Eouse has made experiments in the laboratory of the respiratory changes in pigeons when subjected to smell stimuli. To such odors as oil of bergamot and lily of the valley no appreciable change in the respiratory curve could be noted. Only slight change was noted when asafoetida was used as the stimulus. On the other hand, '' marked reaction " was produced by turpentine and ammonia. Such substances affect the fifth nerve rather than the olfac- tory, as Rouse clearly recognized. Beebe, on the other hand, feels that he has obtained slight positive evidence that birds use the olfactory organs: Three boxes were placed some distance apart in the floor of a cage containing turkey and black vultures. After several days of fasting a piece of tainted meat was placed under the central box. Care was taken to raise the other boxes and to go through the farce of placing something under them. The vultures were very hungry, yet they did not leave their perches and come to the ground, although they had watched his movements. He next reentered and threw two or three, bits of meat to the ground.
Within a second or two, almost as the meat left the hand of the keeper, every vulture swooped to the ground and was hissing and struggling for a portion of the food. Twice the black vultures walked close about the meat box without appearing to notice the odor which was clearly perceptible, even to persons outside the cage. A turkey vulture walked to leeward, instantly turned and made his way to the box, which he examined on all sides. He was soon joined by two others of the same species, and all three took up their stations close to the source of the odor. Soon two black vultures came up, apparently impelled more by imitation than by actual discovery of the smell. All five birds remained for a long time grouped close to the box, going to it now and then, and examin- ing it carefully. Thus even in the turkey vulture the sense of smell is certainly not highly developed, and compared with the sense of sight, is defective indeed.
R. M. Strong in this country has paid most attention to the olfactory responses in birds. His observational ma- terial is mainl}^ inconclusive or negative. In one observa- SENSE OF SMELL IN FISHES 407 tion he states that a turkey vulture was flushed from the entrance of a gopher-turtle hole: " The bird showed a great disinclination to leave the spot, al- though other individuals which were seen by the writer outside of cities were disposed to be wild, A dead gopher-turtle was found inside the burrow. It was impossible to view the turtle except when in a position to look down the oblique burrow, and it did not seem probable that a bird when flying overhead could see the body. A very strong odor of carrion prevailed for some distance on the lee side of the burrow.
" The writer could not rule out the possibility that the vulture had found the turtle outside of the hole through its sense of vision and had later pushed the body inside, but it seemed unlikely that this had happened. The circumstances all appeared to favor the conclusion that the carrion had been smelled, even though the evi- dence was far from conclusive."
In a long series of laboratory experiments where ring doves were used as subjects Strong obtained some evidence that these birds were stimulated by oil of bergamot. The apparatus was somewhat elaborate. The animal under test was put down in such a position that it could be stimulated by a gentle air current passing over oil of bergamot. If the animal followed up this odor it could obtain food (seeds). The apparatus was so arranged that the food could be placed in any one of four food compartments. The birds, while showing definite improvement in forming such a habit, never became perfect in their responses. Un- fortunately Strong did not control his work by sectioning the olfactory bulbs. It is of course possible that the birds were reacting to cutaneous stimulation (fifth nerve).
The. sense of smell in fishes. — Until within recent times there has been little scientific evidence that fish react to foods at a distance. It has been the general practice of fishermen, however, to bait a locality preparatory to fishing in it. This has worked well in the case of the shark and cer- tain other fish, but in such cases it is difficult to say whether the shark senses the food at a distance and then follows the trace or whether chance swimming in the neighborhood leads him to the place where the food can be seen (or, if at night, tasted). We know from anatomical studies that the olfactory bulb and its central connections are well de- 408 SMELL veloped, and that the peripheral organs are also well developed. Furthermore, we know that the fish has well- developed gustatory structures. The neural systems in- volved in the gustatory organs are different from those which care for the olfactory mechanism. Notwithstand- ing these well-marked morphological differences certain authors have maintained that the so-called olfactory organ is only a finely developed taste organ. The difficulty the human being has in smelling a substance contained in a fluid when the fluid is actually in contact with the olfactory membrane has led him to say that he smells only gaseous and vaporous materials, while, on the other hand, he tastes fluids and solids. No one can say positively that gaseous particles have to be dissolved by the mucus present on the surface of the regio olfactoria before activity is initiated in that receptor, yet it is generally so admitted. We know positively, on the other hand, that no substance which can- not be dissolved by the fluids of the mouth can be tasted. The only difference between smell and taste on this view would be in the concentration of the fluids in the two cases. The olfactory mechanism is sensitive to fluids of slight concentration. On such an hypothesis we can conceive of a very small particle of meat being carried out by currents and stimulating the olfactory receptors, e.g., of sharks, even at great distances. If such is the case the olfactory mechanism becomes a distance receptor ranking with the ear and with the eye. To Baglioni, apparently, belongs the credit of advancing the flrst scientific experimental evidence that food substances dissolved in a fluid medium can act upon the olfactory organs of marine forms in the way we have just described, and yet the distinction between animals which ' ' scent ' ' their food and those which see it was made by Bateson many years ago. Baglioni 's experiments were carried out in 1908 upon the octopus and fish (Balistes capriscus). He first blinded the animals in order to make observation more sure. He states that the effect of this operation was to be seen mainly in the limitation of their spontaneous movements. The blind fish especially re- mained quietly resting behind some object in the aquarium.
SENSE OF SMELL IN FISHES 409 This limitation of movements favors the observation of their response to olfactory substances. Baglioni describes his experiments mainly with the octopus and states that the behavior of the fish is the same. We describe one of his experiments upon the octopus: An octopus which had been blinded for more than a month was placed in a tank, the water of which was constantly being renewed from a faucet placed near the middle of the aquarium. The tank was 1.75 meters in length, 40 mm. in breadth, and 25 mm. in height. When the octopus was quietly resting at one end a small dead fish was thrown into the opposite end which came to rest on the bottom. In 1.25 minutes the octopus became agitated. Move- ments of the arm began and then of the whole body. In 3.5 minutes the animal had reached the middle of the basin. At the end of 5 more seconds one arm touched the fish. The fish was seized and carried to the mouth.
Experiments by Parker and Sheldon. — Parker and Sheldon have obtained still more striking evidence of the essentially olfactory nature of such responses. They have worked mainly with the smooth dogfish (Mustelus canis), the killifish (Fundulus heteroclitus), and the fresh-water catfish {Ameiurus nehulosus). All three forms, while differing in the details of their responses, show ability to respond to stimuli which affect the olfactory apparatus. "We present Parker 's excellent summary of his experiments upon the catfish: " Ameiurus nehulosus is a bottom-feeding fish possessing fair powers of sight and unusual gustatory organs located not only in the mouth and on the general outer surface of the body, but es- pecially on the eight barblets about the mouth (Herrick, 1903). It is a hardy fish, living well in confinement and undergoing operations with success. It possesses near its anterior end a pair of nasal chambers, each of which is provided with two apertures, one anterior, the other posterior. The anterior aperture is nearly circular in outline and is located on a slight conical elevation somewhat anterior to the root of the dorsal barblet. The posterior aperture is slit-like in form and lies immediately posterior to the same barblet. The anterior aperture is apparently always open; the posterior one seems capable of slight closure, but is usually freely open.
" By keeping catfishes a few days without food they can be made most eager for it, and if into an assemblage of such individuals a few fragments of fresh earthworms are dropped, the excitement that ensues will last some time after the final piece of worm has been swallowed. During this period the fishes swim about actively 410 SMELL 410 SMELL in the lower part of the aquarium, now in this direction, now in that, and frequently sweep the bottom with their barblets. As can be noticed when feeding actually occurs, the fishes seldom seize a frag- ment of worm until their Jbarblets have come in contact with it. Yet before they have thus touched any food they show a marked degree of excitement and it is this initial nervous state that would lead an observer to suspect that they scented their food. This phase of their activity was, therefore, taken as the one to be tested in connection with their olfactory organs.
" The nasal chambers of the catfish contain ciliated epithelium, the action of which is to draw water in at the anterior olfactory opening and discharge it at the posterior one. As can be demon- strated with carmine suspended in water, the passage through the chamber is accomplished in 8 to 10 seconds.
" As a preliminary step in testing the catfishes, five normal fishes were placed in a large aquarium over night that they might become accustomed to their surroundings. In this aquarium were then hung two wads of cheesecloth, in one of which was concealed some minced earthworm. The fishes, which were swimming about near these wads, were then watched for an hour and their reactions in reference to the wads were recorded. The wad without worms was passed by the fishes many times and did not excite any noticeable reaction. The wad containing the worms was seized and tugged at eleven times in the course of the hour, notwithstanding the fact that from time to time this and the other wad were interchanged in position. Not only did the fishes thus openly seize this wad, but when in its neighborhood they would often turn sharply as though seeking something without success, a form of reaction seldom observed near the wad which contained no worms. Two other sets, of five normal fishes each, were tested in this manner and with similar re- sults. It was perfectly clear to any one watching these reactions that the fishes sensed the difference between the wad of cloth with worms and that without worms.
" To ascertain what receptive organs were concerned in the reac- tions just described, two sets of 5 fishes each were taken from among the 15 normal fishes already tested, and each set was pre- pared differently by subjecting its members to a special operation. One set was etnerized, and, through a small incision between the eyes, their olfactory tracts were cut, thus rendering their olfactory apparatus functionless. From fishes of the other set all the barblets were removed, whereby their external gustatory organs were partly, though not wholly, eliminated. After these operations both sets of fishes were liberated in the large aquarium, where they remained for over two days At the expiration of this time, they were care- fully inspected and tested. They swam about in an essentially normal way and members of both sets snapped bits of worm from the end of a hooked wire much as a normal fish does. Presumably they were in a satisfactory condition for experimentation, " The tests were begun by introducing into the large aquarium containing the 10 fishes a wad of cheesecloth within which were hidden some minced earthworms and recording the kind of fish that visited it and the nature of their reactions. During the first hour the wad was seized 34 times by fishes without barblets but SENSE OF SMELL IN FISHES 411 with normal olfactory organs and, though often passed by fishes with cut olfactory tracts, it was ' nosed ' only once by one of these. A wad of cheesecloth without worms was next substituted for that with worms and the reactions of the fishes were recorded for a second hour. Though members of both sets frequently swam by this wad, none at any time during the hour seized it or even nosed it. These tests were repeated on the same fishes for two succeeding days and with essentially similar results. On the second day the wad with worms was seized 16 times during the test hour by fishes with normal olfactory organs and on the third day 54 times. On both these days the fishes with their olfactory tracts cut made no attempts on the wad with worms nor did any fish at any time nose the wormless wad. The movements of the two sets of fishes when in the neighborhood of the wad containing minced worms were characteristically difi'erent. The fishes with their olfactory tracts cut swam by the Avads without noticeable change; those without barblets, but with their olfactory apparatus intact almost always made several sharp turns when near the wad as though seeking something, and then either moved slowly away or swam more or less directly to the wad and began to nose and nibble it. These reactions were so clear and so characteristic that when taken in connection with the conditions of the fishes, they lead inevitably to the conclusion that the olfactory apparatus of the catfish is service- able in sensing food at a distance much beyond that at which the organs of taste are capable. of acting; in other words, catfish truly scent their food."
The reactions of the killifish under such conditions offer enough points of difference to require noting. It, in con- trast both to the catfish and the dogfish, uses its eyes as well as its nose in seeking food. If a small piece of dogfish is dropped into a pool containing the killifish they imme- diately spring toward it — too suddenly to suppose that diffusion could have taken place to such an extent that the olfactory system could be stimulated. Even if a ball of filter paper is thrown into the tank they dart forward and seize it, but soon discard it. It is the eye likewise which makes them dart forward and seize a packet of plain cheese- cloth. On account of their use of vision it is hard to make conclusive experiments upon their olfactory sensitivity un- less they are blinded at the same time that they are made anosmic. On account of the fact that in normal fish plain cheesecloth packets are soon discarded, it is possible to test the effect of cutting the olfactorj^ nerves. Anosmic animals, when tested with two packets, the one containing hidden meat and the other without food, nibble at both packets 412 TASTE 412 TASTE in such a way that an uninformed observer could not tell which contains food. When these same two packets are placed before normal animals the food packet becomes immediately surrounded by a vigorous assemblage of con- testing individuals.^ II. Taste Location of gustatory organs. — The gustatory organs in certain fishes have a widespread distribution. In addi- tion to the taste buds, which are found in the mouths of fishes, we find them frequently upon the lips, the outer skin of the head and on the trunk. In Ameiiirus it has been shown that the terminal buds occur in the skin of practically the whole body surface but especially on the barblets. These terminal buds must be sharply distin- guished from the sensory structures in the lateral line sys- tem, which they resemble rather closely (p. 394). The sensory nerve supply to the terminal buds of the mouth is shared in by X, IX, and VII pairs of cranial nerves; those of the bodily surface (Ameiurus) largely by the lat- eral accessory branch of the VII cranial nerve.
Herrick's experiments upon the functional signifi- cance of the taste buds. — In the discussion of the func- tional significance of the taste buds we shall consider mainly the experiments which have been made upon the catfish (Ameiurus). As has been stated, this animal rarely uses its eyes in the search for food. If, in the course of random and aimless movements, the fish pass near the sides and bottom of the aquarium which contains food, they touch the food with the lips or barblets, and instantly seize and swallow it. They are easily frightened and ex- perimentation is difficult. If one arranges the aquarium so that the fish may partially conceal themselves, leaving certain of the fins and other parts uncovered, it is possible to make experiments which show the functions of these ^ More recently still Copeland has found that the puffer {Spheroides maculatus) shows the same ability to respond to con- cealed food. His method of determining this fact was essentially like that of Parker.
FUNCTION OF TASTE BUDS 413 taste buds. If, under these conditions, a piece of meat is lowered so that it touches the barblet, the meat is instantly seized and swallowed. Touching any part of the body, head or tail, produces the same reaction. Vision is not necessary to set off the reflex of seizing and swallowing. From the above experiments we are not able to decide whether touch or taste is the functional sense. When they are touched with cotton wool on any part of the body under similar conditions, they will likewise turn and seize the wool, but soon drop it. If, now, one continues stimu- lating them with cotton wool, they cease to respond. When they have become habituated to the contact with the wool, one again tries them with meat. Invariably the response is called out. It is evident that both senses participate. The response is both a tactual and a gustatory reflex. The gustatory reflex is obtained (after tactual accommoda- tion) if the cotton wool is soaked with meat juice. In order to produce typical gustatory reflexes, the stimulus must be strong and localized. Diffusion of juices called forth only seeking movements.* While the olfactory receptors may have participated in these reactions, it is quite prob- able that they would have occurred in anosmic animals.
While Herrick did not section the olfactory nerves in the catfish, he did perform such an operation on several tom- cods {Microgadus tomcod). From the second day after the operation the animals fed normally. Furthermore, "After the third or fourth day the fishes took their food in all respects Hke uninjured fishes, so far as could be observed. They gave all of the characteristic reflexes that have been mentioned above, including the discrimination between cotton wool and cotton dipped in clam juice, and between sea water and clam juice applied with a pipette, etc. The operated fish would locate a concealed bait by means of pelvic fins exactly as the normal fish does, and he would similarly root it out and eat it. In short, the gustatory reflexes, so far as I have observed them, were absolutely unmodified by the * Similar tests upon the pollack (Pollackhius vires), hake {Uropliycis te^mis), and the tomcod {Microgadus tomcod) have yielded similar results. The pollack has very well developed eyes and darts for the food when he sees it. Sight plays a large part in its daily life. It is not well suited to experiments upon taste. The fins are very sensitive to meat.
414 TASTE operation. That the olfactory apparatus was totally destroyed was verified by autopsy dissections made after the close of the observa- tions." ^ Parker's experiments upon the gustatory responses of fishes. — ^Parker's own experiments confirm the work of Herrick and afford some additional facts. He investigated the sense of taste in the common hornpout. The taste buds in this form occur not only in the mouth and in the bar- bules but also over most of the external surface. Those on the exterior of the fish are innervated by the branches of the seventh nerve. Hornpouts will snap at a bait when it is presented to the flank of the body as well as when it is close to the mouth. The exterior of the body is sensitive to sour, saline, and alkaline solutions, the head being more sensitive than the trunk. In hornpouts, if the branches of the seventh nerve distributed to the skin have been cut, thus destroying the sensitivity of the skin, there is no longer a response to bait brought near to the trunk though the same bait is eagerly taken when brought near to the mouth. Such animals, however, retain their full sen- sitiveness to sour, saline, and alkaline solutions when ap- plied to the skin of the trunk. The loss of sensitivity to bait is not due to shock. They still snap at the bait if the lateral branch of the tenth nerve has been cut. After cut- ting the lateral branches of the seventh and tenth nerves the only sensory nerves left intact on the flanks of the body are the branches of the spinal nerves. Fish in this condition still respond to sour, saline, and alkaline solu- tions. It follows that these solutions must stimulate the terminals of the spinal nerves and that these ^ While there is a seeming contradiction between these experiments and those cited from Parlier and Sheldon on p. 409, it may partially be removed by supposing that the tomcod has an extraordinarily finely developed sense of taste in its pelvic fins — so highly developed, indeed, that it functions like a distance receptor. Neither Parker nor Sheldon worked with this form. Parker worked with the catfish, having its olfactory mechanism as the chief point of interest, while Herrick was interested mainly in the gustatory mechanism. Their results are not strictly in harmony. It is unfortunate that Herrick did not work with anosmic catfish. Even allowing for the apparent contradiction it would seem that both investigators have made their respective points, viz., that the terminal buds distributed over the bodily surface are true taste structures (Herrick) and that the olfactory mechanism serves as a distance receptor (Parker).
'' COMMON CHEMICAL SENSE " IN PISHES 415 nerves must be regarded as chemical in function, though they are not primarily concerned in the response to bait. In hornpouts, the posterior half of whose spinal cord has been destroyed but whose seventh nerve is intact, there is no response when sour, saline, and alkaline solutions and bait are applied to the flanks and nes.r the tail. The absence of response to hait in this experiment Parker be- lieves to be due to the loss of the motor mechanism of the cord whereby the fish turns to snap at the bait and not to the loss of the spinal sensory fibers. The loss of these fibers would account for the absence of response to sour, saline, and alkaline solutions. Parker believes that the sense of taste is complex and involves not only the seventh nerve but also the spinal nerves; i.e., that there is both a sense of taste and a " common chemical sense." This conception, which does some violence to the common notion of the function of cutaneous receptors, should be carefully con- sidered.
III. The '^ Common Chemical Sense " in Fishes Introduction. — There is a growing tendency to use the term " chemical sense " generically and to make it include smell, taste, and a sense which may be described as the " common chemical sense." While this usage seems unde- sirable it is impossible to discuss certain experimental results without using at least the term '' common chemical sense." This is especially true when we come to treat of the responses of fish, amphibia, and the invertebrates to chemical substances. It is asserted (Sheldon) that the '* common chemical sense " is entirely different from the ordinary cutaneous system, and that all vertebrates from the lancelet to man possess it (as well as all invertebrates). In man the receptors for this organ lie especially on the exposed mucous surfaces such as those of the nasal cham- bers, the mouth cavity, and the moist surfaces of the eye- lids. Parker states that " the receptors on these surfaces are normally stimulated by the chemical action of the ma- terial in direct contact with them and they represent col- 416 " COMMON CHEMICAL SENSE " lectively a sense as distinct and well defined as smell or taste." The work on human sensory physiology within recent years has tended to show that there are four dis- tinct kinds of cutaneous receptors — stimulated respectively by (1) pressure, (2) cold, (3) warm, and (4) noxious sub- stances. These may be stimulated separately or simul- taneously. Furthermore, there is some positive evidence (von Frey and others) that the hairs and Meissner cor- puscles mediate pressure responses, while the end bulb type of corpuscle mediates cold, and the Ruffini cylinder and re- lated types mediate warm stimuli. The free nerve endings are supposed to be stimulated by noxious substances, acids, cuts, burns, bruises, etc. This whole system is collectively called the " cutaneous sense." It would seem that our present concept of this sense is adequate and that the use of the term ' ' common chemical sense ' ' is confusing. There has been no distinction made between the external skin and the skin bathed by mucus. The experimental work of Head and Sherren (and of Rivers), however clearly it may show that our former ideas of localization of impulses in the cord were in error and that motor nerve trunks carry sensory fibers, does not seriously modify our ideas of the nature of the external cutaneous system. That the cutaneous system, as we understand that term in man, exists unchanged through the vertebrate series, we do not for a moment claim (especially in view of the fact that highly organized sensory terminals like those found in man apparently do not exist in the skin of the fish). The experiments of Parker and of Sheldon on fish tend to establish the view that differences do exist between man and certain of the lower vertebrates. Sheldon's experiments upon the smooth dogfish. — The older experiments upon the " common chemical sense " in fish (e.g., Nagel's) are unsatisfactory because they failed to make any distinction between those forms which possess taste buds only in the mouth cavities, and those which possess such terminals on the bodily surfaces. Sheldon worked only upon the smooth dogfish {Musteliis canis), which possesses taste buds only in the mouth cavities. The sensitivity of the whole bodily surface to chemical stimuli " COMMON CHEMICAL SENSE " IN FISHES 417