STUDIES IN REPTILIAN COLOUR CHANGES

Transcription

1 7 STUDIES IN REPTILIAN COLOUR CHANGES II. THE PITUITARY AND ADRENAL GLANDS IN THE REGULA- TION OF THE MELANOPHORES OF ANOLIS CAROLINENSIS BY L. H. KLEINHOLZ, PH.D. Biological Laboratories, Harvard University (Received 2 January 938) (With Four Plates and Three Text-figures) I. INTRODUCTION THE early literature of reptilian colour changes deals almost exclusively with the African chameleon. In recent years the investigation of metachrosis in the chameleontid reptiles has been leading to a comprehensive knowledge of the physiological mechanisms involved in this activity, chiefly through the studies of Hogben & Mirvish (928) and of Zoond & Eyre (93). Investigations of the pigmentary activity of lizards native to North America began at the beginning of the present century. Our knowledge of the chromatic function in these forms has progressed through the publications of Parker and his associates. The early reports of Carlton (903) and of Parker & Starratt (90) on Anolis and of Parker (906) and of Redfield (98) on Phrynosoma served to define the bionomic aspects of colour change for these forms. Studies of pigmentary activity in the New World lizards were greatly influenced by the views of those nineteenth-century investigators who had engaged in the analysis of colour mutability. Briicke (852), Krukenberg (880) and Keller (895) found that the chromatic responses of African chameleons were under the control of the nervous system. Later investigators of chromatophoral activity were so convinced of this mode of control that for a long time no other regulating mechanism was suspected. Redfield (98), investigating the effects of adrenalin on the melanophores of Phrynosoma, found this hormone to be a potent agent in inducing chromatophoral changes. Redfield's explanation of colour changes was a novel one but soon gained apparent support by the demonstration that the chromatic function in amphibians was under endocrinal control (Smith, 96; Allen, 97; Atwell, 99, 92; Hogben & Winton, 922a, 9226, 923). These recent developments in the physiology of chromatophoral behaviour were a distinct advance in establishing a hormonal basis for the activity of such effectors. Little, however, is known of the role played by blood-borne substances in reptilian colour changes. Redfield (98) believed that in Phrynosoma "the melanophores are co-ordinated by two distinct mechanisms, the adrenal secretion and the direct action of nerves", either mechanism alone being capable of effecting

2 Studies in Reptilian Colour Changes 75 melanophore concentration (p. 309). But Hogben & Mirvish (928) and Zoond & Eyre (93), working with African chameleons, disputed Redfield's results. Carlton (903) and May (92) believed that chromatophoral responses in Anolis were regulated by nerves. Noble & Bradley (933), incident to their study of the moulting process in lizards, observed that hypophysectomized specimens of Hermdactyhs were pale in colour. While examining the hypophysis of Anolis carolinensis, I was prompted by the highly secretory appearance of the intermediate lobe (unpublished results) to investigate the part played by the pituitary gland in the colour changes of this lizard. Results from preliminary experiments (Kleinholz, 935, 936), coupled with the need of revising several concepts that were extant in the literature, led to an extension of this work to include the various factors, nervous as well as endocrine, that might be involved in the chromatic responses of this lizard. This study is a report of the results obtained from experiments designed for that purpose. It is a great pleasure to acknowledge my most appreciative thanks to Prof. Alden B. Dawson for his interest and helpful criticism and to Prof. G. H. Parker for his friendly encouragement during the course of this study. II. MATERIALS AND METHODS A plentiful supply of Anolis carolinensis was obtained from collectors in the south-eastern part of the United States, which is one of the natural habitats of this lizard. In the laboratory the animals were fed with meal-worm larvae and blowflies. Water was supplied by sprinkling the cages daily. Use was made of a heavy cardboard carton in studying adaptations to backgrounds. One hah 0 of the carton was painted white and was separated by a partition from the second half, which was painted black. Lizards were kept in individual glass jars which could be moved from one coloured background to the other without actual handling of the animals. The backgrounds were illuminated by a 00 W. electric lamp at a distance of 30 in. from the bottom of the box. Surgical operations were performed under light ether anaesthesia, with no particular regard for sterile conditions. Hypophysectomy was accomplished through the oral route, the pituitary being visible through the floor of the skull as a pinkish-white mass, located about midway between the level of the orbits and the posterior margin of the skull (Text-fig. ). After displacing the bone in this region, the membranous investment of the gland was slit and the entire pituitary removed with the aid of a suction pipette. The slight amount of haemorrhage which occurs is negligible and soon stops; recovery from the effects of the operation is prompt. Lizards were blinded by cauterizing both eyes. Optic enucleation proved impractical because of a large blood sinus situated close to each eye, so that any such attempts resulted in fatal haemorrhage. Denervation of given regions of skin was accomplished in several ways, one of which was by making transplants. In autotransplantations the areas of the grafts and those of their new sites, changing in shape and size (presumably because of

3 76 L. H. KLEINHOLZ elastic fibres in the dermis), resulted in unsatisfactory "takes". The use of homoiotransplants overcame this difficulty. A receptive area was cut out on one specimen and a large square of skin from a second lizard was transplanted to the new site. After smoothing out the graft, the edges were trimmed to fit exactly with those of the host skin and the adjoining margins lightly covered with celloidin. A second method of denervation was to cut given spinal nerves. This was readily performed by making an incision through the ventral body wall into the coelom. The viscera were displaced, a small slit made in the dorsal peritoneum, and several successive spinal nerves sectioned. The incision was subsequently stitched together with silk N Text-fig. i. The skull of Anolii carotineniit in ventral view, showing the location of the pituitary gland. N, internal nares; O, orbit; P, pituitary gland. thread. In several animals one of the hind legs was denervated by cutting the sciatic nerve, care being taken not to injure the accompanying blood vessel. In the third method of denervation the spinal cord was sectioned at various levels from within the body cavity by means of a sharp chisel. By extending the incision laterally either or both sympathetic chains could be cut. Within two days after the initial cutting the posterior cord was pithed with a fine wire. Electrical stimulation, when used, was obtained by a Harvard induction coil connected to a 2 V. D.C. wall outlet. Leads from the secondary posts of the induction coil were devised to meet the needs of various experiments. Adrenalectomy was performed after preliminary bilateral castration and appropriate ligation of the adrenal blood supply. The glands in Anolis are compact,

4 Studies in Reptilian Colour Changes 77 no scattered chromaffine tissue being macroscopically evident. Recovery from the operation took place within 2 hr., and individuals survived from 0 to 2 hr. III. THE HISTOLOGY OF THE MELANOPHORES IN COLOUR CHANGES In its colour play Anotis ranges from dark brown to bright green, intermediate shades including light brown, yellow, yellowish green and emerald-green. Individual lizards are not uniformly coloured but show considerable variation in certain regions of the body. A dewlap, or throat fan, is present, being most highly developed in males. When erected in combat or in courtship it is brilliant red, due, in part, to granules of pigment in the skin of that region. In some animals a mid-dorsal longitudinal band is present, extending from the cervical region for varying distances toward or even along the tail (Text-fig. 3); in many lizards it is completely absent. When an animal is in the dark condition, a regular darker pattern is discernible against the brown colour of the rest of the body; such markings vary among specimens but are constant in form for a particular individual. The darker areas consist of groups of from two to a score or more of scales scattered over the dorsal and lateral aspect of the body and appendages. In addition to this body pattern all lizards in the dark state have a darker quadrangular patch immediately posterior to each orbit. In some individuals a similar area is located over each scapula. These localized markings will be considered again below and will be referred to as the post-orbital and the scapular regions. In the light state the dorsal and lateral surfaces of the body are green. The ventral surface ranges from cream colour to brownish white in appearance and is frequently peppered over with black spots. A brief description of the important histological features of the skin will be reported here in order to provide a better understanding of the processes involved in the mechanics of metachrosis. Von Geldern's (92) paper gives a good description of the finer details. Skin from lizards in various colour states was fixed by immersion in hot water (80 C.) for 0 sec. to fix the position of the pigment. Embedding in paraffin was effected by the rapid dioxane method. Some sections were stained with Mallory's azan while others were mounted unstained. The epidermis is a transparent layer, typically divisible into the stratum corneum and the stratum germinativum. Just beneath the epidermis is a layer of yellow oil droplets and xanthophores. These are not present in ordinary histological preparations, having been dissolved by the fat solvents usually employed in such techniques, but can be seen in sections cut by the freezing method. Internal to the oil droplet layer lies a much thicker region known as the leucophore layer; this is composed of a number of irregularly spaced blocks or plates, the long axes of which are parallel to each other. In unstained sections the leucophore region appears brownish by transmitted light but is bluish white when examined by reflected light. Below the leucophore layer are situated the melanophores, which are the cells

5 78 L. H. KLEINHOLZ actively concerned in the colour changes of Anolis. The cell bodies of the melanophores lie embedded partly in the leucophore layer and partly in the deeper connective tissue of the dermis. A varying number of branches, extending vertically upward from the cell bodies of the melanophores (PL I, fig. ), pass through the interstices of the leucophore layer and are subdivided into many finer branches which terminate beneath the inner surface of the epidermis. When the skin is in the brown state the pigment granules are dispersed through the fine terminal branches of the melanophores, resulting in a layer of pigment just beneath the epidermis (PL I, figs. and 2). In the green phase of skin colour the pigment is present only in the proximal portions of the branches, having been withdrawn from beneath the epidermis (PL I, figs. 3 and ). The finer distal branches are practically transparent except for a few scattered granules which may have failed to migrate proximally with the main mass of the pigment. It is evident that the melanophores are the cells actively concerned in the various colour phases of Anolis. The other components of the skin, namely the oil droplets, the xanthophores and the leucophore layer, play a passive or at most a very restricted part in metachrosis, acting as filters or reflectors. IV. THE EFFECT OF LIGHT ON THE COLOUR CHANGES OF ANOLIS A. Responses to changes in backgrounds It is now almost universally recognized that conditions of illumination and of temperature are the principal environmental factors concerned in reptilian colour changes. Normal specimens of Anolis carolinensis are brown at room temperature ( C.) in the light and are green in darkness. These observations, first reported by Carlton (903), have been amply confirmed by Parker & Starratt (90) and by von Geldern (92). Hadley (929, 93) extended the observations to include Anolis porcatus and A. todurus. Similar behaviour of the melanophores of the horned toad was reported by Parker (906) for Phrynosoma blainvillei and by Redfield (98) for P. comutum. Colour responses of Anolis to changes in background have not been extensively investigated. Hadley (93) found that Anolis todurus underwent adaptive changes to white and to black backgrounds, but failed to state the time required for these processes. The species of Anolis used in this study showed active responses to such environmental changes. Lizards in the dark state, when placed upon a brightly illuminated white background, become green; conversely, green lizards, when placed upon an illuminated black background, rapidly become brown. The time curves for such chromatic adaptations (Text-fig. 2) were plotted by assigning a numerical value (Table I) to each of several stages in the progressive changes of colour from the light to the dark condition (Hogben & Slome, 93). The time relations shown in Text-fig. 2 demonstrate that the onset of the dark state (curve A) is rapid, approaching a maximum after the lizard has been on a black background for approximately 6 min. The rate of change in colour of a lizard in the dark condition when it is placed on an illuminated white background (curve

6 Studies in Reptilian Colour Changes 79 B) is subject to variation. It is interesting to record in this respect that some twenty Anolis (the greater part of a shipment of lizards from a dealer in Florida) showed no chromatic responses to changes in backgrounds. The animals remained in the dark condition irrespective of the backgrounds on which they were kept; they did, in an irregular manner, undergo pallor in darkness. Such individuals which are refractory in their colour behaviour have also been found among fishes TIME IN MINUTES Text-fig. 2. Time curves for metachrosis of normal and of blinded animals. A (circles), nonna specimens in the light state placed on an illuminated black background; the curve represents the averages of observations on nine lizards. B (triangles), normal dark lizards placed on a white background; averages of observations on nine individuals. C (squares), blinded specimens in the dark state placed in darkness; the curve represents the averages of twelve observations on four lizards. Table I Colour Green Green with light brown regions Uniformly brownish green Light brown Dark brown Index 2 3 S and amphibians. The wide scatter in the time curve for response to a white background may be due to some varying internal factor. It was also observed that a lizard which had darkened on an illuminated black background did not always remain in the completely dark phase, but showed appreciable fluctuations in its chromatic condition; these fluctuations, however, did not occur in a sufficiently regular manner to be interpreted as rhythmic behaviour. B. Responses of blinded lizards The importance of the eyes in pigmentary responses to changes in backgrounds and to changes in light intensity is obviously suggested as a factor to be deter- JEB-XViv 32

7 80 L. H. KLEINHOLZ mined experimentally. So far as I know, the effects of blinding upon colour mutability have not been thoroughly studied in any of the American lizards. The generalized statements of some investigators on the colour mutability resulting from blinding are confusing, because in most cases no distinctions are drawn between responses to backgrounds and responses to light and to darkness after such operations. Von Buddenbrock (928) says (p. 397): "Bei vielen Krebsen, den Fischen und den Amphibienlarven ist der Erfolg dieses Eingriffs stets eine starke Verdunklung. Anderseits hat Blendung keinen Einfluss auf die Farbung der Tintenfische, und Reptilien." Parker (936, p. 2) says: "The eyes of animals are essential to their colour changes, for when these organs are removed or effectually covered all signs of such changes disappeared, and the given animal so far as an alteration of its tint was concerned was largely incapacitated"; although in the same publication (p. 23) a distinction is made between background responses and the responses to change in light intensity: "Another feature which should be noted in passing is that a Fundulus from which the eyes have been removed is pale in darkness, as normal individuals are, but moderately dark in bright illumination." Four specimens of Anolis, after they had been tested and found to be normal in their colour responses to background changes, were blinded by cauterization of the eyes. Twenty-four hours were allowed for recovery from the shock of the operation before the animals were used in the following experiments. When these otherwise normal lizards were tested, it was found that the colour responses to backgrounds had disappeared, but that the typical chromatic reactions to light and to darkness persisted apparently unimpaired. The animals were placed in the dark room while they were brown. At the end of 7 min. one individual was entirely green and the remaining three were perceptibly lighter in colour. A second lizard was green at the end of 7 min., while the other two required 38 min. to attain the stage of complete pallor. When the animals were kept on either black or white backgrounds which were illuminated from above, they attained the dark state within 6 min. The average time required for darkening of blinded individuals when illuminated, computed from eighteen observations on these four lizards, was 7 min. The average time for attaining a condition of complete pallor in darkness (Text-fig. 2, curve C) was 22- min., calculated from twelve readings on the same animals. These time periods agree with those reported by Carlton (903) and by Parker & Starratt (90) for normal animals (Table II). The fact that blinded lizards, although they have lost the ability to adapt their colours to backgrounds, are still able to undergo metachrosis in response to changes in light intensity implies either that the melanophores in the skin are directly responsive to light and to darkness, or that some photoreceptive mechanism, in addition to the eyes, is involved in the system responsible for colour changes. Such possibilities will be considered more critically later.

8 Studies in Reptilian Colour Changes 8 V. THE DARK PHASE A. The effects of hypophysectomy Within an hour after hypophysectomy such successfully operated animals turn green and remain permanently in the light condition (Text-fig. 3); colour responses to background changes disappear. In several cases where removal of the pituitary Text-fig. 3. On the left is a normal Anolit in the dark condition; the white mid-dorsal stripe extends from the cervical region on to the tail. The animal on the right has been hypophysectomized and is consequently in the light state. was apparently incomplete the lizards remained in the light state for 0 to 20 days, after which they darkened slightly when kept on an illuminated black background; post-mortem examination of the pituitary region of such individuals showed that portions of the intermediate lobe were still present, protected by the ridge of bone that forms the posterior wall of the sella turcica. The pale condition of hypois-*

9 82 L. H. KLEINHOLZ physectomized specimens of Anolis is not a result of mechanical manipulation or injury undergone during the operation because lizards in which the bone surrounding the pituitary had been chipped away, and animals from which the greater part of the anterior lobe had been removed, underwent normal colour changes. Lizards from which the pituitary gland has been removed, and which have been kept in darkness, do not respond to illumination by darkening but remain in a condition of pallor. It was shown above that blinded individuals, although having lost the ability to respond to changes in backgrounds, undergo chromatic changes in response t6 changes in light intensity, becoming dark when illuminated and pale when kept in darkness. Such behaviour implies either a direct reactivity of the melanophores Table II Average times to change From brown to green 250 min. IQ From green to brown 0 min Temperature? 20 C. 22 Investigator Carlton Parker-Starratt This paper to light or the presence of photoreceptors other than the eyes. If blinded lizards are hypophysectomized, however, they become green and remain so, notwithstanding that they are maintained in an illuminated environment. Equally noteworthy is the fact that specimens of Anolis which have been successfully hypophysectomized do not darken, either in direct or in diffuse light, after they have been blinded. These experiments preclude the possibility that melanophores in the intact skin of Anolis carolinensis are directly responsive to light. Such results differ from those reported by Hadley (93, p. 325), who studied the behaviour of isolated skin. If the circulation is excluded from any part of the skin of Anolis, that region becomes green and remains in this light condition regardless of the fact that the animal is kept on an illuminated black background. This can be readily demonstrated by ligating one of the legs, the skin of which becomes green within 5 min. Removal of the ligature while the lizard is still on its black background soon results in a darkening of the hitherto green skin. Similar results were obtained after the femoral artery of a green lizard was exposed and ligated. It follows from these results that a portion of skin entirely removed from the body should, since the chromatophorotropic hormone can no longer act on the melanophores through the blood stream, turn green. Such is actually the case. B. Injection of pituitary extracts To supplement the experiments involving hypophysectomy, extracts of the pituitary glands of various animals were injected into lizards that were in the light state, either as a result of hypophysectomy, or of having been kept in darkness or

10 Studies in Reptilian Colour Changes 83 on an illuminated white background. Injections were in all cases made intraperitoneally. This is necessary because subcutaneous injections of distilled water or of Ringer's solution are capable of producing a dark spot at the site of injection, possibly because of the pressure exerted on the melanophores by the wheal of fluid. Carlton (903) found that a piece of isolated skin turned green but could be made to darken by tapping it gently with a blunt instrument. The results of such injections are shown in Table III. In teleost fishes the hypophyses are not distinctly divided into definite lobes; consequently the entire gland was used in the case of Fundulus heteroclitus. Injection of such extracts into normal specimens of Anolis which were in the light state (through having been kept in darkness) and into hypophysectomized lizards (Kleinholz, 935) resulted in a pronounced darkening. In frogs and in Anolis only the neuro-intermediate lobes were used in preparing extracts. Experiments B and C illustrate in the first case the rapidity with which the darkening response was evoked and in the latter that the induced dispersion of the melanophores may persist for several hours. In experiment E a commercial preparation of mammalian posterior lobe in 0-5 % chloretone was used; control injections with chloretone solution were not made. C. Early evidence for nervous control of darkening The results from hypophysectomy and from injecting pituitary extracts prove clearly that the dispersed condition of the melanophores in the skin of Anolis is effected by a hormone from the hypophysis. The evidence of earlier investigators that melanophore dispersion in Anolis was under direct nervous control is fragmentary and inconclusive. Carlton (903), describing the darkening that resulted when a green Anolis was brought from a dark box into the light, offered two possible explanations for such behaviour; the change might be due either to the direct action of light on the melanophores or to the mediation of the nervous system which was stimulated by light. By proving that light which fell on one part of the animal's body induced darkening in the other non-illuminated portions, he eliminated the possibility of the direct action of light on the melanophores, and came to the alternative view because he knew "of no way of explaining the induced changes except on the assumption that nerves served as intermediate organs". Cutting and destruction of the posterior cord had no effect on the ability of a lizard to darken completely under environmental conditions which normally evoked dispersion of the melanophores. By a process of elimination Carlton came to the conclusion that this phase of melanophore activity was controlled by the sympathetic nervous system. His sole evidence for this view was the fact that brown lizards turned green after injection of nicotine. It has been commonly observed, however, that ordinary handling of this lizard readily induces the green state. At that time the possibility of an endocrine factor in metachrosis was practically unsuspected. May's belief (92) that the colour changes of Anolis were due to the transmission of stimuli over nerve fibres supplying the melanophores resulted from a

11 Table III A. An extract was prepared by triturating the pituitaries of four Funthdus heteroclitw in 0-5 c.c. of amphibian Ringer's solution. The first two columns show the colour responses of two pale normal lizards which were kept in darkness after having been injected each with 0-25 c.c. of the extract; the third animal was a control lizard which was injected with an equal volume of brain tissue suspen- Time after injection Colour index 0 min B. The neuro-intermediate lobes of eleven Rana pipiens were ground up in 0-9 c.c. of Ringer's. Three hypophysectomized lizards were injected, each with o - i c.c. of the extract: Time after injection Colour index 0 min S C. An extract of eight frog's pituitariea in o-6 c.c. was prepared as in B. Each of three hypophysectomized lizards received o-i c.c. of extract: Time after injection Colour index 25 min ' ' ' D. Each of three hypophysectomized lizards was injected with C5 c.c. of Ringer's solution containing one intermediate lobe extract of Anolis gland: Time after injection 0 min. 2 5 IO I Colour index E. One part of Parke, Davis and Co. " pituitrin " was diluted with nine parts of Ringer's solution. Normal lizards, lightened by adaptation on a white background, received 0-05 c.c. of this preparation: Time after injection I 5 Colour index 0 min i '5 5 '5 5 '5 5

12 Studies in Reptilian Colour Changes 85 study of the chromatic responses of skin in auto- and in homoio-transplantations. In his experiments, areas of skin which had been denervated (by the operation of transplantation) lost their capacity for colour changes over a period of 2 to 3 weeks, although during this period the skin of the host was normal in its chromatic behaviour. At the end of this time the engrafted skin resumed pigmentary activity in concurrence with the skin of the recipient. May attributed these results to the regeneration of nerve fibres into the hitherto denervated region. D. Behaviour of denervated skin I am able to confirm Carlton's results obtained from cutting nerves. Lizards whose spinal cords had been sectioned at various levels were able to become uniformly brown if kept on black backgrounds; similarly, animals whose posterior cords had been pithed did not differ in their colour changes from normal individuals. Cutting the sciatic nerve of one leg had no effect on the pigmentary condition, the skin of that leg behaving in all respects like that of the control member. Uni- or bi-lateral section of the sympathetic chain had no visible effect on the ability of the lizard to darken. It is also interesting to note that neither von Geldern (92) nor May (92) were successful in their attempts to demonstrate innervated melanophores in skin treated by the usual neuro-histological methods. By making skin transplants and then testing for colour changes on black and on white backgrounds, it was found that, although considerable variation existed, the grafts showed unmistakable melanophore activity. As can be seen from Table IV, the grafts in several cases of auto-transplantation evidenced metachrosis within 2 to 8 hr. after the " denervating " operation of transplantation; in other instances the denervated patches did not change from the initial green colour until the graft was 7 to 8 days old. In animal 6 (Table IV) when the reciprocal auto-transplants were one day old the graft on one side required 75 min. longer than the other to become brown, while several days later the two skin patches darkened at approximately the same time. These variations are undoubtedly due to differences in the vascularization of the grafts, since the transplants did not "take" in good contact with the receptive areas in many cases. That this explanation is correct is seen from the fact that homoioplastically transplanted skin (which was fitted smoothly into the region excised for it on the host) in all cases turned brown when the host was kept on an illuminated black background. Further support is given this view when the effects of injections of "pituitrin" on host and on graft skin colour are studied. In all cases there was an obvious temporal difference in the response, the transplanted regions being appreciably slower than the skin of the host in their rates of darkening (Table IV and PI. II). E. Effects of electrical stimulation To determine whether the dark state might in addition be induced by a nervous agency, various methods of electrical stimulation were used. Leads from the secondary posts of the induction coil were inserted one into the cloaca and the second into the animal's mouth; the passage of current caused the lizard to clamp

13 86 L. H. KLEINHOLZ its jaws upon the mouth electrode, thereby maintaining a closed circuit. The source of current was a 2 V. D.c. wall outlet. Stimulation was kept constant in all cases, the secondary coil being at a distance of 6 cm. from the primary and the current being allowed to pass through the animal for exactly 2 min. Table IV A. Auto-transplants. The numbers in parentheses after the method of treatment indicate the age of the graft in days. The time periods which are marked with a cross (+ ) are only approximate, the mean values having been taken of the two observations between which the colour change occurred. The "pituitrin" was diluted with four parts of distilled water. Serial number Method of treatment From brown to green Host Graft Time to change From green to brown Host Graft Background (7) Background (2) Background (9) Background (i) Background (7) Background () Background (7) Background (0) Pituitrin inj. (5) Pituitrin inj. (2) Pituitrin inj. (8) 60 min min. 5 5 min min B. Homoio-transplants. The pituitrin was diluted with nine parts of distilled water. Serial number Method of treatment Background () Background () Background () Background (2) Background () Background () Background (3) Background (3) Pituitrin inj. (3) Pituitrin inj. (3) Pituitrin inj. (3) From brown to green Host 5 min Graft Time to change 30 min From green to brown Host 5 S min. 9 Graft 9hr.+ 2iomin.+ 60 Five normal animals were placed in darkness to attain the light state. Each was then stimulated in turn with a tetanizing current. At the end of the period of stimulation, when examined for an instant by the rays of a flashlight, the lizards were found to be darkening; several minutes later they were uniformly dark brown in colour. On the average, it required -2 min. after cessation of electrical stimulation for the five lizards to become maximally dark (in agreement with the time for melanophore dispersion shown in Text-fig. 2 and in Table II). 3 0

14 Studies in Reptilian Colour Changes 87 These results might be interpreted as being due to reflex stimulation of the pituitary gland with consequent liberation of the melanophore-dispersing hormone into the circulation. When, however, the method of stimulation was slightly changed, the same results were not obtained. In the first modification both points of a bipolar pencil electrode were placed in the cloaca; with the second method one electrode was inserted in the cloaca and the other into the spinal cord. Although the intensity and duration of the stimulation were the same as those which had caused generalized darkening in the original experiments, in neither of these cases was generalized darkening effected; the only observable change was a darkening of the post-orbital regions and of scattered groups of scales over the otherwise green body. It appears unlikely that the generalized darkening was due to reflex stimulation of the pituitary; such melanophore dispersion resulting from electrical excitation was probably due to direct stimulation of the pituitary. If this be the case it is difficult to understand the darkening of scattered groups of scales which occurred with the modified methods of electrical stimulation. Such results could be explained on the basis of a lower concentration threshold of these regions for the melanophoredispersing principle. To test this last possibility more critically, four hypophysectomized lizards were stimulated electrically, one of the electrodes being placed in the cloaca and the other in the mouth; the duration and the intensity of the stimulation were the same as those used in the earlier trials. Within two minutes after cessation of electrical stimulation the post-orbital regions of these individuals became intensely black; in addition, clusters of black spots appeared in an irregular pattern against the generalized green colour of the skin (PI. III). The black spots consisted of groups of from two to twenty or more scales. This pigmentary condition of the lizard I shall call mottled. The mottling pattern persists in an hypophysectomized animal for at least 20 min. after cessation of electrical stimulation. Within an hour it has usually disappeared and the lizard is again uniformly green. This experiment proves definitely that the generalized darkening obtained originally by electrical stimulation is dependent upon the pituitary and not at all upon a direct nervous agency; it also shows that the mottling pattern may be mediated by an entirely different agency, since the response was evoked in hypophysectomized lizards. The nature of the controlling mechanism for the mottling response was next investigated. VI. THE MOTTLING RESPONSE A. Behaviour of denervated skin Since the mottling pattern and darkening of the post-orbital areas are evoked under the same conditions, the latter regions were selected as convenient ones for experimental study. Denervation of the skin of this area on one side of an hypophysectomized lizard by excision and subsequent replacement in its original position did not prevent darkening on appropriate stimulation. Two minutes after the cessation of stimulation the control side began to darken while the graft on the left

15 L. H. KLEINHOLZ side was still green, but min. later the post-orbital regions of both sides were black. This condition persisted for at least an hour; at the end of 2 hr. both sides were again green. When the responses of denervated post-orbital patches to electrical stimulation were observed in several lizards on the third, fifth, tenth and eleventh days after transplantation (ample time for degeneration of severed nerve fibres) they were found not to differ from the control regions of the unoperated sides. Darkening of the post-orbital regions is therefore not under direct nervous control. B. The role of the adrenals In the absence of nervous mediation, it appeared that the mottling response might be regulated through the circulatory system. Darkening of the post-orbital regions and mottling of the body also appear when the animal is excited by mechanical manipulation, such as prodding the head with a glass rod, or pinching the snout with forceps. Von Geldern (92) has reported a similar colour change in the skin of males preparing for combat; Dr L. T. Evans, while studying territorial defence and the fighting behaviour of Anolis carolinensis, has observed the same pigmentary condition (personal communication). This reaction is apparently associated with conditions of emotional excitement in these lizards. Redfield (98) found that in Phrynosoma tetanization of the roof of the mouth caused generalized pallor, a condition which he called excitement pallor and which he attributed to the secretion of adrenalin. Sand (935) found that emotional excitement of the African chameleon, Lophosaura, provoked by pinching the feet of the animal with forceps, induced darkening of the skin within 2 min. To determine whether the mottling response might be evoked by adrenalin, five hypophysectomized lizards were each injected with 0-05 c.c. of Parke, Davis and Co. solution of adrenalin chloride, in a concentration of :0,000. The mottling pattern became evident with min., and 2 min. later stood out strikingly against the otherwise green colour of the body. A Ringer-solution extract of both adrenals from one Anolis was similarly effective in inducing the mottling pattern in a second lizard. Because these results are exact duplications of the pigmentary conditions resulting from' electrical and from mechanical stimulation of hypophysectomized specimens, they indicate that the hormone which appears in the circulation of stimulated lizards may be adrenalin. The assumption that the mottling reaction of Anolis is due to the secretion of adrenalin is substantiated by the following series of experiments. A lizard was hypophysectomized, and 6 hr. later, when recovery from the effects of the operation was complete, the animal was stimulated electrically in the standardized manner. The mottling pattern appeared within min. after cessation of electrical stimulation. On the following day the animal was anaesthetized and a bilateral adrenalectomy performed. Three hours later, when the mottling colouration provoked by the operation had worn off completely and the lizard was uniformly green, the animal was again stimulated. No mottling appeared, even for as long as 5 min. following stimulation. The animal was then injected with 0-05 c.c. of :0,000 adrenalin.

16 Studies in Reptilian Colour Changes 89 Four minutes later the mottling pattern stood out in strong contrast against the generalized green colour of the skin (PI. IV). Repetition of the same procedure on similarly prepared lizards consistently yielded the same results: electrical stimulation of individuals which had been both hypophysectomized and adrenalectomized failed to evoke any pigmentary change, while injection of adrenalin, in all cases, called forth the mottling response. In the third section of this paper, describing the colour variations in different regions of the body, it was stated that an animal in the dark state shows a dark quadrangular patch posterior to the orbit and a scattering of dark spots against the brown skin of the body. It is now evident that these darker regions constitute the mottling pattern. (Except for being slightly larger and more densely pigmented, the melanophores in the scales of the mottling pattern do not appear much different from those in non-mottling skin; see PI. I, figs. 5 and 6.) The above experiments demonstrated adrenal control of mottling, but since there is no reason to believe that adrenalin is being secreted when the lizard is adapting itself to a black background, it follows that the pituitary hormone which causes dispersion of the melanophore pigment of the body has a similar effect on the melanophores of the scales composing the mottling pattern. This conclusion is supported by the fact that the mottling pattern is distinctly recognizable in adrenalectomized lizards which darken in adaptation to an illuminated black background. VII. SUMMARY. Anolis becomes dark brown on a light-absorbing background and bright green on a light-dispersing background, the changes being accomplished by a dispersion or concentration of pigment within melanophores. 2. Blinded lizards lose the ability to respond to changes in backgrounds, but become brown when in light and green when kept in darkness. 3. Hypophysectomy results in permanent pallor. The brown colour can be temporarily elicited in such lizards by injection of appropriate extracts of the pituitary gland.. Denervated regions of skin undergo normal colour changes. 5. Electrical stimulation of hypophysectomized lizards evokes a mottled pattern. 6. Mottling is under hormonal and not under nervous regulation. REFERENCES ALLEN, B. M. (97). Biol. Bull. Wood's Hole, 32, 7. ATWBLL, W. J. (99). Science, 9, 8. (92). Endocrinology, 6, 22. BROCKE, E. (852). Denkickr. Akad. Wiss. Wien,, 79. VON BUDDENBROCK, W. (928). Grundriss der vergleichenden Pkytiologie. Berlin. CARLTON, F. C. (903). Proc. Amer. Acad. Arts Sci. 39, 259. VON GELDERN, C. E. (92). Proc. Calif. Acad. Sci. 0, 77. HADLEY, C. E. (929). Bull. Mus. comp. Zool. Harv. 69, 07.

17 90 L. H. KLEINHOLZ HADLEY, C. E. (93). J. exp. Zool. 58, 3*. HOGBEN, L. T. & MraviSH, L. (928). J. exp. Biol. 5, 295. HOGBEN, L. T. & SLOME, D. (93). Proe. roy. Soc. B, 08, 0. HOGBEN, L. T. & WINTON, F. R. (922 a). Proc. roy. Soc. B, 93, 38. (9226). Proc. roy. Soc. B, 9, 5. (923). Proc. roy. Soc. B, 96, 5. KELLEK, R. (895). PflUg. Arch. ges. Phytiol. 6, 2. KLEINHOLZ, L. H. (935). Biol. Bull. Wood's Holt, 69, 379. (936). Proc. Nat. Acad. Set., Wash., 22, 5. KHUKENBERG, C. F. W. (880). Vergl. physiol. Stud. Heidelberg,, 23. MAY, R. M. (92). J. exp. Biol., 539. NOBLE, G. K. & BRADLEY, H. T. (933). Biol. Bull. Wood's Hole, 6, 289. PARKER, G. H. (906). J. exp. Zool. 3, 0. (936). Color Changes of Animals in Relation to Nervous Activity. Philadelphia. PARKER, G. H. & STARRATT, S. A. (90). Proc. Amer. Acad. Arts Sci. 0, 57. REDFIELD, A. C. (98). J. exp. Zool. 26, 275. SAND, A. (935). Biol. Rev. 0, 36. SMITH, P. E. (96). Anat. Rec., 57. ZOOND, A. & EYRE, J. (93). Philos. Trans. B, 223, 27. EXPLANATION OF PLATES I-IV PLATE I Vertical sections of scales from lizards in their colour phases. Sections stained with Mallory's azan. Fig.. From a lizard in the brown state. The stratum corneum has been displaced in sectioning; the branches of the melanophores extend to the stratum germinativum. Fig." 2. In the brown phase the terminal processes of the melanophores form a thin line underneath the stratum germinativum. The lamellae of the leucophore layer are visible. Fig. 3. In the green condition the pigment of the melanophores has become concentrated in the cell bodies. Fig.. The distal processes of the melanophores, even though devoid of pigment, can be seen extending toward the epidermis. Fig. 5. From the post-orbital region of a lizard in the mottled state. Fig. 6. A post-orbital scale in the green phase. PLATE II Colour changes of transplanted skin. All figures are of the same lizard (serial number 9, Table IV B), photographed within the first 8 hr. after transplantation. Fig.. Twelve hours after the operation, the animal is brown and the graft is green. Fig. 2. About 0 hr. later, after having been kept on an illuminated black background; host»kin and graft are brown. Fig. 3. The lizard was in the dark phase when placed on a white background; 8 min. later the lizard became green but the graft remained brown. Fig.. Still being kept on the illuminated white background, the lizard became uniformly green less than an hour later. PLATE III The mottling pattern. All figures are photographs of the same animal. Fig.. Dorsal view of an hypophysectomized lizard. Fig. 2. Lateral view. Fig. 3. Dorsal view of the lizard, photographed immediately after electrical stimulation for 2 min. The mottling pattern has appeared. Fig.. Lateral view showing the intense darkening of the post-orbital region.

18 JOURNAL OF EXPERIMENTAL BIOLOGY, XV,. PLATE I KLEINHOLZ. STUDIES IN REPTILIAN COLOUR CHANGES (pp. 7 9).

19

20 JOURNAL OF EXPERIMENTAL BIOLOGY, XV,. PLATE II KLEINHOLZ. STUDIES IN REPTILIAN COLOUR CHANGES (pp. 7 9!).

21

22 JOURNAL OF EXPERIMENTAL BIOLOGY, XV,. PLATE III ]9 0 KLEINHOLZ. STUDIES IX REPTILIAN COLOUR CHANGES (pp. 7 9*-

23

24 JOURNAL OF EXPERIMENTAL BIOLOGY, XV,. PLATE IV #*'. \ KLEINHOLZ. STUDIES IN REPTILIAN COLOUR CHANGES (pp. 7 9).

25

26 Studies in Reptilian Colour Ctianges 9 Fig. 5. Lateral view of the lizard, 3 days after the skin of the post-orbital region had been denervated. The transplant and the host skin are green. Fig. 6. Lateral view of the animal after 2 min. of electrical stimulation. The mottling pattern has appeared and the denervated post-orbital patch is black. Fig. 7. The post-orbital region of the animal in fig. 2. Fig. 8. The darkened region after stimulation. Fig. 9. The denervated post-orbital patch in the green condition. Fig. 0. The same denervated post-orbital patch after electrical stimulation. PLATE IV Adrenal control of the mottling response. All figures are photographs of the same individual. Fig.. Hypophysectomized lizard in the green state. Fig. 2. The mottling response to electrical stimulation. Fig. 3. Photographed 3J hr. after the removal of both adrenals. The tail and pelvic region show a slight darkening, probably because of ligation and injury in that region. Fig.. The hvpophysectomized and adrenalectomized lizard after standard electrical stimulation. The mottling pattern has failed to appear. Fig. 5. Four minutes after the injection of 0-05 c.c. of :0,000 adrenalin. The mottling pattern has appeared.