Professional Documents
Culture Documents
Textbook Methods and Applications For Modeling and Simulation of Complex Systems 18Th Asia Simulation Conference Asiasim 2018 Kyoto Japan October 27 29 2018 Proceedings Liang Li Ebook All Chapter PDF
Textbook Methods and Applications For Modeling and Simulation of Complex Systems 18Th Asia Simulation Conference Asiasim 2018 Kyoto Japan October 27 29 2018 Proceedings Liang Li Ebook All Chapter PDF
https://textbookfull.com/product/prima-2018-principles-and-
practice-of-multi-agent-systems-21st-international-conference-
tokyo-japan-october-29-november-2-2018-proceedings-tim-miller/
https://textbookfull.com/product/information-retrieval-24th-
china-conference-ccir-2018-guilin-china-
september-27-29-2018-proceedings-shichao-zhang/
https://textbookfull.com/product/knowledge-and-systems-
sciences-19th-international-symposium-kss-2018-tokyo-japan-
november-25-27-2018-proceedings-jian-chen/
Liang Li
Kyoko Hasegawa
Satoshi Tanaka (Eds.)
123
Communications
in Computer and Information Science 946
Commenced Publication in 2007
Founding and Former Series Editors:
Phoebe Chen, Alfredo Cuzzocrea, Xiaoyong Du, Orhun Kara, Ting Liu,
Dominik Ślęzak, and Xiaokang Yang
Editorial Board
Simone Diniz Junqueira Barbosa
Pontifical Catholic University of Rio de Janeiro (PUC-Rio),
Rio de Janeiro, Brazil
Joaquim Filipe
Polytechnic Institute of Setúbal, Setúbal, Portugal
Igor Kotenko
St. Petersburg Institute for Informatics and Automation of the Russian
Academy of Sciences, St. Petersburg, Russia
Krishna M. Sivalingam
Indian Institute of Technology Madras, Chennai, India
Takashi Washio
Osaka University, Osaka, Japan
Junsong Yuan
University at Buffalo, The State University of New York, Buffalo, USA
Lizhu Zhou
Tsinghua University, Beijing, China
More information about this series at http://www.springer.com/series/7899
Liang Li Kyoko Hasegawa
•
123
Editors
Liang Li Satoshi Tanaka
Ritsumeikan University Ritsumeikan University
Kusatsu, Shiga, Japan Kusatsu, Shiga, Japan
Kyoko Hasegawa
Ritsumeikan University
Kusatsu, Shiga, Japan
This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721,
Singapore
Preface
The Asia Simulation Conference (AsiaSim) is the international conference that has the
longest history as a conference of modeling and simulation in Asia. It is an annual
conference organized by the Federation of Asia Simulation Societies (ASIASIM),
whose current member societies are: CSF (China Simulation Federation), JSST (Japan
Society for Simulation Technology), KSS (Korea Society for Simulation), SSAGsg
(Society of Simulation and Gaming of Singapore), and MSS (Malaysian Simulation
Society). This conference provides a forum for scientists and engineers from around the
world to promote the advancement of modeling and simulation in academic and
industrial communities.
AsiaSim 2018 was held in Kyoto, Japan. We received about 100 full papers.
Submissions came from China, Japan, Korea, Singapore, Malaysia, Columbia, and
Italy. After an intensive review process by the internationally assembled Program
Committee, where each paper was reviewed by multiple reviewers, we finally accepted
45 full papers. Due to the high quality of the submitted papers, the paper selection was
very difficult and we were forced to reject many interesting papers. The accepted papers
are now consolidated in this volume of the Communications in Computer and Infor-
mation Science (CCIS) series of Springer, and are divided into relevant topics. The
diversity of topics is a unique and important feature of the AsiaSim conference. Giving
researchers of different fields opportunities to get together and exchange ideas has
inspired many interesting research activities. We hope the publication of this volume
will further promote this nice feature of the AsiaSim conference.
We thank the members of the Program Committee for their valuable effort in
reviewing submitted papers. We also thank the Organizing Committee, which sup-
ported our editorial work in various aspects. We also express our special thanks to the
College of Information Science and Engineering and ICT Medical Healthcare Center of
Ritsumeikan University, the co-sponsor of the conference. Finally, we thank all the
authors and participants of AsiaSim 2018.
General Chairs
Kazuo Furuta University of Tokyo, JSST President, Japan
Satoshi Tanaka Ritsumeikan University, Japan
Steering Chair
Kyoko Hasegawa Ritsumeikan University, Japan
Program Chairs
Liang Li Ritsumeikan University, Japan
Naohisa Sakamoto Kobe University, Japan
Technical Co-sponsors
Co-organizers
Co-sponsors
Program Committee
Satoshi Tanaka Ritsumeikan University, Japan
Liang Li Ritsumeikan University, Japan
Kyoko Hasegawa Ritsumeikan University, Japan
Hiroshi Tamura Chuo University, Japan
Norifumi Yamada Fukui University, Japan
Akinori Kimura Ashikaga Institute of Technology, Japan
Hiroaki Nakamura National Institute for Fusion Science
Naohisa Sakamoto Kobe University, Japan
Yoshiyuki Miyamoto AIST, Japan
Yifa Tang Chinese Academy of Sciences, China
Taku Itoh Nihon University, Japan
Shigeru Shimamoto Waseda University, Japan
Soo-Hyun Park Kookmin University, South Korea
Shin Muroya Matsumoto University, Japan
Kazuo Furuta The University of Tokyo, Japan
Masami Iwase Tokyo Denki University, Japan
Katsuhisa Ozaki Shibaura Institute of Technology, Japan
Shafishuhaza Sahlan Universiti Teknologi Malaysia
Fumiaki Araki JAMSTEC, Japan
Yosuke Onoue Nihon University, Japan
Susumu Nakata Ritsumeikan University, Japan
AsiaSim 2018 Organization IX
Demand and Supply Model for the Natural Gas Supply Chain in Colombia . . . 220
Mauricio Becerra Fernández, Elsa Cristina González La Rotta,
Federico Cosenz, and Isaac Dyner Rezonzew
Flow Simulation
1 Introduction
Simulation model is a complex object, which usually has multiple inputs/outputs and
sophisticated behaviors. The credibility of a simulation model is influenced by many
indicators (factors) related to the nature and outputs of the model. It is necessary to
build a factor space [1, 2] to describe these factors and their relationship, and further
aggregate the total credibility of the model through the factors. Traditional way of
building a factor space is using an AHP (Analytic Hierarchy Process) tree [3] and
weighted average function to aggregate the grand credibility from leaf nodes. However,
the relation between factors is not always linear, and the influence of the factor is not
always transferred through layers one by one. Model validation needs a better method
to reveal the factors and their influence to the credibility of the model.
Many generic methodologies develop the requirements, planning, architecture,
process and recommended techniques to perform verification and validation for M&S
(Modeling and Simulation) [4, 5]. Unfortunately, these methodologies all suggest
hierarchical tree to present and organize credibility indicators of model validation. Even
the project exercises and VV&A software tools adopt the traditional hierarchy to
perform simulation validation [6]. This cannot satisfy the practical requirements of the
model validation, and often leads to a credibility assessment result lack of objectivity
and confidence.
The remainder of this paper is organized as follows. We state the three challenging
problems of the factor space which are not resolved well by the hierarchical tree in
© Springer Nature Singapore Pte Ltd. 2018
L. Li et al. (Eds.): AsiaSim 2018, CCIS 946, pp. 3–17, 2018.
https://doi.org/10.1007/978-981-13-2853-4_1
4 K. Fang et al.
Sect. 2. Section 3 proposes the network method for building a factor space of model
validation. Section 4 presents a case study of the electromagnetic rail gun model
validation to explain the usage of the network method. Concluding remarks are given
in Sect. 5.
2 Challenging Problems
The basic function of the factor space in model validation is to determine the total
credibility of the model. In order to achieve the goal, factor space needs to reveal the
credibility indicators and their relationship, and use them to aggregate the total cred-
ibility. There are some challenging problems in the function of the factor space,
especially the factor decomposition, credibility aggregation and the model defect
tracing.
Actually the indicators of four motions are decomposed from the root by the model
structure, and this is what the hierarchical tree does in AHP. However when the
A Novel Method to Build a Factor Space for Model Validation 5
decomposition goes down further it is stuck, because the physical quantities are linked
by the model resolving process but not the structure. Obviously the outputs (both
intermediate and final ones) are not always irrelevant and the relationship is not always
linear, thus it is incorrect to make structural decomposition further.
Meanwhile, in case that the outputs x; y; z are so important that if they failed to meet
some acceptability criteria, no matter what the performance of other indicators are, the
model credibility will be unaccepted. This is an ultra-connection from x; y; z to the root
node, which a traditional hierarchy cannot deal with. So we need a better way to
perform factor decomposition.
X
q Y
ki
v0 ¼ ðvi wj Þ; ð1Þ
i¼1 j¼1
where v0 is the root value (model credibility), vi is the leaf value, wj is the weight and
should be summed to 1 among brother nodes, q is the number of leaf nodes, and ki is
the number of ancestor nodes which belong to the leaf node possessing vi . Apparently,
the weighted average function is linear, and demands the brother nodes are irrelevant
each other at different levels. Actually this is almost impossible to be satisfied in the
model validation. See the factor space part expanded from the “Motion of mass center”
node of Fig. 2 below.
Fig. 2. Factor space part expanded from the “Motion of mass center” node of Fig. 1.
6 K. Fang et al.
Because m changes during the simulation, Fqx1 is not linear to Wx . If use normal
AHP hierarchy and weighted average function to aggregate the partial credibility of Wx
from Fqx1 and m, it will be an incorrect result. Actually the relationship of Fqx1 and m to
Wx is a derivation, but not a composition. So the credibility aggregation needs better
solution than the traditional way.
The hierarchical tree has a limitation of connections inside, that is, it only allows
relationship between factors in adjacent layers. In order to express more complicated
connections between credibility indicators, we propose a network method to build the
factor space in model validation.
where F is a factor space, N and V are the node set and value set, L and A are the link
set and attribute set mapped with each other. Set the link direction as from attribute-
holding node to attribute-receiving node. For example, a structural link of the
A Novel Method to Build a Factor Space for Model Validation 7
traditional hierarchy has the direction from child node to parent node. According to the
requirements of model validation, further develop element definitions of the network as
the following:
Definition 1: Define N ¼ fn1 ; n2 ; . . .; nk g as the node set. N ¼ N [N~ where N is the
certain node set and N~ is the uncertain node set. If ni 2 N,
use tðni Þ to indicate the node
~
type. tðni Þ 2 fregular; sufficient; inherited g. If ni 2 N, use cðni Þ as the transit condi-
tion, and cðni Þ 2 f0; 1g.
Definition 2: Define V ¼ fv1 ; v2 ; . . .; vk g as the value set. If tðNÞ 2 fregular;
sufficientg, V and N are mapped with each other. Use vi ¼ vðni Þ to indicate the value of
ni which is mapped with vi .
[L
Definition 3: Define L ¼ fl1 ; l2 ; . . .; lk g as the link set. L ¼ L ~ where L is the certain
link set and L ~ is the uncertain link set. Use tðli Þ to indicate the link type, and
tðli Þ 2 fregular; sufficient; ultra;equivalent; contraditory; inherited; tracedg. Use to l ¼
ðni ; nj Þ express a link from node ni to node nj .
Definition 4: Define A ¼ fa1 ; a2 ; . . .; ak g as the attribute set, and A is one-one mapped
with L. Use ai ¼ aðli Þ to indicate the attribute of link li which is mapped with ai . If
tðli Þ 2 fequivalent; inherited; traced g, ai 2 £.
~ and cðnÞ ¼ 1, then n 2 N.
Definition 5: If n 2 N If cðnÞ ¼ 0, then n 2 ;, and define n
as rubbish node, which needs to be deleted from the network.
tðnÞ ¼ sufficient and vðnÞ ¼ 1, then the sufficient link from n
Definition 6: If n 2 N,
breaks. If vðnÞ ¼ 0, then n needs to supplement additional brother node, which is
defined as shadow node and values 0.
Definition 7: If l 2 L, tðlÞ ¼ regular, and l ¼ ðno ; nd Þ, define aðlÞ 2 ½0; 1 as the weight
and tðLc Þ ¼ regular, then
distributed from nd to no . If Nc ¼ fnjðn; nd Þ 2 Lc g, Lc 2 L,
Pk
ak ðlk Þ ¼ 1, lk 2 Lc and k ¼ dðLc Þ.
i¼1
Definition 12: If l 2 L, l ¼ ðno ; nd Þ, and tðlÞ ¼ traced, define nd as traced node. vðnd Þ
is irrelevant to vðno Þ.
Definition 13: If n ¼ fxjx 2 N ^ ðx; nÞ 2 Lg, define n as the pre-set of n. If
n ¼ fxjx 2 N ^ ðn; xÞ 2 Lg, then define n as the post-set of n.
Definition 14: If ðni ; nj Þ 2 L, define ni as the child node of nj , and nj as the father node
of ni . If ðni ; np Þ 2 L and ðnj ; np Þ 2 L, define ni and nj are brother nodes. If n ¼ ;,
define n as root node. If n 6¼ ; and n 6¼ ;, define n as branch node. If n ¼ ;, define n
as leaf node.
Definition 15: Set power operator satisfies ðn Þ0 ¼ n, ðn Þ1 ¼ n , ðn Þ2 ¼ ðn Þ … If
nj 2 ðni Þs1 , nj 2 ðni Þs2 ,…, nj 2 ðni Þsk , define Sðni ! nj Þ ¼ fs1 ; s2 ; . . .; sk g as the dis-
tance set from ni to nj . Non-negative integers s1 sk are all distances from ni to nj . If n0
is a root node, abbreviate Sðni ! n0 Þ as Sni .
Definition 16: If NL ¼ fn1 ; n2 ; . . .; nk g, 8n 2 NL makes n \ NL ¼ ; and n \ NL ¼ ;,
and 8ni ; nj 2 NL , 8sk 2 Sni , 8sm 2 Snj makes Maxðsk Þ ¼ Maxðsm Þ, define NL as a layer
(order) of the factor space network, which is the Maxðsk Þ–th layer.
Definition 17: If 9s; nj 2 ðni Þs , define ni to nj as reachable. If 8s; nj 62 ðni Þs , define ni
to nj as unreachable, and set sðni ! nj Þ ¼ 1.
Definition 18: If sðni ! nj Þ [ 1 and sðni ! nj Þ 6¼ 1, define ni as the offspring node
of nj , nj is the ancestor node of ni , and ni & nj are lineal relative nodes. If sðni !
nj Þ ¼ 1 and ni & nj are not brother nodes, define ni & nj are collateral relative nodes.
When the dynamic analysis is done, the credibility aggregation can be achieved by
the following procedure:
(1) Go through the factor space downwards, and stop at the destinations of traced
links. Perform similarity analysis by comparing the simulation outputs and real
world outputs, and get partial credibility on these nodes after result transformation
[8]. Generally, the partial credibility can be achieved by:
where IðnÞ and OðnÞ are the simulation input and output of the node, Iðn0 Þ and
Oðn0 Þ are the corresponding real world input and output, and jj jj represents the
norm of the variant. According to the technique used, CðnÞ can be achieved by
statistics analysis [9], time domain analysis [10] or frequency domain analysis
[11] methods together with their result transformation formulas.
(2) If the destination of the traced link is lack of real world output, then go down
along the path further to find a node whose real world output is valid, and get the
value of the node by similarity analysis. The partial credibility of the upper node
which has no real world output can be achieved by error analysis via the com-
putational process of the model:
where n1 is the node which has real world output, Cðn1 Þ is partial credibility, and
Oðn1 Þ is simulation output. n2 is the node which has no real world output, Cðn2 Þ is
partial credibility, and Oðn2 Þ is simulation output. f is the computational function
of the model from n1 to n2 .
(3) Use appropriate algorithm to aggregate the total credibility on the root node, by
gathering the partial credibility on the destinations of the traced links along the
decomposition paths of the factor space:
C ¼ f ðvi ; vi þ 1 ; . . .; vi þ k Þ; ð6Þ
(2) Along the validation path in the factor space which contains the defect factors
already located, make further validation to get partial credibility on the destination
of traced links, and determine if it is a defect node by comparing with the
acceptability criteria.
(3) If there is a destination of iteration paths, make validation of its initial input, related
constants, and other variants which are irrelevant to the iteration variant [12].
(4) When the validation path reaches the leaf node, the defect tracing is over.
(5) Collect all the defect nodes in the tracing, and take the nodes with lowest order as
the origin that induces the deficiency of the model credibility.
4 Case Study
The electromagnetic rail gun model is composed by the sub-models of power supplies,
paralleled rails, armature, wave modulation inductor, projectile etc. The armature is
located between the two rails, and conducts the current and delivers Lorentz force. The
projectile is located in front of the armature, which is forced to move forward together
with the armature. The model reveals the whole physical process of the real world
object.
Fig. 4. The factor space network of the electromagnetic rail gun model.
8
< 1Cth ðc c Þ þ C ; c 2 ½c ; 1
1c m th th m th
Cðcm Þ ¼ th
Cth cm ; ð8Þ
: ; cm 2 ½0; cth
cth
simulation data and observed data are shown in Table 1. The data curves of the
projectile displacement x are shown in Fig. 5.
Set cth ¼ 0:5 & Cth ¼ 0:8, and use the function in Formula (7, 8), we can obtain the
partial credibility of the projectile displacement x is Cx ¼ 0:7049. Because Cth ¼ 0:8,
the node is not accepted. Meanwhile we can see the root node is directly linked by the
projectile displacement x only, so the model’s total credibility is also 0:7049 and
unaccepted. We need to perform defect tracing to find which part of the model causes
the credibility deficiency.
Table 1. The simulation data and observed data of the projectile displacement x.
Time (ms) Simulation data (m) Observed data (m)
0.1000 0.0001 0.0001
0.2000 0.0009 0.0009
0.3000 0.0048 0.0050
0.4000 0.0148 0.0155
0.5000 0.0335 0.0356
…… …… ……
1.1000 0.3404 0.3796
1.2000 0.4246 0.4746
…… …… ……
4.0000 4.6349 5.2051
4.1000 4.8064 5.3973
4.2000 4.9782 5.5899
4.3000 5.1503 5.7828
4.4000 5.3227 5.9760
Table 2. The simulation data and observed data of the No. 9 individual circuit current i9 .
Time (ms) Simulation data (kA) Observed data (kA)
0.8000 0.0000 0.0000
0.9000 63.4997 63.3014
1.0000 118.0873 118.6114
1.1000 156.2808 159.1810
1.2000 172.6993 179.9446
…… …… ……
2.6000 71.6777 75.8113
2.7000 67.5994 71.4827
…… …… ……
4.0000 33.8028 35.8419
4.1000 32.1987 34.1577
4.2000 30.6890 32.5729
4.3000 29.2671 31.0806
4.4000 27.9271 29.6744
Use the function in Formula (7, 8), we can obtain the partial credibility of the
No. 9 individual circuit current i9 is Ci9 ¼ 0:7393. The node is not accepted and we
need further defect tracing along the traced links downwards.
Table 3. The simulation data and observed data of the rail voltage Ur .
Time (ms) Simulation data (V) Observed data (V)
0 0 0
0.1000 0.6055 0.6286
0.2000 6.6897 7.2459
0.3000 29.8628 33.4997
0.4000 72.0103 84.9831
0.5000 110.6497 134.7681
…… …… ……
1.1000 402.5833 493.5161
1.2000 410.8884 509.3004
…… …… ……
4.0000 −42.0534 −47.6174
4.1000 −43.5441 −49.2517
4.2000 −44.7558 −50.5756
4.3000 −45.7219 −51.6273
4.4000 −46.4721 −52.4406
Use the function in Formula (7, 8), we can obtain the partial credibility of the rail
voltage Ur is CUr ¼ 0:6849. The node is not accepted.
5 Conclusion
The factor space plays an important role in model validation. It reveals all the indicators
that influence the total credibility of the model, and makes the work of model defect
tracing possible. However, the traditional way of factor space building only analyses
the model by structural decomposition, and relies on the linear aggregation function to
get the total credibility. This cannot satisfy the factor space requirements in the engi-
neering of model validation. The network based factor space overcomes the limitation
of the traditional hierarchy, and serves for the model validation better. Case studies
prove that the method is practical and effective.
Meanwhile we should notice that, although the factor space lists all the credibility
indicators and their relationship, it does not mean the validation can be performed
thoroughly without doubt. Some physical quantity is not observable or very hard to
measure in the real world, such as the aerodynamic force, the projectile mass etc. The
node possessing this kind of physical quantity is called “black node”. We cannot
perform similarity analysis from comparing the simulation data and observed data and
get the partial credibility, except for knowledge based expert evaluation. Therefore, if
there is black node in the factor space, the validation result may be less of objectivity.
The definition of the network based factor space is extendable. The types of each
quadruple element can be extended to express new indicators and relation-
ship. According to the practical requirements in validation engineering, the network
method for building a factor space should be more adaptive and accuracy.
Another random document with
no related content on Scribd:
Polybia, nests of, 81, 82, 83
Polyctenes fumarius, 560
Polyctenidae, 560
Polyergus lucidus, 151;
P. rufescens, 150 f.
Polymorpha (Coleoptera), 189, 190, 213 f.
Polymorphism, 139 f., 143
Polyphylla fullo, antenna, 191
Polyploca, 386
Polyplocidae, 386
Polyplocotes, 248
Polyrhachis, 155;
P. pandurus, 156;
P. spinigera, 138
Pompilidae, 93, 101 f.
Pompilus, 103;
P. polistoides, 104;
P. sericeus, 106
Ponera contracta, P. ergatandria, P. punctatissima, 172
Ponerides, 132, 170 f.
Porphyraspis tristis, 284
Porphyrophora polonica, 597
Porrorhynchus, 216
Portschinsky, on habits and development of Muscidae, 507, 512
Post-metamorphic growth, 141
Post-scutellum, 307, 312
Potamogeton pectinatus, beetle on, 280
Potamophilus acuminatus, 244
Potassium hydroxide, 328, 384
Potato-beetle, 278
Pouch, abdominal, 350, 362
Poulton, on colours, 336, 339
Praecostal nervures, 319
Praescutum, 312
Pratt, on imaginal discs, 453 n.;
on Melophagus, 519
Prepona, larva, 354
Prey, of Fossores, table, 92
Pria dulcamarae, 232
Primitive, beetles, 251, 252;
Diptera, 475
Priocnemis affinis, 5;
P. bicolor, 107
Priodont, 193
Prionides, 287
Proboscis, 13, 14, 16, 17, 304, 307, 309, 311, 443, 482, 485,
532
Processional maggots, 464
Processionary caterpillars, 376, 408
Prodoxidae, 432;
mouth, 309
Prodoxus, 433
Pro-legs, 323
Prolific, Aphids, 589;
Coccids, 594;
Hepialus, 397;
Lice, 601;
Meloe, 274;
Stylops, 301
Prominents, 383
Pronuba, 321;
P. yuccasella, 432;
P. synthetica, 432
Pronymph, 453
Propodeum, 131, 133
Propolis, 63
Propygidium, 187
Prosopis, 21, 22;
hair of, 11;
proboscis of, 17;
P. signata, 21
Protection, 43, 413;
of trees by ants, 158;
of plant by ants, 168
Proterhinidae, 298
Proterhinus lecontei, 298
Protolepidoptera, 336
Protoparce carolina, 309
Protopaussus, 214
Psammorycter vermileo, 481
Pselaphidae, 223;
and ants, 182
Psen atratus, P. concolor, 127
Psephenus, 244
Pseudholoptic, 440 n.
Pseudocorylophidae, 228
Pseudodicthadia, 180;
P. incerta, 177
Pseudodoxia limulus, 431
Pseudomeria graeca, 99
Pseudomorphides, 205, 206
Pseudomyrma bicolor, 168
Pseudomyrmini, 168
Pseudoneuroptera, 527
Pseudopaedogenesis, 303
Pseudopod, 188, 264, 267, 290, 449, 492
Pseudopontia paradoxa, 357
Pseudo-pupa, 271, 273
Pseudotetramera, 190
Pseudotrimera, 239
Pseudovespa, 88
Pseudovarium and Pseudovum, 584
Pseudovitellus, 588
Psilidae, 504
Psiliglossa, larva, 8
Psilocephala, 484
Psilura monacha, 407
Psithyrus, 53, 57, 59 f.;
P. campestris, 60;
P. vestalis, 60
Psocidae, 248
Psyche helix, 394
Psychidae, 369, 392
Psychina, 394, 395, 404
Psychoda, 466
Psychodidae, 470
Psylla pyricola, 579, 580;
P. succincta, 579;
P. buxi, 580
Psyllidae, 578
Pterocheilus, 76
Pterodecta, 400
Pterophoridae, 340, 371, 425, 426
Pterophorinae, 426
Pterostichus, 205
Pterothysanidae, 369, 406
Pterygodes, 312
Pterygogenea, 542
Ptilinum, 442, 503, 520
Ptilinus pectinicornis, 253
Ptilocnemus sidnicus, 557
Ptilomacra, 395
Ptilomera laticaudata, 553
Ptinidae, 246
Ptomaphila lacrymosa, 222
Ptychoptera, 466
Ptychopterinae, 472
Ptyelus goudoti, 577
Pugs, 411
Pulex avium, P. irritans, P. serraticeps, 525;
P. obtusiceps, 525
Pulicidae, 522 f.
Puliciphora lucifera, 495
Pulvillus, 446
Pupa, of beetles, 188 f.;
coarctata, 452;
obtecta, 227, 327, 451;
depositing eggs, 469;
hairy, 407, 426;
of Lepidoptera, 326 f.;
with mandibles, 436
Puparium, 452
Pupation, of Badamia, 365;
of Parnassius and Thais, 363
Pupipara, 456, 513, 517
Purple emperor, 344
Pusci, 504
Puss-moth, 328, 383, 406
Pygidium, 187
Pyralidae, 340, 370, 420
Pyralidina, 420, 426
Pyrameis atalanta, 353;
P. cardui, 353;
P. tameamea, 353
Pyraustidae, 421
Pyrochroa rubens, 266
Pyrochroidae, 266
Pyrophorus, 258;
P. noctilucus, 259
Pyrrhocoridae, 549
Pyrrhocoris apterus, 535, 549
Pyrrhopyge, 364
Pythidae, 265
Pytho depressus, 266
Quartenia, 89
Queen, 66, 67, 69, 140
Races, of Apis, 68
Radial nervures, 319
Raffray, on Pogonostoma, 204;
on classification of Paussidae, 214;
on classification of Pselaphidae, 224
Railway-beetle, 251
Ranatra linearis, 563
Raptorial legs, 493, 554, 556
Réaumur, on Xylocopa, 33
Receptaculum seminis, 140
Rectal cauda, 538
Red admiral, 352
Red ant—see Formica rufa
Reduviidae, 555 f. 537
Reduvius personatus, 558
Régimbart, on Gyrinidae, 216
Resemblance, between ant, wasp, and spider, 169;
between Anthophora and Bombus, 33;
between Arctia villica and Eusemia villicoides, 410;
between beetle-larva and Termite, 206;
between Bombus and Eulema, 35;
between Bombyliidae and Hymenoptera, 489;
between bug and ant, 556, 557;
bug and Tipulid, 556, 558;
bug and fly, 547;
in butterflies, 348;
between Callidulidae and Lycaenidae, 400;
between Celyphus and Hemiptera, 505;
between Cleridae and insects they destroy, 254;
between Dioptinae and Ithomiides, 409;
between Diptera and Hymenoptera, 499, 500;
between Epicopeia and Papilio, 418;
between Flatides and Lepidoptera, 576;
of flies and bees, 502;
to galls, 403;
between host and parasite-bees, 30;
between Insects of different Orders, 339;
of Ithomiides to other butterflies, 346;
between lady-bird and Endomychid, 237;
of larva to a colony of larvae, 418;
of larva of Odynerus and of Chrysis, 4;
between larvae, 162;
in Limacodidae, 401;
of Lobster caterpillar, 385;
between Longicorns and Hymenoptera, 287;
of moth to bird-excrement, 401;
of parasite and prey, 95;
of Pericopinae and Heliconiides, 409;
between protected butterflies, 345;
between Psithyrus and Bombus, 59;
of Reduviid and Pepsis, 558;
between Rhyphus- and Mycetobia-larvae, 463;
of Syntomids to other Insects, 388 f.;
of Tipulids and Hymenoptera, 475;
between two kinds of ants, 162
Resting-larva, 306
Retinaculum, 316, 319, 420
Retort-shaped bodies, in mouth of Hemiptera, 535
Reuter, E., on classification of butterflies, 343 n.
Rhachicerus, 480
Rhagovelia plumbea, 552
Rhaphiorhynchus, 483
Rhegmatophila alpina, 305, 386
Rheumatobates bergrothi, 553
Rhinomacerides, 291
Rhinopsis, 115;
R. ruficornis, 169
Rhinosimus, 266
Rhipicera mystacina, 256
Rhipiceridae, 256
Rhipidiini, 267
Rhipidioptera, 543
Rhipiphoridae, 267
Rhipiphorides, 268
Rhipiptera, 298 [in error for Rhiphiptera]
Rhizophagus, 232
Rhizotrogus, 191
Rhodoneura, 405
Rhogmus, 179, 180
Rhopalocera, 340, 341 f.
Rhopalomelus angusticollis, 206
Rhopalomeridae, 504
Rhopalosoma poeyi, 100
Rhopalosomides, 100
Rhygchium brunneum, R. carnaticum, R. nitidulum, R. oculatum,
77
Rhynchites betulae, 291, 292
Rhynchitides, 291
Rhynchophora, 190, 277, 288 f.
Rhynchophorous series, 240
Rhynchopsylla pulex, 526
Rhynchota, 532
Rhyphidae, 478
Rhyphus, 463
Rhysodidae, 201, 234
Riley, on Epicauta, 270, 271;
on spermathecal bodies, 321;
on Yucca-moth, 432
Ripidius pectinicornis, 269
Rippe, of Schäffer, 334
Robber-flies, 491
Rose-chafers, 200
Rosema, 401
Rostrum, 291, 472
Rothney, on Ampulex, 115;
on Sima, 169;
on Sphex, 110
Rothschild, N. C., on morphology of fleas, 523 n.
Royal jelly, 66
Rozites gongylophora, 167
Ruby-wasps, 1 f.
Rurales, 342
Rutelides, 195, 198
Saccoderes, 558;
S. tuberculatus, 537
Saccus, 314
Sagra splendida, 279
Sagrides, 279
Salda, 560
Saldidae, 544, 562
Salivary, duct, 320;
gland, 320
Sand-flea, 525;
-flies, 477
Saperda populnea, 285, 286
Sapromyzidae, 504
Sapyga quinquepunctata, 100
Sapygides, 99
Sarcophaga, 510;
S. carnaria, 510
Sarcophagidae, 510
Sarcophila magnifica, S. wohlfahrti, 510
Sarcopsylla gallinacea, S. penetrans, 525
Sarginae, 479
Saropoda, 32
Sarrothripus, 410
Sasaki, on parasite of silkworm, 508
Saturnia, 307, 310;
S. pavonia, 313, 374
Saturniidae, 368, 372
Satyrides, 347
Sauba, or Sauva, ant, 137
Saunders, E., on feathered hairs, 11;
on proboscis of bees, 16
Saunders, Sir S., on Hymenopterous larvae, 8
Scale, 131, 133, 315, 467;
development of, 329 f.;
-formation, 333;
-holder, 331
Scale-Insects, 592 f.;
enemy of, 356, 357, 417, 425, 430
Scalpella, 443
Scape, 441
Scaphidiidae, 229
Scaphisoma agaricinum, 229
Scaphium, 314
Scapulae, 312
Scarabaeidae, 194 f.
Scarabaeini, 196
Scarabaeus sacer, 196, 197
Scatomyzidae, 504
Scatophagidae, 504
Scatopse, 477
Sceliphron, 112;
S. nigripes, 91
Scenopinidae, 484
Schäffer, on structure of wings and nervures, 330
Schiödte, on Anoplura, 599 f.;
on Hemiptera, 543
Schistocerca peregrina, enemy of, 506, 514
Schizocarpus mingaudi, 220
Schizometopa, 504
Schizoneura, 589
Schizophora, 455
Schmidt-Schwedt, on Donacia, 280
Schoenbergia, 360, 361
Schoenobiinae, 425
Sciapteron, 387
Sciara militaris, 464
Sciomyzidae, 504
Sciophila unimaculata, 462
Scolia bifasciata, S. haemorrhoidalis, 97
Scoliidae, 93, 94 f.
Scoliides, 97 f.
Scolytidae, 294
Scopariidae, 421
Scopariinae, 421
Scopelodes, 401
Scopulipedes, 20, 32 f.
Screw-worm, 512
Scudder, on butterflies, 543
Scutata or Scutati, 546
Scutellerides, 545
Scutelligera, 501
Scutellum, 307, 312, 537
Scutum, 307;
S. proboscidis, 443
Scydmaenidae, 223
Scymnites, 238
Scymnus minimus, 238
Seasonal dimorphism or variation, 335
Seitz, on Syntomid resemblances, 388, 389
Semi-loopers, 415
Seminal duct, 321
Semi-pupa, 271
Semper, on development of wing, 333
Sense-organ, 442, 448;
thoracico-abdominal, 414;
in Uraniidae, 419
Sepsidae, 504
Sericaria mori, 375
Sericteria, 325
Seroot-fly, 482
Serricornia, 189, 213, 255
Sesia, 387;
S. scoliaeformis, 321
Sesiidae, 386, 388, 370 (for Syntomidae), 389
Setae, 534 f.;
aërostatic, 408
Setina, 410
Seventeen-year Cicada, 569
Sex, differences, 92, 95;
production of, 32, 67;
in larva, 325
Sexuparous, 586
Sharp, on classification of Dytiscidae, 213 n.
Sheep, bot-fly, 517;
-tick, 518
Shield, 592
Shoulder, -lappets, 312;
-tufts, 312;
of wing, 316, 319
Siagona, 206
Siculodidae, 423
Siebold, von, on Strepsiptera, 301
Sigara minutissima, 568
Sight, of Pompilus, 104
Silk-glands, 325
Silkworm, 375;
affected by parasitic fly, 507, 508;
Madagascar-, 405
Silpha, 221;
Silpha atrata, S. laevigata, S. lapponica, 222;
S. obscura, 222;
S. opaca, S. thoracica, 222
Silphidae, 221 f., 252, 256
Sima, 168;
S. leviceps, stridulating organ, 169;
S. rufonigra, 169
Simuliidae, 477
Simulium columbaczense, 477
Sinodendron cylindricum, 194
Siphon, 563, 581, 589
Siphonaptera, 522 f.
Siphonophora, 239
Siphunculata, 600
Sitaris, 33;
S. humeralis, 272
Sitodrepa, 247
Skippers, 363
Slave-making ant, 149, 150, 163
Sloth, 430
Slug-worms, 402
Smallest Insect, 228
Smerinthini, 380
Smerinthus populi, 309, 381
Smith, F., on Mellinus arvensis, 123
Snails, enemies of, 205, 222, 510;
parasite of, 495
Social, bees, 35;
wasps, 78, 84
Sociales, 20, 53 f.
Solanum dulcamara, beetle on, 232
Soldiers, 132
Soldier-ant, 150
Solenobia, 395, 430
Solenopsis fugax, 137
Solitary wasps—see Fossores and Eumenidae
Song, of Cicada, 572
Soronia, 232
Sound-organs, 448;
of Ageronia, 354;
of Hecatesia, 371;
of Cicada, 573, 574—see also stridulating organs
Sound-production, 155, 156;
by Aegocera, 411;
by Arctiidae, 410;
by Sphingidae, 382—see also Stridulation
Spalacopsini, 288
Spatula, 459
Spencer, Herbert, on Weismann, 143
Spercheus emarginatus, 218, 219
Spermatheca, 320, 321
Spermophila, 506
Sphaeridiides, 219
Sphaeriidae, 227
Sphaerites, 223
Sphaeritides, 229
Sphaerius acaroides, 227
Sphaerocarides, 279
Sphecia, 387
Sphecius speciosus, 123
Sphecodes, 21, 22;
S. gibbus, 23, 23;
S. rubicundus, 22;
S. subquadratus, 23
Sphegidae, 93, 107 f.
Sphegides, 107 f.
Sphex coeruleus, 110;
S. flavipennis, 108;
S. lobatus, 110;
S. maxillosus, 108
Sphindidae, 245
Sphingidae, 309, 315, 316, 368, 380 f.
Sphinx ligustri, 380
Spider parasite, 490
Spilosoma, 408
Spinneret, 324, 325, 403, 417
Spirachtha, 227
Spiracles, 188, 191;
of Diptera, 449 f.;
of Hippoboscidae, 519;
of Lepidoptera, 313, 314;
of Lipara, 451;
of Nepa, 564;
of Thrips, 528
Spondyliaspis, 581
Spondylidae, 288
Springing plant-lice, 579
Spuler, on nervures, 317 n.
Squama, 448
Squeakers, 209
Staetherinia, 401
Stag-beetles, 193
Stalk, 317, 319
Staphylinidae, 223, 224 f.
Staudinger, Schatz and Röber, on butterflies, 343 n.
Stauronotus maroccanus, 489
Stauropus fagi, 385
Stelis minuta, 29;
S. nasuta, 30, 43;
S. signata, 30
Stelocyttares, 81
Stenamma westwoodi, 159
Stenopteryx hirundinis, 519
Stephostethus, 240
Sternorhyncha, 544
Sterrhopteryx, 394
Stethopathidae, 496
Stigmata—see Spiracles
Stigmatomma, 180
Stigmus pendulus, 128
Sting, 4, 5, 6, 58, 144;
development of, 8, 9
Stinging, 98;
by Calicurgus, 102;
by Pompilidae, 104;
by Sphex, 109
Stingless bees, 61
Stink-gland, 257, 533;
-vessel, 225
Stipes, 309
Stizinae, 123
Stomach, 320
Stomoxys calcitrans, 512
Stratiomyidae, 478
Stratiomys, 452
Straus-Durckheim, on Melolontha, 198
Strawberries, eaten by beetles, 205
Streblidae, 521
Strepsiptera, 189, 298 f.
Streptoperas, 401
Stridulating organ, of Myrmica, 133;
of Heterocerus, 243;
of Passalus-larva, 192;
of Sima leviceps, 169
Stridulation, of ants, 134;
of Corixa, 568;
of Criocerides, 281;
of Dynastides, 199;
of Geotrupes, 195;
of Ipides, 232;
of Lomaptera, 200;
of Longicorns, 287;
of larva of Lucanus cervus, 194;
of Megalopides, 282;
of Melolontha-larva, 198;
of Mutilla, 94;
of Pelobius, 208;
of Phonapate, 246;
of Phyllomorpha, 548;
of Praogena, 264;
of Siagona, 206;
of Trox, 195
Strigil, 568
Striphnopterygidae, 376
Strongylognathus huberi, S. testaceus, 162
Strumigenys, 170
Style, 442
Stylopidae, 298
Stylopised bees, 26
Stylops dalii, 299
Styx infernalis, 340, 358
Suana, 405
Subcostal nervure, 318
Submedian nervure, 318
Suboesophageal ganglion, 320
Sucking-stomach, 311, 449
Suction by Lepidoptera, 311
Suctoria, 526
Supericornia, 546
Swallow-flies, 519
Swarming of wasps, 70 n.
Swarms, 62, 65, 67, 80, 135, 467, 505, 584
Swift-moths, 396
Symbiosis, of ants and plants, 139;
of bee and Acarid, 70.
See also Ants'-nest Insects, and Association
Symbius blattarum, 269
Symphily, 183
Synecthry, 183
Synemon, 371
Synoeca cyanea, nest, 82
Syntelia westwoodi, 229
Synteliidae, 229
Syntomidae, 339 n., 369, 388
Syntomis phegea, 390
Syringe, 535, 536
Syrphidae, 439, 498 f.
Systoechus oreas, 489
Systropus crudelis, 489