Bulletin STATE OF ILLINOIS HENRY HORNER, Gosenwr DEPARTMENT OF REGISTRATION AND EDUCATION DIVISION OF THE NATURAL HISTORY SURVEY THEODORE H. PRISON. Chief Volume 21 BULLETIN Article 2 Responses of the Large-mouth Black Bass to Colors FRANK A. BROWN, JR. PRINTED BY AUTHORITY OF THE STATE OF ILLINOIS URBANA, ILLINOIS i\Iav 1937 STATE OF ILLINOIS Henry Horner, Governor DEPARTMENT OF REGISTRATION AND EDUCATION John J. Hallihan, Director BOARD OF NATURAL RESOURCES AND CONSERVATION John J. Hallihan, Chairman William Trelease, D.Sc, LL.D., Biology William A. NoyeS; Ph.D., LL.D., Henry C. Cowle.s, Ph.D., D.Sc, Forestry Chem.D., D.Sc, Chemistry John W. Alvord, C.E., Engineering Edson C. Bastin, Ph.D., Geology Arthur Cutts Willard, D.Eng., LL.D., President of the University of Illinois NATURAL HISTORY SURVEY DIVISION URBANA, ILLINOIS Scientific and Technical Staff Theodore H. Prison, Ph.D., Chief SECTION OF ECONOMIC ENTOMOLOGY W. p. Flint, B.S., Chief Entomologist C. C. CoMPTON, M.S., Associate Ento- mologist M. D. Farrar, Ph.D., Research Ento- mologist S. C. Chandler, B.S., Southern Field Entomologist J. H. Bigger, B.S., Central Field Ento- mologist L. H. Shropshire, M.S., Northern Field Entomologist W. E. McCauley, M.S., Assistant Entomologist C. W. Kearns, Ph.D., Research Fellow in Entomology ]. F. Alsterlund, M.S., Research Fellow in Entomology DwiGHT Powell, M.S., Research Fellow in Entomology L. R. Tehon, Ph.D., Botanist J. C. Carter, Ph.D., Assistant Botanist G. H. Boewe, M.S., Field Botanist SECTION OF FORESTRY James E. Davis, M.F., Extension For- ester SECTION OF AQUATIC BIOLOGY David H. Thompson, Ph.D., Zoologist F. D. Hunt, Field Naturalist D. F. Hansen, A.M., Assistant Zoologist D. J. O'Donnell, M.S., Assistant Zo- ologist SECTION OF INSECT SURVEY H. H. Ross, Ph.D., Systematic Ento- mologist Carl O. Mohr, Ph.D., Associate Ento- mologist, Artist B. D. Burks, M.A., Assistant Ento- mologist R. E. Yeatter, Ph.D., Game Specialist W. H. Leigh, M.A., Assistant Zoologist SECTION OF PUBLICATIONS Carroll Chouinard, M.A., Editor This paper is a t xtribulionfrom the Sectii (2S1S4—1200—3-37) I of .•igttatic Biology Contents Problem of Color Vision in Fishes 33 Survey of literature. Significance of determining color vision in black bass. Acknowledgments. Materials for the Experiments 35 Fish. Apparatus. Colored pipettes. Black, gray and white pipettes. Color rods. Training and Responses of Large-mouth Black Bass 37 Method. Reactions of untrained fish to colors and grays. Training to each ot tour colors. Responses of fish trained to four colors. Training to each of four colors plus tints. Responses ot fish trained to four colors plus tints. Interpretation of the Responses 50 Summary 53 Bibliography 54 ^,/f?f^ ILLINOIS NATURAL HISTORY SURVEY BULLETIN VOL. 21, ART. 2 MAY 1937 Responses of the Large-mouth Black Bass* to Colors FRANK A. BROWN, JR.^ PROBLEM OF COLOR VISION IN FISHES ECAUSE an excellent review of havior is in response to wave length and not intensity, then the fish may be spoken ot as having color vision for that particular wave length of light. Those few who are not contented with such a broad interpretation of color vision must certainly be content with the re- sults derived from the formation of association between colors and foods. It is generally accepted that the form and function of animals are primarily concerned with feeding, protection of self, and reproduction. The methods of training fishes used by most investiga- tors utilize the first, second, or both of these. Thus we may reason that any differential responses to wave lengths of light which involve such fundamental factors as these, together with modifica- tion of behavior in the face of a new situation, will concern the most highly evolved functional regions of the central nervous system. Even here, however, we need not consider such speculative aspects as the degree of consciousness of the fish with respect to the stimulus quality, or whether light of any particu- lar wave length produces the same sen- sation in a fish as in man. Graber (1884, 1885), utilizing the phototactic responses ot the fishes, de- termined that Corbitns barbattda, Albiir- nus spectabilis, Gasterostcus spinchia, and Syngnalhus acus selected areas illumi- nated by light of one color in preference to areas lighted by another color. Bauer (1910, 1911) similarly used a B color vision in fishes has been written recently by Warner (1931), a detailed historical introduction here is considered unnecessary. A brief statement of the work done in this field, however, is included to permit a clearer understanding ot the results and in- terpretations that are to be found in this report. Survey of Literature The problem of color vision in fishes was one which rather early attracted the attention ot zoologists. Four general modes ot attack upon this problem have been used, that of colored light prefer- ence, that ot choice of variously colored foods, that ot the responses of the in- tegumentary pigmentary system to colored backgrounds, and finally, that of training fishes to discriminate among colors through the formation of asso- ciations. These methods are listed in the order in which they have received the most of their popularity, the oldest experiments being almost exclusively of the preference type while the latest ones are almost always of the association type. Thus tar nearly all zoologists are agreed that if a fish responds in a specific manner to light of a particular color in contrast to light of all other colors and intensities of white, i.e., if it can be established that the difference of be- ^ A pities salmoides (Lacepede). sAssistant zoologist. Illinois Natural Historv SURVEV. July-August 1935. 34 ILLINOIS NATURAL HISTORY SURVEY BULLETIN Vol. 21,Art. 2 preference method with the fishes Charaz piintazzo, Athcrina hepsetns. Box salpa and Mugil sp., and came to the con- clusion that these fishes likewise had a color vision. On the other hand, Hess (1910, 1911, 1913, 1914) published a series of papers upon color vision in a number of kinds of young fishes. He too used chiefly the preference method. His interpretation of his experiments in- dicated that fish vision was approximate- ly that of a color-blind man, and that the apparent ciifferences in response to different colors were based solely upon differences in intensities. Hess severely criticised the work of Graber and Bauer in not having eliminated the possibility that the effects they obtained were due to brightnesses. Upon similar grounds Hess discredited as evidence establishing color vision the observations of Zolotnitsky (1901) and Reighard (1908). Zolotnitsky had found that fish that had been fed for some time upon red chironomid larvae would snap only at red pieces of yarn when a number of variously colored bits of yarn were attached to the outside of the glass container. Reighard (1908) had noted that Lutianus griseus selected baits of one color in preference to baits of another. Sumner (1911) and Mast (1916), working upon the adaptations of flat- fishes to colored backgrounds, arrived at the conclusion that their color vision was much like that of man. Haempel & Kolmer (1914), experimenting with Cottus gobio and Phoxinus laevis, found that these fishes became reddish when upon a red background. Connolly (1925) showed that Fundidus hetero- ditus would modify its pigmentary sys- tem in a different fashion in red and yellow lights than in blue light. That color change in fishes in response to colored backgrounds proves color vision to exist has been seriously questioned by Schnurmann (1920). He shows that the color adaptations in the minnow Phoxinus appear to be the reactions of a color-blind fish with approximately the characteristics of vision of a color-blind man looking through a yellow-tinted screen, h. screen of this sort would absorb the short wave lengths much more than the long ones and thereby the differential response may result from intensity. A rather conclusive work upon the color vision of fishes was performed by von Frisch (1913) who trained Phoxinus laevis to respond positively to a particu- lar color and then had the fish select that color out of a long series of grays. He also demonstrated that red, yellow and purple-red were much alike to the fish. Green and blue were distinguish- able from one another and from red and yellow. Burkamp (1923), using in principle the same kind of technic, arrived at essentially similar conclusions for Phoxinus laevis, Rhodeus amarus, Idus melanotus and Tinea vulgaris. Both von Frisch and Burkamp con- trolled the intensities of their colors principally in using a confusing series of grays. The latter worker also used colors diluted with black and white. Washburn & Bentley (1906) trained Semotilus atromaculatus to accept food from a red pair of forceps and to refuse it when offered from green. An attempt was made to control the intensity by reversing the order of brightness of the colors as they appeared to the human eye. White (1919) and Hineline (1927), employing color filters, showed that Umbra limi and Eucalia appear to distinguish between all colors with the exception of the combination green and blue. The latter work has also been criticized in that it has not sufficient control of the intensity factor. Reeves (1919) was the only one to equalize two colors in the brightnesses that they appeared to the fishes them- selves and then to have the fishes dis- tinguish between them. Miss Reeves has demonstrated quite conclusively \ that Semotilus atromaculatus and Eupo- motis gibbosus distinguish red from blue by virtue of the color values. Schiemenz (1924) used a method in which he trained Phoxinus to respond positively to a white enameled stick upon which was shone light of a par- ticular wave length and to refuse to respond to other colors. These colors were obtained by screening off small portions of a spectrum obtained with a prism. The intensity factor was con- trolled by varying the variously colored Mf 1937 BROWN: COLOR RESPONSE OF LA RGE - MOUTH BASS 35 lights through a broad range of in- tensities. \VolfF(1926) permitted trained fish to select their training color from a series of spectral colors presented simultaneously. These two researches have demonstrated quite conclusively that Phoxinus laevis is able to distinguish among about 20 colors of the visible spectrum and ultraviolet by virtue of color as opposed to brightness. These last two are perhaps the most complete experiments which have been carried out in the field of color vision of fishes. A critical summary of the field now indicates that the minnow Phoxinus laevis is able to distinguish about 20 colors from one another by virtue of the wave length of the light. Furthermore, it appears to be proved that Seuwtilus atromaculatus and Eupomolis gibbosiis are able to distinguish red from blue solely by differential effects of the colors as distinct from intensity. Be- yond these statements all we can say is that it is quite probable that about 15 or 20 more genera of fishes also have a kind ot color vision although this has not been decisively proved. Significance of Determining Color Vision in Black Bass In view of the small amount of con- clusive data upon the color vision of fishes and the fact that the best of what there is has been for all practical pur- poses confined to few genera of fishes, it was believed profitable to investigate the color vision of the large-mouth black bass, Aplites salmoides (Lacepede). The fact that there is a color vision in some fishes does not preclude the possi- bility of some other fishes being color blind or having modified color vision. In addition to the purely scientific outlook upon this problem, there are some economic aspects to the study of this particular fish. Thousands of dollars are spent annually for the pur- chase of highly colored bass lures by sportsmen. Bass in some regions are believed to prefer one color or combi- nation and those ot other regions, others. Furthermore, there is often voiced the belief that at different times of day or in different weathers there are changes in the choice of colors by this fish. Thus this work has a double interest. Acknowledgments This investigation was carried out during the summer of 1935. I should like to take this opportunity to ac- knowledge my indebtedness to Dr. David H. Thompson, of the Illinois State Natural History Survey, at whose suggestion this work was under- taken and who was extremely generous with constructive criticism throughout the course ot the work. Furthermore, I wish to express my gratitude to Francis X. Lueth, student assistant with the Survey, who served as an invaluable aid, and to O. C. McMillin, who kindly constructed certain pieces of the appar- ratus. MATERIALS FOR THE EXPERIMENTS Fish .All the fish used in the following experiments were large-mouth black bass which were caught in Crystal lake, Urbana, 111. One lot of animals was taken on July 5 and the second lot on July 23. The fish varied in length from 2.5 to 4.8 cm. at the time they were caught. Upon being brought into the laboratory the fish were first allowed to remain in large aquaria for a dav or more in order to accustom them to the laboratory temperatures and water supply. Thereafter they were placed in individual white enameled basins, where they remained for the rest of the experimentation. Apparatus Twelve white enameled basins were lined up along one side of a room. There were no windows along this side and as a result the illumination was independent of the changing sky light to a consider- able degree. Never at any time did the fish receive direct sunlight. Throughout the actual experimentation the fish were lighted by a number of incandes- cent lamps in such a way that the dishes were all quite uniformly illuminated to the extent of 12 to 20 foot-candles, depending upon the amount of day- light entering through the single window of the room. The water in the basins was main- tained at a depth of 2.5 cm. and was changed every second day. 36 ILLINOIS NATURAL HISTORY SURVEY BULLETIN Vol. 21,Art. 2 The circular bottoms of the pans were 15 cm. in diameter. These bottoms were marked ofF in arcs of circles differing 1 cm. in radius from one another and all having a common center at a point upon the circumference of the bottom of the pan nearest the observer, fig. 1. The marking of the bottoms of the pans in this manner enabled an observer quite accurately to record the distance of the bass at any given moment from the given point upon the periphery of the dish where training pipettes were to be presented and the fish fed or shocked in association. For shocking the fish, each dish was provided with a pair of fixed copper electrodes 12 cm. apart, dipping into the water at the edge of the pan ap- proximately equidistant on either side of the common center of the arcs of the circles, fig. 1, EE. A 12-way switch per- mitted the experimenter to shock any fish without disturbing the others. Thus, everything in the pans was constant order that the observer could conven- iently carry out the whole training pro- cedure. Colored Pipettes It was found that an ordinary pipette (medicine dropper) covered neatly with white adhesive tape could be readily Fig. 1.— Diagram of the markings upon the bottom of a training basin, and the position of the electrodes, EE. except for the pipettes, which were presented at a given spot from time to time. The electric current for the shock was furnished by two dry-cell batteries and a Harvard inductorium with the secon- dary coil set at about 7 upon the scale of the instrument. A button switch was attached to an extension cord in Fig. 2.—Training pipette, showing the man- ner in which the white adhesive tape was a))- plied. colored by wax crayons or water colors and at the same time act as an opaque container in which the Dapbnia or mosquito larvae which were used as food material could be held concealed until the time for feeding arrived. The adhesive tape was applied to all the pipettes as shown in fig. 2. This was done in such a way as to cover the whole of the glass even to the tip. The colors involved in the experiment were standard according to Ridgway (1912). Wax crayon was used to color the tape on the pipettes for the follow- ing: Rose red Scarlet Flame scarlet Lemon yellow Scheele's green Meadow green Helvetia blue Seal brown The following could be obtained only by using water colors: Orange rufous Analine yellow Calliste green Thulite pink Pale greenish yellow Pale turquoise green Light Columbia blue Black, White and Gray Pipettes For the series of gray pipettes the adhesive tape was first applied to the outside of the pipettes in the same manner as for the colored ones. White was constructed by first whitening the surface of the tape with shoe whitening and, after drying, the whole was rubbed with hard paraffin. Grays 1-6 inclusive I May 1937 BROWN: COLOR RESPONSE OF LARGE-MOUTH BASS 37 were obtained by dipping the tape- covered pipettes into black Inciia ink ot different dilutions. Paraffin was applied to the dried outer surface ot the pipettes. Black 7 was the result of rubbing a covered pipette with a black wax crayon and then paraffining it. Measurements of the relative light reflection of these gray pipettes were made with a MacBeth illuminometer. A petri dish filled with water was placed upon a milk glass plate and 2 inches above the surface of the water the illu- minometer was directed in such a manner that light, emitted from a light source making an angle of 60 degrees with the water surface, was reflected up into the illuminometer at the same angle, fig. 3. The pipettes were lowered one by one just beneath the surface of the water of the petri dish and the re- lative values ot the light reflected from the side ot each was measured. The relative brightnesses of the pipettes as measured by this method were as follows: White 63 Gray 1 49 Gray 2 32 Gray 3 26 Gray 4 21 Gray 5 15 Grav 6 7 Black 7 1 It is thus seen that the range ot in- tensity of the gray pipettes was con- siderable and it was very reasonable to expect that it would cover the range of intensities of the colors used. From a good quality, 1 mm. thick, water-color board, small 6 by 65 mm. strips were cut. Twenty-eight of these were painted with water colors in such a way that the strips could be arranged into a complete visible spectrum divided into 28 grades of color of roughly equal degrees of difference. The stages were made by imitating the pure spectral colors given by Ridgway in his "Color Standards and Color Nomenclature." The colored paper rods were finally dipped in warm paraffin to render them waterproof and to obviate any possible { odor ot the pigments. The paraffin coating had practically no detrimental effect as far as changing the original color was concerned. \. Fig. 3.—Method used to determine the re- lative brightness of the gray pipettes. The following is a in order from red to 1. Spectrum red 2. Scarlet red 3. Scarlet 4. Grenadine red 5. Flame scarlet 6. Orange chrome 7. Cadmium orange 8. Orange 9. Cadmium yellow 10. Light cadmium 11. Lemon chrome 12. Lemon yellow 13. Greenish yellow 1-i. Bright green- vellow list vio 15. 16. 17. 18. 19. 20. 21. 12. 23. 24. 25. 26. 27. of the 28 colors let: Neva green Emerald green Vivid green SkobelofF green Benzol green Italian blue Cerulean blue Methyl blue Spectrum blue Bradley's blue Phenyf blue Blue-violet Bluish violet Spectrum violet TRAINING AND RESPONSES OF LARGE-MOUTH BLACK BASS Method In all the training experiments the fish were shown a number ot pipettes one after another. From a particular pipette the fish were fed either Daphuia or mosquito larvae, whereas trom all the others nothing was ted, but instead, it the fish approached closer than a cer- tain distance trom the pipette a shock was given. In some of the training ex- periments that distance was 6 cm. and in others it was 3 cm. Whenever the same series ot pipettes was shown to a fish time after time, the sequence was con- tinually changed in order that the responses of the fish could not possibly be a result of a learned sequence, but rather must be a response unmodified by anything other than color or inten- sity. That the presence ot the living food in the pipette might have had some attracting influence upon the fish was highly improbable. It was impossible that the effect could be chemical tor the fish responded within a second or two 38 ILLINOIS NATURAL HISTORY SURVEY BULLETIN \^ol. 21 , Alt. 2 and from a distance ranging from two to 15 cm. from the entering pipette. The other possibility, that the bass might have heard the swimming move- ments of the Bapluiia or mosquito larvae, was eliminated by repeatedly presenting an empty pipette of the color to which the fish was trained and obtaining exactly the same results as when food was present. Furthermore, of the pipette and then retreat. The distance they would retreat was inverse- ly related to the frequency with which they struck. A well trained fish would strike repeatedly at its own color, retreating very little between strikes. Occasionally it would strike at another color but in this case it would rapidly retreat considerably farther and appear very timid. Striking a color other than Table 1.—The distance away in cm. that each of 12 untrained black bass remained from colored, black and light gray pipettes. Each value is the average of four traials. Fish May 1937 BROWN: COLOR RESPONSE OF LARGE-MOUTH BASS 39 Table 2.—Records obtained during the training of bass to approach one cf the four colors, rose red, lemon yellow, Scheele's green, or Helvetia blue, and at the same time retreat from the other three colors. Each value for an individual fish is the average number of cm. a fish remained away from the color during five training trials. eat rapidly and continuously from 15 to 20 of these. In all the experiments at training and measuring response this capacity was borne in mind and a safe margin of hunger was always main- tained. With practically no exceptions the bass eyed attentively all the pipettes and rods which were shown them and their other general movements and be- havior also indicated clearly to an observer that the fish were actively responding to them. Reactions of Untrained Fish to Colors and Grays Bass Nos. 19 to 30 inclusive were placed in the individual, white enameled training dishes and Daphiia were added to the water. At first the fish refused to eat and were very nervous. They were permitted to remain unmolested until they had become accustomed to their new surroundings, which required two or three days. At the end of this time they would dash to a pipette and strike futilely at a Daphnia which was clearly visible through the glass wall and then promptly seize the Daphnia when it was ejected from the end. These fish were now considered suffic- iently at ease in their laboratory en- vironment to permit a fair test of the responses of the untrained fish to rose red, lemon yellow, Scheele's green, Helvetia blue, gray 1 and black 7. The same procedure as in training was used except that here there was neither food nor shock associated with the pipettes. The series was shown to each fish four times over a period of two days. The average distance away from each pipette for each fish tor the tour trials was record- ed to the nearest halt of a centimeter I and can be seen in Table 1. These results indicated that red was the most attractive color; yellow was next; then in order came light gray, green, blue and black. This experiment presented reactions upon which the modification of response in the bass Fish 40 ILLINOIS NATURAL HISTORY SURVEY BULLETIN \'ol. 2 1 , Art. 2 - 4 f A S il t ^^ S /1 1 4 S ^1 ti Q ^ c ^ ,Q ^ .Co :? ,"5 ^ h S f^ ^ ^ ^9 ^iS ^i. i- ^S |(^ ^!v b/ ^1 l^'S ^!v ti ^§ i^ ^ yy hO ^' oj. ^ ,C TRAINING TO YELLOW TRAINING TO GREEN 5 4 TRAINING TO BLUE Fig. 4.—Average data for groups of three fish trained to each color, showing stages in the learning process. See Table 2. May 1937 BROWN: COLOR RESPONSE OF LARGE-MOUTH BASS 41 was to be superimposed. It was borne in mind, however, that the attractive- ness of red for the fish might have been the result of the feeding upon Baphuia which were reddish brown. Training to Each of Four Colors Twelve bass were now subjected to training to four colors. Numbers 7, 11 and 14 were trained to rose red as con- trasted with Scheele's green, lemon yellow and Helvetia blue; Nos. 8, 12 and 15 were trained to lemon yellow as contrasted with the other three colors; Nos. 9, 13 and 16 were trained to Scheele's green, and Nos. 6, 10 and 17 to Helvetia blue as contrasted with the remaining three colors. Throughout this experiment the bass were fed Daphnia from the pipette of their train- ing color and were shocked whenever they were less than 6 cm. from the end of any other pipette after three seconds had elapsed. Each fish was shown the series of four pipettes five times each half day for two and a halt days. At the end of this period the fish were con- sidered trained tor the purposes of the experiment. Table 2 and fig. 4 show the records that were obtained during the course ot the training of these fish. The average of each group ot five trials on each color has been calculated and every record in the table is thus an average of five Table 3.—The average distance in cm. away from colors and grays that color-trained fish remained. Each value is the average of nine trials taken with three fish. Trained Trained Trained Trained to to to to Red Yellow Green Blue Rose Red 1.22 4,38 5. 4.5 Scarlet 1.56 3.38 5.33 4.17 Shrimp pink 1.75 2.33 5.17 5.5 Flame scarlet 1.22 5.13 5.5 5.34 Lemon yellow 5.8 1.88 4.66 6. Scheele's green 7.23 5.63 1.84 1.5 Meadow green 7. 6.13 1.5 1.5 Helvetia blue 7.23 6.88 1.33 1.33 Seal brown 6.78 7.38 1.33 1.5 Gray 1 5.9 1.12 5.33 5.33 Gray 2 5.77 1.62 4.17 4.67 Gray 3 5.9 3.62 3.66 3.66 Gray 5 6.66 6.5 2.67 2.17 Gray 6 7. 6.13 2.5 1.33 Gray 7 7.23 6.63 2. 1.5 trials. From this data a series of curves illustrating the learning process has been constructed. These indicate that even at the end ot six to eight trials the bass have become considerably modified in their behavior towards the colors. The learning with respect to the red and yellow has been most striking and that to green and blue much less so. Thus, tor the same kind of training for the same length of time the results of training to the four colors are by no Table 4.—Results obtained in an attempt to train yellow-trained fish to approach lemon yellow and retreat from gray 1 and 2 and shrimp pink. Values for individual fish are average distances in cm. from the pipettes for 10 training trials. Fish No. 42 ILLINOIS NATURAL HISTORY SURVEY BULLETIN V'ol. 2 1 , Art. 2 Fig. 5.—Responses of trained bass Nos. 6-17 to variously colored pipettes. The pipettes are the same as in Table 3. May 1937 BROWN: COLOR RESPONSE OF L A RGE - MOUTH BASS 43 reddish tint in strong contrast to all other colors and shades of gray, and secondly, that among the grays the light grays were slightly more attrac- tive to these fish than were the dark. Fish trained to yellow were not nearly so color specific but responded equally well to the lightest gray and to the yellow to which they were trained. They were also somewhat positive to shrimp pink, a very pale-colored pipette. Lastly, fish trained to green or blue were almost equally positive to green, blue and dark grays. Unlike the fish trained to red and yellow, these fish were quite negative to the light grays. In the light of the similarity of yellow, light grays and shrimp pink to animals trained to yellow it was deemed ad- visable to determine whether the fish had been trained to yellow only as a bright pipette because no other equally bright color was present in the original training series, or whether it was be- cause the fish distinguished very little between yellow and light gray. Five trials training to lemon yellow in con- trast to gray 1, gray 2 and shrimp pink pipettes were given to each ot the fish trained to yellow and then the average of the next 10 trials (still training trials, i.e., feeding upon yellow and shocking upon the other pipettes) was taken as a criterion of the degree of learning ot the fish. The original train- ing had shown that five to eight trials were sufficient to produce a strikingly modified behavior to pipettes the colors or intensities of which were quite dis- tinct for the fish. Table 4 gives the averages of the 10 trials for each ot the fish. Still a light gray and the yellow were quite indistinguishable and the very pale reddish pipette, shrimp pink, was becoming the most different from the yellow though to the human eye it appeared of a brightness intermediate between the two lightest shades of gray. In other words, here again red was dis- tinctive. Bass were next shown two pipettes simultaneously in order to deterniine what the effects of choice would be. The pipettes were placed 7 cm. apart in a metal rack and then the rack lower- ed in such a fashion that the pipettes broke the surface ot the water at the same instant. When the fish struck at one ot the pipettes or remained before one for two seconds, that pipette was Table 5.—Choice of flsh trained to each of four colors when presented simultaneously with two colors. Given Choice of Pi Pipette Selected 1 Trained to Ro^t Tliulite pink Thulile pink Tiiulite pink Shrimp pink Rose red F'lame scarlet Orange rufous Analine yellow Rose red Flame scarlet Red. Fish Nc Gray 4 Gray 3 Gray 5 Gray 4 Gray 6 Lemon yellow Lemon yellow Lemon yellow Seal brown Seal brown Trained to Lemon Yellow, Fish Nos. Lemon yellow Flame scarlet Lemon yellow Orange rufous Lemon yellow Analine yellow Lemon yellow Gray 1 Lemon yellow Pale turquoise green Lemon yellow Calliste green Lemon yellow Thulite pink Lemon yellow Gray 3 Lemon yellow Shrimp pink Trained to Si heele's Green, Fish Nos. Pale greenish yellow Scheele's green Scheele's green Pale turquoise green Meadow green Scheele's green Calliste green Scheele's green Trained to Helveti Light Columbia blue Light Columbia blue Light Columbia blue Light Columbia blue Light Columbia blue Light Coll blue Helveti iblue Helvetia blue Helvetia blue Helvetia blue Helvetia blue Helvetia blue Gray 3 Black 7 Gray 6 Light Colum- bia blue Helvetia blue Lemon yellow Lemon yellow Helvetia blue Blue, Fish Nos. Gray 1 Gray 3 Gray 5 Lemon yellow Thulite pink Scarlet Pale turquoise green Scheele's green Lemon yellow Calliste green Black 7 Seal brown regarded as the choice of the fish. On the other hand, if the fish approachetl neither or swam back and forth between them, then the response of the fish was termed "undecided." Table 5 gives the results ot these experiments. Glancing through this table we are again impressed with the distinctness 44 ILLINOIS NATURAL HISTORY SURVEY BULLETIN Vol. 21, Art. 2 I of red as a color. Several different reds were selected very decisively from a number of contusing colors and grays. The blues, greens and grays appeared to be more or less confused by the fish, the blues being the most often confused with gravs. from those other colors shown it. The results of 30 training trials made after the completion of the first 10 trials were averaged together and are shown in Table 6 and fig. 6. The results are fully comparable since all the fish had had equivalent training. Table 6.—Results of a second experiment of training fish to approach certain colors and to retreat from others. Each value for individual fish is the average distance from the pipettes in cm. for 30 training trials after the initial 10 training trials have been excluded. Fish No. Mav 1937 BROWN: COLOR RESPONSE OF LARGE-MOUTH BASS 45 shades of gray to the first lot ot bass had given a clue that seemed worth follow- ing up. Did rose red and lemon yellow actually appear to the fish as bright colors and did Helvetia blue and Scheele's green appear as colors of low intensity? To the light-adapted human at about this time, the number of trials that are available for red is somewhat smaller. One red-trained animal was given 15 series, and the only one which survived at the end of the training, No. 27, was given 35 series. The results of this experiment are UJI2 46 ILLINOIS NATURAL HISTORY SURVEY BULLETIN Vol. 21,Art. 2 Table 7.- Data obtained in teaching trained fish to approach the color toward which they had been trained and to retreat from each of three shades of gray. Each value for individual fish is the average distance in cm. the fish re- mained away from the color during five train- ing trials. Fish Ma\- 1937 BROWN: COLOR RESPONSE OF LARGE-MOUTH BASS 47 Table 8.—Data obtained in teaching trained fish to approach the color toward which they had been trained and to retreat from each of eight shades of gray. Each value for individual fish is the average distance in cm. the fish remained away from the color during five training trials. Fish Training White Gray Gray Gray Gray Gray Gray Black No. Color 12 3 4 5 6 7 48 ILLINOIS NATURAL HISTORY SURVEY BULLETIN Vol. 21, Art. 2 <^ 3 I- 2 TRAINING TO ROSE RED TRAINING TO SCHEELE'S GREEN TRAINING TO HELVETIA BLUE