PDF Computational Intelligence and Intelligent Systems 7Th International Symposium Isica 2015 Guangzhou China November 21 22 2015 Revised Selected Papers 1St Edition Kangshun Li Ebook Full Chapter
PDF Computational Intelligence and Intelligent Systems 7Th International Symposium Isica 2015 Guangzhou China November 21 22 2015 Revised Selected Papers 1St Edition Kangshun Li Ebook Full Chapter
PDF Computational Intelligence and Intelligent Systems 7Th International Symposium Isica 2015 Guangzhou China November 21 22 2015 Revised Selected Papers 1St Edition Kangshun Li Ebook Full Chapter
https://textbookfull.com/product/computational-intelligence-and-
intelligent-systems-10th-international-symposium-
isica-2018-jiujiang-china-october-13-14-2018-revised-selected-
papers-hu-peng/
https://textbookfull.com/product/computational-intelligence-and-
intelligent-systems-kangshun-li/
https://textbookfull.com/product/web-and-wireless-geographical-
information-systems-14th-international-
symposium-w2gis-2015-grenoble-france-
Kangshun Li · Jin Li
Yong Liu · Aniello Castiglione (Eds.)
Computational Intelligence
and Intelligent Systems
7th International Symposium, ISICA 2015
Guangzhou, China, November 21–22, 2015
Revised Selected Papers
123
Communications
in Computer and Information Science 575
Commenced Publication in 2007
Founding and Former Series Editors:
Alfredo Cuzzocrea, Dominik Ślęzak, and Xiaokang Yang
Editorial Board
Simone Diniz Junqueira Barbosa
Pontifical Catholic University of Rio de Janeiro (PUC-Rio),
Rio de Janeiro, Brazil
Phoebe Chen
La Trobe University, Melbourne, Australia
Xiaoyong Du
Renmin University of China, Beijing, China
Joaquim Filipe
Polytechnic Institute of Setúbal, Setúbal, Portugal
Orhun Kara
TÜBİTAK BİLGEM and Middle East Technical University, Ankara, Turkey
Igor Kotenko
St. Petersburg Institute for Informatics and Automation of the Russian
Academy of Sciences, St. Petersburg, Russia
Ting Liu
Harbin Institute of Technology (HIT), Harbin, China
Krishna M. Sivalingam
Indian Institute of Technology Madras, Chennai, India
Takashi Washio
Osaka University, Osaka, Japan
More information about this series athttp://www.springer.com/series/7899
Kangshun Li Jin Li
•
Computational Intelligence
and Intelligent Systems
7th International Symposium, ISICA 2015
Guangzhou, China, November 21–22, 2015
Revised Selected Papers
123
Editors
Kangshun Li Yong Liu
College of Mathematics and Informatics School of Computer Science and
The South China Agricultural University Engineering
Guangzhou The University of Aizu
China Aizu-Wakamatsu, Fukushima
Japan
Jin Li
School of Computer Science Aniello Castiglione
Guangzhou University Department of Informatics
Guangzhou University of Salerno
China Fisciano
Italy
The present volume contains the proceedings of the 7th International Symposium on
Intelligence Computation and Applications (ISICA 2015) held in Guangzhou, China,
November 21–22, 2015. ISICA 2015 successfully attracted over 189 submissions.
Through rigorous reviews, 77 high-quality papers were selected for this volume of
Communications in Computer and Information Science (CCIS 575). ISICA confer-
ences are one of the first series of international conferences on computational intelli-
gence that combine elements of learning, adaptation, evolution, and fuzzy logic to
create programs as alternative solutions to artificial intelligence. The past ISICA pro-
ceedings including three volumes of CCIS and four volumes of LNCS have been
accepted in both the Index to Scientific and Technical Proceedings (ISTP) and Engi-
neering Information (EI).
Following the success of the past six ISICA events, ISICA 2015 persisted in
exploring new problems emerging in the fields of computational intelligence. In recent
years, a number of intelligent driving systems for driverless cars have been developed.
For example, at least ten of Google’s self-driving cars, including six Toyota Prius, an
Audi TT, and three Lexus RX450h, have undergone road safety testing. Such
impressive progress makes people think that current techniques have solved all issues
in the design of an intelligent driving system in the sense of overall human perfor-
mance. However, it is simply not the case. There are still many unsolved problems. For
example, Google’s cars are not able to spot a police officer who is waving for traffic to
stop on the side of road. The car’s sensors cannot tell whether a road obstacle is a rock
or a crumpled piece of paper. It is expected that these unsolved problems in such
intelligent systems will become increasingly difficult. While it is difficult to create
intelligence directly, an intelligent system should inherit the simple mechanism of
evolution in which the simple models could produce the evolution of complex
morphologies.
ISICA 2015 featured the most up-to-date research in the analysis and theory of
evolutionary computation, neural network architectures and learning, neuro-dynamics
and neuro-engineering, fuzzy logic and control, collective intelligence and hybrid
systems, deep learning, knowledge discovery, learning, and reasoning. It provided a
venue for fostering technical exchanges, renewing everlasting friendships, and estab-
lishing new connections.
On behalf of the Organizing Committee, we would like to thank warmly the
sponsors, South China Agricultural University, Guangzhou University, Wuhan
University, and China University of Geosciences, who helped in one way or another to
achieve our goals for the conference. We wish to express our appreciation to Springer
for publishing the proceedings of ISICA 2015. We also wish to acknowledge the
dedication and commitment of both the staff at the Springer Beijing office and the CCIS
editorial staff. We would like to thank the authors for submitting their work, as well as
the Program Committee members and reviewers for their enthusiasm, time, and
VI Preface
Program Committee
Shaowei Cai Chinese Academy of Sciences, China
Jiannong Cao The Hong Kong University of Science and
Technology, China
Aniello Castiglione University of Salerno, Italy
Weineng Chen Sun Yat-sen University, China
Yan Chen South China Agricultural University, China
Debiao He Wuhan University, China
Jun He Aberystwyth University, UK
Shuqiang Huang Jinan University, China
Xinyi Huang Fujian Normal University, China
Ying Huang Gannan Normal University, China
Chunfu Jia Nankai University, China
Dazhi Jiang Shantou University, China
Nan Jiang East China Jiao Tong University, China
Jin Li Guangzhou University, China
Kangshun Li South China Agrictural University, China
Ping Li Sun Yat-sen University, China
Wei Li Jiangxi University of Science and Technology, China
Xuan Li Fujian Normal University, China
Zhiqiang Lin Chinese Academy of Sciences, China
Xiaozhang Liu Hainan University, China
Zheli Liu Nankai Universisy, China
Xu Ma Qufu Normal University, China
Wen Sheng Deakin University, Australia
Ke Tang University of Science and Technology of China, China
Ming Tao Dongguan University of Technology, China
Xiang Tao Chongqing University, China
Cong Wang City University of Hong Kong, China
Jiahai Wang Sun Yat-sen University, China
Jianfeng Wang Xidian University, China
Yilei Wang Ludong University, China
Yong Wang Central South University, China
Xianglin Wei Nanjing Telecommunication Technology Research
Institute, China
Di Wu Sun Yat-sen University, China
Lu Xiong Jiangxi University of Science and Technology, China
Honyang Yan Guangzhou University, China
VIII Organization
Additional Reviewers
Chen, Peng
Chen, Yan
Li, Wei
Ye, Jun
Zeng, Ling
Contents
Evolutionary Algorithms
A New Trend Peak Algorithm with X-ray Image for Wheel Hubs Detection
and Recognition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Wei Li, Kangshun Li, Ying Huang, and Xiaoyang Deng
Image Feature Extract and Performance Analysis Based on Slant Transform . . . 489
Jinglong Zuo, Delong Cui, Hui Yu, and Qirui Li
1 Introduction
In recent years, some swarm intelligence optimization algorithms have emerged, such as
genetic algorithm (GA) [1], simulated annealing algorithm (SA) [2], particle swarm
optimization algorithm (PSO) [3], ant colony algorithm (ACO) [4], differential evolution
algorithm (DE) [5] and so on. These optimization models are all based on group search,
and they have made great progress on the theoretical and applied aspects recently.
In 2006, Sheldon et al. [6] proposed a new algorithm of group intelligence, that is
group search optimization (GSO). The GSO is a novel random group search algorithm
based on producer-scrounger model, simulated the behavior of searching resources for
animals. Some investigations have showed that the GSO has better accuracy and
convergence speed compared to PSO and some other evolutionary algorithms (EAs).
Moreover, the GSO not only employs the simplicity and easy implementation com-
pared to other EAs, but also it uses a particular search model. These characteristics
enable the GSO to avoid some unnecessary structure analysis, thereby saving a lot of
time, and it is especially suitable for the optimization design. However, there exists no
a general optimizer being suitable to solve all kinds of optimization problems due to the
theorem of No Free Lunch. That is to say, optimizers may obtain good performance on
some optimization problems and may behave poor for others.
The standard GSO is suitable for solving the complex structure design problems
with multimodal, high-dimension and varying dimensions, but it still employs some
shortcomings such as premature convergence and poor accuracy of solution. In order to
improve the GSO, a hybrid GSO with opposition-based learning and differential
evolution (OBDGSO) is proposed in the paper. The opposition-based learning method
travels the current solution space and the opposition-based solution space simultane-
ously, and the better solution is found to view as the new solution, which can increase
the probability of finding the global optimum. In general, DE has a strong ability of
global search at the early stage of the algorithm. The reason is that the early behavior of
mutation in DE can obtain remarkable differences between individuals of the popu-
lation. However, the differences between the individuals become insignificant gradually
at the later stage because the population tends to converge with the evolution process,
correspondingly, the ability of local search of DE become stronger and stronger. So, it
can be found that some advantages of DE, such as fast convergence and not easy to fall
into local optimum, are precisely what the GSO does not have. In this paper, the
OBGGSO on the one hand improves the diversity at the later stage of algorithm, and
enhances the information exchange of the above optimizers (for example, DE, OBL
and GSO), which facilitates the OBDGSO to jump out of the local optimum area and
accelerate the convergence speed.
2 Background
(1) the Xp will scan at zero degree and then scan laterally by randomly sampling three
points in the scanning area. The way of calculation for the point at zero degree is
as follow:
The point in the right hand side hypercube follows formula (3):
and the point in the left hand side hypercube abides by Eq. (4):
/t þ a ¼ /t ð6Þ
where a is a constant.
At each iteration, a number of group members are selected as scroungers. The
scroungers will keep searching for opportunities to join the resources found by the
producer. At the tth iteration, the area copying behavior of the ith scrounger can be
modeled as a random walk towards the producer followed the Eq. (7)
where amax is the maximum turning angle; and the ith member will choose a random
distance as Eq. (9)
~x ¼ a þ b x ð11Þ
~x ¼ 1 x ð12Þ
~x ¼ ai þ bi xi ; i ¼ 1; 2; . . .; n ð13Þ
where rand(0,1) is a uniform random number within [0,1], and CR is the crossover
constant with outcome 2 [0,1].
3) Selection operation
DE uses a “greedy” selection strategy to ensure the better individual having better
fitness to enter into the next generation. The trial individual will be compared to
parent individuals after the process of mutation and crossover. If the fitness of the
trial individual is better than the parent, it will replace the parent and join into the
next population. Otherwise, the parent remains unchanged and enters into the next
iteration directly.
8 C. Xie et al.
3 Hybrid GSO
A lot of investigations have showed that the shortcomings of GSO, such as poor
convergence speed and easily trapping into local optimum, could attribute to the loss of
diversity of group gradually. So, it is crucial to GSO to improve the diversity to
enhance the efficiency of GSO to solve the complex optimization problems.
As mentioned in Sect. 1, the main merit of DE contains the strong local search
ability, and the method of opposition-based learning is good at global exploration.
Considering the above factors, we combine DE and OBL into GSO to design a hybrid
GSO optimizer.
Let the population size be N, and select 30 percents of the population randomly to
carry out OBL operator to generate a new opposition-based population. Combining the
new population and the original population (whose size is 0.3*N) to sort in descending
order based on the fitness, then the better half of the mixed population would be
selected. Afterwards, the number of 0.4*N of individuals from the rest of the popu-
lation are selected to perform DE operation. At last, the remainder population, whose
size is 0.3*N, will be carried out GSO optimization. The hybrid GSO integrates the
advantages of DE and OBL to better balance the global exploration and local
exploitation to solve the complicated optimization problems effectively.
The flowchart of the hybrid GSO in the paper is presented as follows.
Step 1. Initialize the population P randomly, let the population size be N and the
maximum iteration number be Tmax. Calculate the fitness of each individual, and set
the counter of iteration t = 0.
Step 2. Random select 0.3*N individuals to form a subpopulation SP1, applying
OBL to SP1 to generate an opposition-based population OBP. Combined SP1 and
OBP to select the better half of the mixing population to form subpopulation P1
based on the fitness values.
Step 3. Random select 0.4*N individuals to construct a subpopulation SP2, and
apply DE to SP2 to generate a differential evolution population P2, whose size is
also 0.4*N.
Step 4. Apply GSO to the remainder population whose size is 0.3*N to generate a
population P3, and | P3| = 0.3*N.
Step 5. Combine the subpopulation P1, P2 and P3 to form the next population, and
t = t+1.
Step 6. If t > Tmax, stop; otherwise, goto Step 2.
In order to test the validity of the hybrid GSO, we select two representative optimizers
as peer comparison algorithms, the one is GSO proposed by He S [13]. in 2009, and the
other is the original DE proposed by Storn et al. The comparable experiments in the
paper are all based on 13 benchmark single objective optimization problems [14]. And
the 13 test problems are listed in Table 1. These 13 benchmark functions can be classify
into three kinds as follows. (1) unimodal functions. (2) simple multi-peak functions.
A Hybrid Group Search Optimizer 9
Table 2. The average results for each algorithm on each test function
Function GSO DE OBDGSO fmin
f1 21.2789 0.1803 9.12E-07 0
f2 0.1779 1.0614 7.20E-08 0
f3 4.63E+04 2.19E+03 1.41E+04 0
f4 73.1417 4.5009 62.4728 0
f5’ 4.48E+07 6.14E+06 3.61E+06 0
f6’ 1.31E+03 77.4618 22.3586 0
f7 3.23E-08 1.40E-06 1.49E-11 0
f8 3.91E+03 7.64E+03 3.71E+03 0
f9 1.05E+02 2.18E+02 18.3557 0
f10 11.4299 10.7654 6.0519 0
f11 1.2249 0.3278 3.08E-06 0
f12’ 1.79E+06 6.1801 4.2264 0
f13’ 3.90E+06 18.4162 8.4528 0
The above two types of functions are mainly used to test the optimization accuracy.
(3) non-rotating multi-peak functions. This kind of functions have many local extreme
points, which is hard for the average optimizers to find the global optimum. So, the
third type of test functions are often used to test the global optimization performance
and the ability of avoiding premature convergence.
All experiments in the paper are based on Windows 7 operating system, dual-core
2.50 GHz Intel processor and 4G memory, and Matlab 2010 programming platform.
In order to compare the performances of peer algorithms, we set some identical
running parameters, such as the group (population) size is 100, the dimension of
decision variable is 30, and the maximum number of iterations is set to 1000.
10 C. Xie et al.
Table 3. Statistical mean and standard deviation of obtained by OBDGSO, GSO and DE on 13
test functions over 30 independent runs
Test OBDGSO GSO DE
instance
f1 Mean 1.00E-09 8.1584 0.1432
Std. 1.70786E-06 18.99110904 0.032472599
t-test + +
f2 Mean 7.40E-09 0.0408 0.8419
Std. 4.39996E-08 0.118550676 0.211006374
t-test + +
f3 Mean 9.44E+03 2.96E+04 1.61E+03
Std. 3997.819643 10703.80471 671.7363074
t-test + =
f4 Mean 51.6286 61.9726 3.9978
Std. 8.146587239 7.455841639 0.349443413
t-test + =
f5 Mean 9.41E+05 3.04E+07 5.21E+06
Std. 3270285.171 14123670.35 900927.2766
t-test + +
f6 Mean 5.7405 3.85E+02 50.5079
Std. 11.23317659 964.833203 19.58850474
t-test + +
f7 Mean 8.93E-11 1.00E-10 8.00E-09
Std. 4.36417E-12 5.54543E-08 2.56192E-06
t-test + +
f8 Mean 3.08E+03 3.33E+03 7.33E+03
Std. 387.9434117 387.4388644 555.615874
t-test = +
f9 Mean 0.0029 6.15E+01 2.10E+02
Std. 24.75970942 47.18212305 6.143087229
t-test = =
f10 Mean 2.9365 10.2191 10.3556
Std. 2.376233926 0.664516162 0.203911203
t-test = =
f11 Mean 1.60E-07 1.0465 0.2912
Std. 4.22386E-06 0.229185544 0.066716038
t-test + +
f12’ Mean 2.2658 2.42E+05 5.3495
Std. 1.654364408 1861291.275 1.137839227
t-test + =
f13’ Mean 1.6379 6.09E+05 10.8051
Std. 7.899360368 3198753.277 5.047741332
(Continued)
A Hybrid Group Search Optimizer 11
Table 3. (Continued)
Test OBDGSO GSO DE
instance
t-test + =
Better(+) 10 7
Same(=) 3 6
Worse(−) 0 0
Score 10 7
In addition, the algorithmic parameters of GSO are used followed [15], and the
algorithmic parameters of DE are followed [16] and [17].
Our OBDGSO is compared with the other two optimizers, such as GSO and DE, all
three algorithms are carried out 30 times repeatedly, and we can obtain the statistical
average data as the experimental results, which are listed in Table 2. We can observe
that the OBDGSO has the best accuracy of solution in all 13 test functions among the
three peer algorithms.
Table 3 lists the statistical mean and standard deviation results obtained by
OBDGSO, GSO and DE on 13 test functions over 30 independent runs. It can be seen
that the OBDGSO has significant performance advantages over GSO and DE. So we
can conclude that OBDGSO is promising optimizer in solving multi-modal,
high-dimensional functions.
5 Conclusion
The paper proposed a hybrid GSO with opposition-based learning and differential
evolution, called OBDGSO. The hybrid GSO utilizes the method of opposition-based
learning to enhance the ability of global exploration and uses differential evolution to
improve the local search ability. Three peer comparison algorithms (GSO, DE and
OBDGSO) are performed to comparing experiments on 13test functions, the experi-
mental results show that OBDGSO has significant advantages over GSO and DE in
accuracy of solution and convergence speed. So, the conclusion is that the OBDGSO in
the paper is a promising optimizer in solving multi-modal, high-dimensional functions.
References
1. Holland, J.H.: Adaptation in Natural and Artificial Systems. University of Michigan Press,
AnnArbor (1975)
2. Kirkpatrick, S., Gelatt, C.D., Vecchi, P.M.: Optimization by simulated annealing. Science
220, 671–680 (1983)
3. Kennedy, J., Eberhart, R.C.: Particle swarm optimization. In: Proceedings of IEEE
International Conference on Neural Networks, pp. 1942–1948 (1995)
12 C. Xie et al.
4. Dorigo, M., Maniezzo, V., Colorni, A.: The ant system: optimization by a colony of
cooperating agents. IEEE Trans. Syst. Man Cybern. Part B Cybern. 26, 29–41 (1996)
5. Storn, R., Price, K.: Differential evolution-a simple efficient adaptive scheme for global
optimization. J. Global Optim. 11, 341–359 (1997)
6. He, S., Wu, Q.H., Saunders, J.R.: A novel group search optimizer inspired by animal
behavioural ecology. In: 2006 IEEE Congress on Evolutionary Computation (CEC),
pp. 1272–1278. IEEE Xplore, Vancouver, BC, Canada. New York, 16–21 July 2006
7. He, S., Wu, Q.H., Saunders, J.R.: Group search optimizer:an optimization algorithm inspired
by animal searching behavior. IEEE Trans. Evol. Comput. 13(5), 973–990 (2009)
8. Rahnamayan, S., Tizhoosh, H.R., Salama, M.M.A.: Opposition-based differential evolution
algorithms. In: IEEE Congress on Evolutionary Computation Canada (2006)
9. Giraldeau, L.-A., Lefebvre, L.: Exchangeable producer and scrounger roles in a captive flock
of feral pigeons - a case for the skill pool effect. Anim. Behav. 34(3), 797–803 (1986)
10. Storn, R., Price, K.: Differential evolution-a simple and efficient heuristic for global
optimization over continuous spaces. J. Glob. Optim. 11, 341–359 (1997)
11. Yuan, J., Sun, Z., Qu, G.: Simulation study of differential evolution. J. Syst. Simul. 20,
4646–4647 (2007)
12. Wang, F.-S., Jang, H.-J.: Parameter estimation of a bioreaction model by hybrid differential
evolution. Evol. Comput. 1, 16–19 (2000)
13. He, S., Wu, Q.H., Saunders, J.R.: Group search optimizer: an opimization algorithm inspired
by animal searching behavior. IEEE Trans. Evol. Comput. 13(5), 973–990 (2009)
14. Yao, X., Liu, Y., Liu, G.: Evolutionary programming made faster. IEEE Trans. Evol.
Comput. 3(2), 82–102 (1999)
15. Karaboga, D., Akay, B.: A comparative study of artificial bee colony algorithm. Appl. Math.
Comput. 214(1), 108–132 (2009)
16. Qin, A.K., Huang, V.L., Suganthan, P.N.: Differential evolution algorithm with strategy
adaptation for global numerical optimization. IEEE Trans. Evol. Comput. 13(2), 398–417
(2009)
17. Liu, G., Li, Y., Zhang, Q.: Enhancing the search ability of differential evolution through
orthogonal crossover. Inf. Sci. 185(1), 153–177 (2012)
A New Firefly Algorithm with Local Search
for Numerical Optimization
1 Introduction
modified FA performs better than the standard FA in finding the best solution with
smaller CPU time. Fister et al. [9] proposed a memetic FA (MFA) to solve combi-
natorial optimization problems. In MFA, the parameter a is dynamically adjusted, and
the parameter b is changed in the range [0.2, 1.0] based on the distance between two
fireflies. Additionally, the random part ae for the movement of the attraction is scaled
by the size of the search range. Experimental results show that the MFA is significantly
better than the standard FA. In our previous work [10], the MFA is used as the standard
FA and combined with other strategies. Gandomi et al. [11] introduced chaos into FA
to increase its global search ability for robust global optimization. Different chaotic
maps are utilized to tune the attractive movement of fireflies. Results show that the
chaotic FA (CFA) outperforms the standard FA. In [12], quaternion is used for the
representation of individuals in FA so as to enhance the performance of the firefly
algorithm and to avoid any stagnation. Yu et al. [13] designed a new FA with a wise
step strategy (WSSFA), which considers the information of firefly’s personal and the
global best positions. Results show that the modified algorithm outperforms the stan-
dard FA on twenty benchmark functions. In [14], a variable step size FA (VSSFA) is
proposed, where a dynamical method is used to update the parameter a. Computational
results show that WSSFA and VSSFA achieve better solutions than the standard FA on
a set of low-dimensional benchmark functions (D = 2). However, our experiments
demonstrate that both of them can hardly obtain reasonable solutions for some
high-dimensional problems (D = 30). Compared to WSSFA and VSSFA, MFA can
achieve promising solutions.
In the FA, the fitness function for a given problem is associated with the light
intensity. The brighter the firefly is, the better the firefly is. That means a brighter firefly
has a better fitness value. The search process of FA depends on the attractions between
fireflies. Based on these attractions, a firefly tends to move other brighter fireflies. If a
firefly is brighter than another one, the brighter firefly will not be conducted any search.
In this paper, we propose a new FA (called NFA) to avoid this case. When the current
firefly is brighter than another one, a local search operation is conducted on the current
one to provide more chances of finding more accurate solutions. It is noted that the
proposed NFA is implemented based on the MFA. Therefore, the NFA is a hybrid
algorithm by combining the MFA and the proposed local strategy. To verify the
performance of NFA, a set of well-known benchmark function with D = 30 are tested.
Experimental results show that NFA performs better than the standard FA, MFA, and
VSSFA.
The rest paper is organized as follows. In Sect. 2, the standard FA is briefly
introduced. In Sect. 3, the proposed NFA is described. Experimental results are pre-
sented in Sect. 4. Finally, the work is concluded in Sect. 5.
2 Firefly Algorithm
As mentioned before, the FA mimics the behavior of the social behavior of the flashing
characteristics of fireflies. To simply the behavior of fireflies and construct the search
mode of FA, three rules are used as follows [1]:
Another random document with
no related content on Scribd:
peptic ferment in, 16
Foam structure, 6
Folliculina, 137;
tube of, 152
Food, 35 f. (see also Ingestion);
of Higher Animals, 38;
absorption of, by Plants, 38;
in relation to life-cycle of Ciliata, 147 f.;
of Sponges, 237;
of Hydra, 256 and n.;
of Millepora, 261;
of Siphonophora, 304;
of Charybdea, 319;
of Alcyonium, 339;
of Zoantharia, 373;
of Asterias rubens, 439;
of Ophiothrix fragilis, 486;
of Ophiolepididae, 496;
of Echinus esculentus, 516;
of Echinarachnius parma, 546;
of Echinocardium cordatum, 552;
of Holothuria nigra, 561;
of Dendrochirota, 572;
of Synapta inhaerens, 577;
of Antedon rosacea, 583
Food-vacuole, of Actinosphaerium eichornii, 72;
of Ciliata, 145 f.;
of Carchesium, 146
Foot-plate, of young Pentacrinidae, 592;
of larva of Antedon rosacea, 619
Foraminifera, 40, 49, 50, 58 f.;
relations of, 49;
shell of, 49, 59 f., 60, 61, 63, 65;
habitat of, 59 f.;
literature of, 58 n.;
marine, 60 f.;
nuclei of, 62, 67 f.;
nutrition of, 40, 62;
streaming of granules in, 17;
chromidia of, 62, 68 f.;
collection of, 62;
dimorphism of, 66, 67;
reproduction of, 67 f.;
economic uses of, 69 f.;
palaeontology of, 69 f.
Forbes, 338
Force, dual, of dividing-cell, 26 n.
Forcepia, 223
Forceps, 222
Forcipulata, 462, 473 f.
Forcipulate pedicellaria, 456, 473
Formative vacuole of contractile vacuole, in Flagellata, 110, 115;
in Ciliata, 143
Fossil, Foraminifera, 69 f.;
Radiolaria, 87 f.;
Dinoflagellata, 132;
Peridinium, 132;
Sponges, 192 f., 207 f., 215, 241;
Coelenterates, 270, 281 f., 343 f., 346, 393 f., 406;
Asteroidea, 475 f.;
Ophiuroidea, 501 f.;
Echinoidea, 556 f.;
Crinoidea, 594 f.;
Thecoidea, 596;
Carpoidea, 596 f.;
Cystoidea, 597 f.;
Blastoidea, 599 f.
Fowler, 293 n., 382, 400, 404
Framboesia, 121 n.
France, epidemic of pébrine in, 107
Francé, on structure of funnel of Choanoflagellates, 115 n., 121 n.;
monograph of Choanoflagellates, 123, 182 n.;
on Polytomeae, 119 n.
Freetown, prophylaxis of malaria at, 106
Fringing reef, 390 f.
Frog's blood, Lankesterella in, 102
Frondicularia, 59, 63
Fructification, of Mycetozoa, 90 f.;
of Acrasieae, 90;
of Myxomycetes, 49, 91 f.
Fry, E. and A., on Myxomycetes, 93 n.
Fuligo, 90;
F. varians, 92 f.;
pepsin in, 16
Fungacea, 402
Fungi, cell connexions in, 37 f.;
in relation to Protista, 40;
Gasteromycetous, 91
Fungia, 403;
asexual reproduction of, 388, 389;
F. crassitentaculata, 403
Fungiidae, 403
Funiculina, 359, 362;
F. quadrangularis, 362
Funiculinidae, 362
Funnel, of Craspedomonadidae or Choanoflagellates, 111, 121, 122,
182;
of Phalausteridae, 111;
of choanocytes of Sponges, 171
Fusion of larval Sponges, 174
Fusion-nucleus of Ciliata, 150
—see also Reproduction, Syngamy, Zygotonucleus
Fusulina, 59
Galaxea, 400;
G. esperi, 400
Galeolaria, 307;
G. biloba, 304
Galeolarinae, 307
Galerites, 558
Gamble, 312 n.;
and Keeble, 175 n.
Gametes, 33 f.;
of Trichosphaerium, 54;
of certain Protomastigaceae, 116 n.;
of Volvocidae, 127 f.;
of Pandorina (of three sizes), 128, 129
Gametocyte of Acystosporidae, 104 f.
Gametogonium (= parent-cell of gametes), male, of Acystosporidae,
105
Gametonuclei (= nuclei capable of syngamous fusion), 34
Ganeria, 464
Ganeriidae, 454, 464
Gardiner, 345, 370, 375, 392 n., 404
Garveia, 270
Gasteromycetous fungi, 91
Gastral layer, 171
Gastralia, 201
Gastropores, 257, 258
Gastrozooids, of Millepora, 259, 260;
of Hydractinia, 264;
of Siphonophora, 299;
of Antipatharia, 408
Gastrula, definition of, 603
Gaule, misinterpretation of nature of Haemosporidae, 102
Gegenbaur, 302
Gellius 217, 223;
G. varius, development, 172 f., 173, 174
Gemmantes, 400
Gemmaria, 405
Gemination = Budding, q.v.
Gemmiform, pedicellariae, of Echinus esculentus, 506;
of E. acutus, 509;
of E. elegans, 510;
of Cidaridae, 534;
of Echinarachnius parma, 544;
of Echinocardium cordatum, 550
Gemmule, 177, 178, 179, 230
Generation, spontaneous, 42 f.
Generations, alternation of, 44, 250
Genital base of Holothuria nigra, 567
Genital bursa, of Ophiothrix fragilis, 485
compared with hydrospires of Blastoidea, 600
Genital canal of Antedon rosacea, 586
Genital organs (including ducts), of Asterias rubens, 451 f.;
of Ophiothrix fragilis, 490;
of Ophiarachna, 491;
of Ophiuroidea, 494;
of Amphiura squamata, 494;
of Echinus esculentus, 528;
of Echinocardium cordatum, 552;
of Hemiaster philippi, 552;
of Holothuria nigra, 567;
of Antedon rosacea, 586
Genital plate, of Ophiothrix fragilis, 485;
of Echinus esculentus, 512, 513
Genital rachis, of Asterias rubens, 452;
of Ophiothrix fragilis, 490;
of Echinus esculentus, 528;
of Antedon rosacea, 586
Genital scale of Ophiothrix fragilis, 485
Genital stolon, of Asterias rubens, 451;
of Ophiothrix fragilis, 489;
of Echinus esculentus, 528;
of Holothuria nigra, 567;
of Antedon rosacea, 585, 586;
of larva of A. rosacea, 619
Geodia, 211
Geographical distribution of Protozoa, 47
Geotaxy (= barotaxy), 20
Gephyra dohrnii, 382, 408
Gephyrea, 577
Gerarde, 167
Gerardia savalia, 406
Gerbillus indicus infested by a Haemosporidian, 102 n.
Germinal spot (= nucleole of ovum), 7
Germinal vesicle (= nucleus of ovum), 7
Germination, 32;
of Myxosporidian spores, 107
Germ-plasm, 28 f.;
continuity of, 172
Germ theory, 44
Germs, invisible air-borne, 43
Geryonia, 290, 295
Gilchrist, 338
Gill of Echinus, 514, 527
Gill-cleft, of Echinus, 514;
of Sphaerechinus, 540 f.;
of Strongylocentrotus, 541
Ginkgo, spermatozoa of, 38
Glaucoma, 137, 153;
G. scintillans, rate of fission of, 147 f.
Glauconite, 70
Globiceps, 272
Globiferae of Centrostephanus longispinosus, 532
Globigerina, 59, 63, 242;
spines of, 61, 66;
-ooze, 61 f.;
G. bulloides, 68, 69
Glossina morsitans, intermediate host of Trypanosoma brucei, 119;
G. palpalis, intermediate host of T. gambiense, 120
Glossograptus, 282
Glycerin, 15
Glycogen, of Ciliata, 144;
-vesicles of Pelomyxa palustris, 53
Gnat (Anopheles), intermediate host of Haemamoeba and
Laverania, 103 f.;
(Culex) intermediate host of Haemoproteus, 103;
of Trypanosoma, 120
Golgi, on relation of Acystosporidian life-cycle and stages of
intermittent fever, 103
Gonactinia, 371, 372, 377
Gonangium, 276
Goniaster, 471;
fossil, 475
Goniastraea, 375, 401
Goniocidaris, 534;
G. canaliculata, 535
Gonionema, 288, 291;
G. murbachii, 232, 290, 291, 292
Gonium, 111
Gonophore, of Gymnoblastea, 265;
of Calyptoblastea, 277;
of Stylasterina, 284;
of Siphonophora, 302
Gonotheca, 276, 281
Gonozooids, of Siphonophora, 302;
of Antipatharia, 408
Gorgonacea, 350 f.
Gorgonella, 357;
spicule, 336
Gorgonellidae, 337, 357
Gorgonia, 356;
G. cavolinii, 340;
G. flabellum, 357;
G. verrucosa, 356
Gorgoniidae, 334, 337, 356
Gorgonocephalus, 491, 501
Gosse, 273
Goto, 291, 293;
on development of Bipinnaria, 612
Grammaria, 278
Granatocrinus, 599;
G. norwoodi, 600
Grant, 167
Grantiidae, 192
Grantiopsis, 191
Granular disintegration of Protista, 14 f.
Granules, in protoplasm, 6;
excretory, 6, 144;
aleurone, 37;
basal, of cilia, etc., 138 n., 141 (see also Blepharoplast);
proteid, of Suctoria, 161
Graphiohexaster, 203
Graptolitoidea, 281
Grassi, on malarial parasites, 103
Gravity, stimulus of, 19 f.
Greasy film, outer clear layer of protoplasm behaves like, 17
Greeff, on Protozoa, 46
Green Flagellates, relations of, 48
Greensand, 70;
Cambridge, 208
Green water often due to Euglena viridis, 124
Greenwood, M., on peptic digestion in Protozoa, 16;
on feeding of Carchesium polypinum, 45 f., 146 f.
Gregarina, 97, 98, 99;
G. blattarum, 98
Gregarines, habitat, 99;
syngamy, 99
Gregarinidaceae, 95 f., 97 f.
Gregory, 346
Grew, 166
Grey chalk, 61
Gromia, 52;
G. oviformis, 59 n.
—see also Allogromia
Grooves, longitudinal and transverse, of Dinoflagellata, 110, 130,
131, 132;
of Peridinium, 131;
of Polykrikos, 132;
oral, of Noctiluca, 133
Grosvenor, 249 n.
Growth, 19 f.;
Spencer's limit of, 23, 31
Gruber, on regeneration in Protozoa, 35 n.;
on diffused nucleus in marine Ciliata, 144 n.;
on tubicolous marine Ciliata, 152
Gruppe, deposit of Radiolaria, 87
Guinea Coast, 106
Gullet (= pharynx) of Paramecium caudatum, 151
Gut, supposed, of Ciliata, 145
—see also Alimentary canal
Gutter, oral, of Vorticellidae, 156, 158
Gymnamoebae, 51 n.
Gymnasteridae, 471
Gymnoblastea, 262 f.
Gymnodinium, 110;
G. pulvisculus, parasitic in Appendicularia, 132
Gymnomyxa, 49 n.
Gymnophrys, 58
Gymnosphaera, 70, 73
Gymnostomaceae, 137;
predaceous, trichocysts of, 143;
mouth and pharynx of, 145;
noteworthy members of, 152
Gyractis, 380
Ianthella, 220
Ichthyophtheirius, 137;
noxious parasite of fish, 152
Iciligorgia, 351
Idioplasm, 29
Ijima, 199, 206, 231, 234
Ileonema, 137, 152
Ilyanthus mitchellii, 380
Ilyodaemon, 571, 572;
I. maculatus, 571
Imperforate, Foraminifera, 58 f.;
Corals, 371
Inadunata, 595
Incurrent canal, 170
India, diseases of Trypanosomic origin, 119 f.
Induction shocks, action on Protozoa, 7, 22
Infero-marginal ossicle of Asteroidea, 436
Inflammation, 8
Infra-basal plate, of Crinoidea, 588;
of fossil Crinoidea, 594;
of larval Antedon rosacea, 619
Infundibulum, 415
Infusions, appearance of organisms in, 42 f.;
organisms of, 136
Infusoria, 40, 48, 50, 136 f.;
specific gravity of, 13 n.;
zygote does not encyst, 34.
Ingestion, of food, by Amoeba limax, 9;
by Choanoflagellates, 122;
by Dinoflagellates, 131;
by Carchesium, 146;
by Coleps, 150
—vacuole of, in Flagellates, 113;
in Oikomonas, 112;
in Choanoflagellates, 122
Inner perihaemal ring-canal, of Asterias rubens, 448;
development of, in Asterina gibbosa, 612
Inoculation of malarial fever in man through a mosquito, 105 f.
Insectivorous plants, 38
Insects, metamorphoses of, 44;
as hosts of Trichonymphidae, 123
Interambulacral area, of Echinarachnius parma, 544;
of Echinocardium cordatum, 550
Interambulacral plate, of Echinus esculentus, 511;
of Cidaridae, 533 f.;
of Echinarachnius parma, 544 f.
Interbrachial septa—see Interradial septa
Interchanges between cell and medium, 14
Intermediate dorsal process of ciliated band of Auricularia, 608
Intermediate (= supplemental) skeleton of Perforate Foraminiferal
shell, 63, 66
Intermittent fever, malarial, produced by Acystosporidae, 103 f.
Internal budding of Suctoria, 160 f., 162;
of Ephydatia, 177
Internal gills—see Stewart's organs
Internal movements of protoplasm, 17
Interradial plates, of calcareous ring of Holothuria nigra, 566;
of Holothuroidea, 569;
of Synaptida, 569;
of Dendrochirota, 569;
of calyx of Crinoidea, 589;
of Thaumatocrinus, 589;
of Hyocrinus, 590;
of Rhizocrinidae, 591;
of corona of Echinoidea—see Interambulacral plate
Interradial septa, of Asterias rubens, 437;
of Heliasteridae, 474;
absent in Brisingidae, 475
Interradius, 428;
of Asterias rubens, 434;
of Echinus esculentus, 504;
of Holothuria nigra, 562
Interstitial growth, 10
Intestine, 415;
of Echinus esculentus, 516;
of Holothuria nigra, 563;
of Antedon rosacea, 583;
of Actinometra, 589;
of Dipleurula, 605;
of Protocoelomata, 616
Intracapsular protoplasm of Radiolaria, 80 f.
Intramolecular respiration, 14 n.
Intranuclear spindle of Euylypha, 29
Invertebrata, hosts of Gregarines, 97 f.
Iodine, 239
Iophon, 223
Iridogorgia, 355
Isaurus, 405
Ischadites, 207
Ischikawa, on syngamy of Cystoflagellates, 135;
on structure of Ephelota, 162
Isidae, 337, 353
Isidella, 354
Isis, 353
Ismailia, prophylaxis of malaria at, 106
Isochela (a chela divisible by each of two planes into two equal
parts, the two ends being equally developed), 222
Isocrinus—see Pentacrinus
Isogamy, 33 f.;
of Rhizopoda, 56 f.;
of Stephanosphaera, 128
—see also Syngamy
Isospores, 85;
of Radiolaria, 76;
of Collozoum inerme, 76
Karyogamy, 34 n.
—see also Syngamy of Ciliata
Karyokinesis, 25, 26, 27;
function of, 28 f.;
of micronuclei of Ciliata, 144 f.
—see also Mitosis
Karyolysus, 97
Karyosome, 24
Keeble, 175 n.
Keller, 233
Kemna, on stylopodium of Foraminifera, 60
Kent, Saville, on Choanoflagellates, 122 f., 182;
on Infusoria and Flagellates, 136 n.
Keroeides, 351
Kerona, 138;
K. polyporum, 158 n.
Kieselguhr, 87
Kirkpatrick, 215
Kishinouye, 313 n., 321 n., 333, 352
Klebs, on Flagellates, 119;
on Dinoflagellates, 130
Koch, von, on methods of cultivation of lower organisms, 44;
on malarial parasites, 103
Kölliker, on Sporozoa, 94 f.
Kophobelemnon, 362
Kophobelemnonidae, 362
Köppen, on Sticholonche and its parasite, Amoebophrya, 87 n.
Korethraster, 453, 463
Kowalevsky, 341, 422
Krukenberg, on pepsin in a Myxomycete, 16
Kükenthal, 363