is under control of
sensing elements when compared to natural noses. the user and can be adapted to a particular application. SBE
uses a similar principle but contains a much larger adsor-
1.3.2 Sampling bent surface allowing much greater quantities of volatiles
to be entrapped. Just like the purge-and-trap method, the
Apart from sensing aspects associated with “electronic nos- SPME or SBE is heated to a high temperature to desorb
es,” the sampling aspects are often neglected. In order to get volatiles into the electronic nose detector. A variation of
10 PART | I The Electronic Nose
FIGURE 1.4 Solid-phase microextraction of a headspace and desorption process of an SPME fiber on to an array of sensors.
In principle, “electronic tongues” function in a similar where ai and aj are the activity of the primary ion and in-
way to the “electronic nose” (Fig. 1.5). A sensor array terfering ion, respectively; k is the selectivity coefficient;
produces signals that are not necessarily specific for any E0 is the sum of the standard potential of the electrode and
particular chemical species. A pattern of signals is gener- the junction potential; E is the potential difference for the
ated, that can be correlated to certain features or qualities electrochemical cell composed of the ion selective and ref-
of the sample. In this case the sensors operate in an aque- erence electrode; and zi and zj are charge numbers of the
ous environment and they have different cross-sensitivities primary and interfering ion, respectively.
to various chemical species. Just like the “electronic nose,” Because the matrix containing the analytes of interest
an appropriate method of multivariate analysis or pattern may be very complex, the biggest problem encountered with
recognition is used to process the signals from the array in “electronic tongues” is that of electrode fouling. Potentio-
order to discriminate or classify different samples. As with metric measurements are temperature dependent, and the
“electronic noses,” a wide variety of sensor technologies signals may be influenced by solution changes, as well as
may be employed. These include conventional ion selec- adsorption of solution components on membrane surfaces,
tive electrodes, chalcogenide and oxide glasses, noble met- that can affect the nature of the charge transfer. Hence, care
als, organic polymers, biosensors, ISFETS, optical sensors, needs to be taken to control the temperature of the measure-
mass sensors, impedimetric sensors, and others (Baldwin ment, washing of the electrodes with solvents to minimize
et al., 2011; Legin et al., 2004, 2005; Vlasov et al., 2008, the effects of adsorption, using antifouling membranes.
2010; Winquist et al., 2004). The most common sensors However, potentiometric measurements are widely used
used are potentiometric in nature, producing an electri- because of their simplicity. Toko and coworkers (Tahara
cal potential described by the Nernst equation, Eq. (1.1), et al., 2013; Tahara and Toko, 2013; Toko, 2004, 2014;
Electronic Noses and Tongues in the Food Industry Chapter | 1 11
FIGURE 1.5 Electronic tongue concepts: sampling, sensor array, signal processing, pattern recognition.
Toko et al., 2014) developed a taste sensor array based on fundamental statistical distributions of data. Two types of
electrodes using a lipid/polymer membrane for the trans- nonparametric learning or classification methods are avail-
ducer. The composition of the membrane is designed con- able: supervised and unsupervised. Supervised methods in-
sidering the charges on the membrane surface and hydro- volve the learning of data based on advance knowledge of
phobicity on the basis of physicochemical properties of the classification, whereas unsupervised methods make no
substances with each basic taste, and as a result the system prior assumption about the data.
correlates well with basic taste perception. However, it must Preprocessing of data is often necessary before pattern
be realized that these systems in no way resemble the bio- recognition algorithms can be applied. This may involve
logical tongue. Unlike the case with potentiometric sensors, scaling of data, normalization, removal of redundant data,
voltammetric measurements are performed when equilib- extraction of features that are important to classification, and
rium is not reached, and the signal obtained is the current– others (Raman et al., 2011; Roeck et al., 2008). If the data
potential relationship. These methods are particularly use- are high dimensional, for example, when there are a large
ful when trying to obtain measurements from samples that number of sensors in an array, or if virtual sensors are be-
contain redox species. ing utilized from a mass spectrometer or gas chromatograph,
Just like with electronic noses, attention needs to be then principal components analysis is a powerful tool for data
paid to sampling—often static measurements are carried reduction. It involves forming a covariance matrix between
out, where the electrodes are inserted into a medium and an variables, and then carrying out a transformation that results
equilibrium is allowed to be established. On the other hand, in extracting eigenvectors and eigenvalues. This analysis pro-
dynamic flow of analyte medium past the electrodes may duces principal components that are a linear combination of
also give useful information. the original variables. Many algorithms are described in the
literature that are applicable to the classification problems en-
countered using electronic noses and tongues—they include
1.4 PATTERN RECOGNITION independent component analysis (ICA) and linear discrimi-
For both “electronic noses” and “tongues,” pattern recog- nant analysis (LDA), learning vector quantization (LVQ),
nition algorithms play a large role. The notable advantage self-organizing map (SOM), artificial neural networks of
is the ability to characterize complex mixtures without the a variety of architectures (feed forward–back propagation
need to identify and quantify individual components, with- networks, radial basis functions, support vector machines,
out the need of highly selective sensors. There are two main genetic algorithms, and others). All of these methods func-
approaches to pattern recognition: parametric and nonpara- tion well, but are dependent on the stability of the sensor re-
metric. Parametric methods rely upon obtaining or esti- sponses over time, and consistency in sampling. The strategy
mating the probability density function of the parameters is to expose the sensor array to a range of different analytes,
used to characterize the response of a system. Conversely, memorize the resulting patterns in a data base, and use this
nonparametric methods require no assumption about the a priori knowledge to recognize unknown samples later on.
12 PART | I The Electronic Nose
(Continued)
Another random document with
no related content on Scribd:
CHAPTER I.
THE MANSE AND ITS INMATES.
HELEN BARRIE
DIED 18TH MAY 1838, AGED 3 YEARS.
——
WITH CHRIST ... FAR BETTER.
“THE
The spot had no more constant visitor than Bell. The
BUTTERFL flowers that in their seasons grew round it were
Y ON A planted by her hand, and tended by her with
GRAVE.” constant care; the only difference being that in
weeding or trimming it there was not the quick,
bustling energy which she exercised in the garden, but a reverent
slowness unusual for her. She never put her foot on the sod under
which Nellie lay; and although for the first few visits she sighed
mournfully as she read the inscription (and she read it aloud to
herself at every visit), it was not long before her face lightened as
she uttered the last two words, and she would add in a cheerful
confirmatory tone, as if Nellie herself had repeated the epitaph, “Yes,
Nellie; yes, Bell’s bairn, far better; far, far better.”
Mrs. Sigourney.
CHAPTER II.
A QUIET EVENING AT THE MANSE.
Crabbe.
I NEED hardly tell that between Mrs. Barrie and Bell the relationship
of mistress and servant was more than cordial, more than intimate,
—I can find no better word to express it than perfect. To say that
Bell knew her place is a term much too bald; she filled it, fulfilled it,
full-filled it. She was devoted to the family’s interest; her heart and
mind were in her work; she had a clear head, a strong arm, a blithe
happy manner, and an uncommonly large stock of common sense.
She had a ready “knack” of dividing the articles under
her care, by a sliding scale of her own, so as to put all BELL’S
to the best use: she laid aside some for the dining- SLIDING
room on “company” days, and even at a sudden call SCALE.
she was seldom found unprepared; some for the
parlour, to suit old and young (for there was no formal nursery in the
manse,—Bell’s room, “off” the kitchen, was best entitled to the name,
although competing claims might have been put forward by the
kitchen itself, the parlour, and even the study); some for the kitchen,
but that had not a high place in her scale; a good deal for the poor,—
plain, handy, and given in good time and with discernment. Of one
thing she was very careful, and that was, that if any food seemed
likely to spoil, it was given away before it went wrong; if any clothing,
it was given clean, and although often well patched, it was fit for
immediate use. There was a corner in the kitchen pantry with a stock
of comforts, and even luxuries, for cases of sickness, old age, or
special need. The dumb animals were studied with thoughtful care,
and they repaid it well. Everything that could be used was used
regularly and methodically.
Bell’s dress varied with her work. In the morning she
“sorted” the live stock, clad in what an artist would THE BUSY
have called a grotesque or picturesque costume, BEE.
according to the season. In winter her upper garment
was an old overcoat of Mr. Barrie’s—a “Spencer;” in summer it was a
loose-fitting jacket of striped cotton, lilac and white; her linsey-
woolsey petticoat was of the right length for such work, and all were
shaken or brushed or beaten daily. She put on her cotton “morning
wrapper,” of blue with small white spots, just before she “set” the
breakfast, and got “redd up” for the day in time to serve up the
dinner. While she had her set times for her regular work, and “turned
her hand” smartly to anything more pressing, she observed no
“Factory Act” restrictions as to her hours of labour. Very early in the
morning the clank of Bell’s “pattens”[1] was heard as she attended to
her home farm, and till far on in the evening she was working away
anxiously and cheerfully. Her rest was a change of work on week-
days. On the Sabbath afternoon she took what seemed likest a rest,
viz. a walk round the whole premises, leisurely, observant,
inquisitive, noticing everything, and mentally noting a good deal for
next week’s attention; varied by an occasional “saunter” into the
gardens of the neighbours for purposes of observation, comparison,
insight, or exchange.
Her respect for Mr. and Mrs. Barrie was profound: they were the
handsomest couple in the parish, and many parishes might have
been gone over before a more comely, gentle, ladylike, person than
Mrs. Barrie could be met with. Bell said, “They were, if that was
possible, better than they were bonnie;” and when Mrs. Barrie told
Bell, as she often did, to rest and take things more leisurely, Bell
would say, “I like to work, mem, I like it; I canna be idle.” Mrs. Barrie’s
remonstrances were firmer on extra occasions, such as a “heavy
washing,” but Bell’s answer was, “It was naething, naething at a’;
and didna we get a grand day for drying the claes?”—or at the
“Spring cleanin’,” when her answer was, “It’s best to get all the
confusion past and by wi’t. It was a nice thing a fresh, clean hoose;
—’deed, mem, it astonishes me to see hoo much cleanin’ every
place needs, although it’s no very bad like before you begin.”
I may be dwelling too long on Bell, and it is not at all unlikely that she
may become the heroine of my story, or rather the central figure
round which the “bits” are grouped. If so, I could not wish a better,
although Bell herself had no idea that she was such a good servant,
or that she did more than her bare duty; she oftener felt she had not
done as well as she wished. She was far too sensible and busy a
woman to think much about herself; and should she read this, she
would be the first to say “she wished she had done better,—he
hasna tell’d my fau’ts.” Worthy, kindly, honest Bell!
Mrs. Barrie’s housekeeping was the admiration, to
many it was the miracle, of the parish and district. She “GIVEN TO
was a good manager, and with such a helpmate as HOSPITALI
Bell, she made her income do wonders. To the poor, TY.”
the manse was always open for judicious help; the
hospitality of the dining-room and parlour was substantial and
becoming. This was all the more astonishing from the fact that Mrs.
Barrie was “such a delightful creature,” “such a charming person,”
“quite a lady,” “a model minister’s wife,” “so accomplished,” “so
amiable,” “so frank,” “so nice,” “so attractive” (these are actual
epithets used by her friends), that the number of visitors, many of
whom were easily persuaded to become guests, was larger than
was desirable, and the consequent calls on the larder and pantry
were heavy. Indeed, this was a subject of frequent remark among
those who enjoyed the hospitality of the manse, all wondering how
ever she could manage, and many “beseeching” Mrs. Barrie not to
trouble herself about them, as they only wished a quiet chat,
although the length of many of their visits made them more like
visitations; Mrs. A. and Mrs. B. suggesting that Mrs. C. and Mrs. D.
might be more considerate, whilst Mrs. C. and Mrs. D. were
surprised at the audacious manner in which Mrs. A. and Mrs. B.
thrust themselves on Mr. and Mrs. Barrie. It was really difficult to
withstand the attractions of the manse, and Mr. and Mrs. Barrie,
more particularly Mrs. Barrie, was made a social martyr because she
was so good, and kind, and true.
It never occurred to Mrs. Barrie that her good nature and good
housekeeping were inconsiderately drawn upon by many who should
have known better. She liked to see, and to contribute to, the
enjoyment of others, preferred being active to being passive in this
matter, and was “given to hospitality” from the genuine sweetness of
her nature; and while the sigh of weariness often escaped her lips at
the close of some of the nice “sociables,” which had been prolonged
so as to interfere with domestic and other duties, she never
murmured; although she and Bell had often to encroach on the hours
of rest or sleep in order to keep everything forward, and as they
would like it.
Mr. Barrie’s broadcloth was invariably fresh-looking, and his linen
faultless. Mrs. Barrie was at all times becomingly dressed, and in the
afternoons quite “the lady, aye sae genty.” The boys and girls were
comfortably and neatly clad every day, specially so on Sabbath days,
and theirs was a happy home.
Before I began to describe the inmates of the manse, I mentioned
that Mrs. Barrie said, on leaving the parlour, she was going to see if
the “bairns were happit.” She seldom spoke Scotch, but when she
did, it was with quaint emphasis and special sweetness. There was
no real need for Mrs. Barrie having any anxiety on this subject of
“happing,” as Bell was always on the alert; but Mrs. Barrie’s motherly
heart could not rest until she had seen, and kissed in their beds, her
“wee croodlin’ doos.” She went first to see Bell about the supper;
then to Bell’s room, where Mary and Flora were fast asleep; then to
her own room, where Lewis was sleeping soundly, but James wide
awake, scheming in his little head whether he could not make a pair
of skates, and wishing that Bell would come up, as her “pattens”
seemed the likeliest raw material to make them of, and he had seen
an old pair in the byre. Mrs. Barrie heard his story, and said they
would never do; but that Mr. Martin was in the parlour, and she would
ask him the price of a pair, if he would sleep like a good boy; and
kissing both, and “tucking” them in, she returned to the parlour.
During her absence, Mr. Barrie spoke to me in quite a
fatherly way. He knew that I had a good business and BE
fair prospects, but that, since my father’s death, I had “JUDEECI
bought a small property called Knowe Park adjoining OUS.”
the village, and that this had absorbed my available
means to such an extent as to render it a little difficult for me to carry
on business comfortably to the extent that my father had done. After
stating that he thought I was taking a wise step in getting married, he
said he found it generally the case, although it sounded like a
contradiction, that a married house was more cheaply and much
better kept than a bachelor’s; and that he was in the custom of
drawing the attention of folks who were about to get married to the
subject of Life Assurance, or, if working men, to Benefit Societies,
and to the necessity of economy and prudence in money matters.
“But,” added he, “you know these things better than I do, and I know
you will act judeeciously,” with a considerable emphasis on the ee.
And as he referred to the various relationships of social life, he
closed each section (for his advices unconsciously ran into “heads
and particulars,” like his sermons), with, in short, “Be judeecious;”
and so clearly did he illustrate the inseparable connection between
wisdom and success or happiness in everything he spoke of, that his
advice seemed then, and seems yet, summed up in, “Be
judeecious.” He will excuse me for telling here, that in the parish he
was not unfrequently spoken of as “Judeecious;” and after the lapse
of fully forty years, he is still occasionally styled, “Worthy old
Judeecious,” by some elderly warm friends, when recalling the sunny
memories of former days, although in general conversation he is
now spoken of as Dr. Barrie.
He related with considerable glee a saying of an old minister, who, in
speaking of money matters, used to maintain that there were only
three ways in which a minister could make money—patrimony,
matrimony, or parsimony. He also told the story, which is long ago
threadbare, of the old merchant, who, when asked why his son had
not done so well in business as he had, replied, “That’s easily
explained: we old folks began with a little house and a plain table,
with porridge and a herring, and got up to tea and a ‘chuckie’
(chicken); but the young folks began with a braw house, and tea and
chuckies and silks, and never buckled up their sleeves to work.”
When Mrs. Barrie joined us, supper was already on the table. After
glancing into the cradle, to see if all was right with “Gordie,” or
Gordon Lennox, as his full name was, she said, “Come away,
gentlemen,” and seating herself at the head of the table, did the
honours in a graceful and homely way.
Bell had brought in the little black kettle, and it kept
singing by the fireside. When the simple meal was “BAIRNS
over, Mr. Barrie and I made a “tumbler” of toddy each, WILL BE
a rare thing for him, but he said it was “New Year BAIRNS.”
time,” and an “occasion;” and my health was drunk,
and that of Agnes, in which Mrs. Barrie joined, a very rare thing for
her; and Mr. Barrie had just said, “Now, my dear, you must give Mr.
Martin the benefit of a little of your experience,” when the door-
handle was slowly turned, evidently by a less firm hand than Bell’s,
and a little head and part of a little white nightgown, appeared at the
half-opened door, and a voice was heard timidly saying, “Mamma,”
followed by Bell’s voice, which, with a mixture of astonishment and
anxiety in its tone, was heard saying, “James—here—at this time o’