1. Melanophore contraction is not effected through a direct nervous agency, nor are the adrenal glands normally necessary to mediate the light state.

  2. The concentrated state of melanophore pigment is ordinarily due to the disappearance of the melanophore-dispersing hormone from the circulation. The rapid onset of pallor which follows electrical stimulation may be due to a vasoconstrictor effect.

  3. Intact skin in hypophysectomized lizards does not respond directly to light, but isolated skin darkens slightly in bright sunlight and becomes bright green again in diffuse light.

  4. Isolated skin responds to hormones, becoming generally dark in solutions of pituitary extract and showing the mottling reaction in solutions of adrenalin.

The evidence presented in the previous paper pointed conclusively to the pituitary control of the dispersed state of the melanophores in Anolis carolinensis. In this section are reported the factors regulating the concentration of melanophore pigment and the chromatic behaviour of isolated skin.

A. Mediating agents for melanophore concentration

Hypophyseal regulation of melanophore dispersion has been reported for amphibians (Hogben & Winton, 1922a, 1922b, 1923), for elasmobranchs (Lundstrom & Bard, 1932; Hogben, 1936), and for cyclostomes (Young, 1935). The explanation of colour changes offered by Hogben and Winton for amphibians and that suggested by Lundstrom and Bard for elasmobranchs (Mustelis canis) were, for a time, accepted as the only control necessary for such pigmentary behaviour. Recently, however, this problem has become controversial. Hogben & Slome (1931) found evidence for a dual endocrine control of colour changes in the African toad, Xenopus laevis; the response to white backgrounds, in their opinion, was due to a hormone from the pars tuberalis and that to black backgrounds was due to the melanophore-dispersing principle of the pars intermedia. Similarly, in a later paper, Hogben (1936) reported that removal of the anterior lobe of the pituitary in elasmobranchs (ScyIlium canicula, S. catulus, and Raia brachiura) appeared to abolish the white-background response. Parker & Porter (1934) were of the opinion that the light phase in Mustelis canis was induced through the action of concentrating nerves and was not merely the result of absence of the melanophore-dispersing hormone. Wykes (1936), on the contrary, could find no evidence for direct innervation of the melanophores in elasmobranchs (Scyllium canícula, Raia brachyura, R. maculata and Rhina squatina), confirming the earlier investigation of Young (1933), who reported the absence of sympathetic innervation of the skin and chromatopho res in seven different species of elasmobranchs.

The controversy has been almost as great in the field of reptilian colour changes. Carlton (1903) believed that in Anolis the contracted melanophore was in the resting condition. Redfield (1918) held that the pale phase in Phrynosoma was effected by nerves or by adrenalin. Hogben & Mirvish (1928) showed that the hormonal mechanism which Redfield found in Phrynosoma was not involved in the pigmentary behaviour of Chameleo, in the African chameleon concentration of melanophore pigment was accomplished by sympathetic nerves.

In analysing the mechanism whereby the fight phase of colour response in Anolis is effected, the various possibilities mentioned above must be considered. Thus, the possible mediating agencies for this response are (1) the direct action of nerves; (2) reflex stimulation of the adrenal glands; (3) both adrenalin and the direct action of nerves (Redfield’s hypothesis), with either alone being capable of causing melanophore contraction; (4) some other humoral control. The evidence presented in the following sections will be applied toward the above-mentioned possibilities.

B. The possibility of a nervous control

The possibility of a direct nervous agency to effect the light phase in Anolis is not supported by any experimental evidence. Given areas of skin which had been denervated in the process of transplantation did not remain dark, as might be expected if the concentration of melanophore pigment were dependent upon nervous action. On the contrary (as may be seen from Table IV in the previous paper of this series), the skin grafts changed from brown to green along with the colour change in host-skin. Although there was a distinct time difference in the changes of the two, this may reasonably be accounted for by the locally disturbed blood supply of the grafted areas.

Transection of the spinal cord at the level of the 10th to the 13th vertebra did not retard concentration of the melanophore pigment in the posterior half of the animal. Proof that the green colour is not effected by local reflex stimulation of the melanophores is given by similar preparations where, in addition to cutting the spinal cord and the sympathetic chains, the posterior half of the cord was destroyed by pithing. In all lizards thus treated the ability to adapt to backgrounds seemed in no way impaired, the animal as a whole (both the denervated and the innervated portions) turning brown on an illuminated black background and green on an illuminated white background.

Lizards in the dark condition were stimulated electrically. In one series, both electrodes from the secondary coil were inserted into the cloaca of a dark Anolis which was kept on an illuminated black background ; in another group one needle-electrode was inserted into the spinal cord and the other into the muscle of the body wall, near the cord. As in the first series, the dark lizards were kept on an illuminated black background, and several minutes were allowed to elapse before beginning stimulation in order to allow the animal, if it had blanched because of handling, to revert to the dark phase. In each case generalized pallor was produced within 2 min. after the beginning of electrical stimulation, later accompanied by the appearance of the mottling pattern. To test more critically whether nerves are directly concerned in effecting this result, lizards in which given regions of the body had been denervated were stimulated as before. Any possibility that nerves are directly involved in mediating the pale phase is negated by the results from such experiments : in all cases the melanophores in the denervated as well as those in the normal regions of the body were contracted by the electric current.

C. The possibility of adrenal control

These results, which exclude nerves as the mediating agents of the green phase, point, consequently, to a circulatory mechanism by which concentration of melanophore pigment is brought about. A possible explanation is offered by the appearance of the mottling pattern (in addition to generalized pallor) in those experiments where normal lizards were stimulated electrically, suggesting the liberation of adrenalin into the blood stream.

Injection of adrenalin intraperitoneally into dark lizards (kept on illuminated black backgrounds) induces rapid concentration of the melanophore pigment. This persists for periods roughly proportional to the concentration and amount of solution injected. Similar results have been reported by Hadley (1931) for Anolis iodurus and by Redfield (1918) for A. carolinensis and for Phrynosoma comutum. To determine whether the adrenals are normally necessary to bring about the green colour, three otherwise normal lizards were adrenalectomized. When these were tested for their ability to change colour, their chromatic responses to background changes were found to be in no way different from those of normal control lizards.

The following protocol supplements the above observations:

The experiments described in this and in the preceding section show definitely that contraction of the melanophores is neither dependent upon the direct action of nerves nor upon adrenal secretion. Redfield’s hypothesis for melanophore concentration in Phrynosoma does not, from the experimental evidence presented here, apply to the pallor-inducing mechanism for A. carolinensis. The experiments do, however, indicate that blanching in Anolis is regulated through the circulation.

D. Circulatory control of the light phase

The regulation of melanophore concentration through the blood stream may be accomplished in either of two ways. Either a hitherto-unidentified hormone is liberated into the circulation under appropriate conditions, or the melanophore-dispersing hormone of the pituitary is inactivated or disappears from the blood.

The existence of the “W” substance, claimed by Hogben to be the agent effecting the white-background response, is not yet conclusively established. In explaining abolition of the white-background response in elasmobranchs, Hogben (1936) says (p. 156): “According to the evidence obtained in…Amphibia ‘W’ is associated directly or indirectly with the pars tuberalis which in Xenopus as in Elasmobranchs is not morphologically distinct from the pars anterior. The results of removing the anterior lobe of Elasmobranch fishes support the same hypothesis though they do not conclusively establish its truth.”

There is little evidence in support of a homology between the pars anterior of elasmobranchs and the pars tuberalis of the higher vertebrates. In fact, De Beer (1926) states (p. 75): “The origin of the ventral lobes of the pituitary from the lateral lobes of the hypophysis suggests that these elements are the homologues of the pars tuberalis of higher forms.”

The evidence that the pars tuberalis is the source of a melanophore-concentrating hormone is as yet incomplete. In adult A. carolinensis there is no anatomically distinct pars tuberalis. It is possible that a portion of some other lobe may be homologous with this structure; the question can be decided by study of appropriate embryonic stages of this lizard.

That melanophores may be concentrated by a hitherto-unidentified hormone has not been disproved, and is still open to proof. But the fact that the melanophores are concentrated after hypophysectomy and in pieces of isolated skin shows that the concentrated condition of the melanophores is the “resting” state and makes it unnecessary to assume the presence of a melanophore-concentrating principle; melanophore pigment becomes concentrated because of the absence of the pituitary pigmentary hormone from the vicinity of the chromatophore. It will be recalled that in studying the pale phase, experiments were described in which electrical stimulation brought about generalized pallor in a very short time, usually within 3 or 4 min. The rapidity of this response is in contrast to the much slower white-background response (see Text-fig. 2 of the preceding paper). The explanation for this lies, I believe, in the nature of stimulation. The lizard pales slowly on a white background because melanophore-dispersing hormone gradually disappears from the blood stream; the rapidly appearing pallor which is characteristic of electrical stimulation is due, in my opinion, to the generalized vaso-constriction resulting from the stimulation, thereby resulting in an effective exclusion of the melanophore-dispersing hormone from the pigmentary effectors.

A. Absence of direct response in intact skin

The contribution which a study of the behaviour of isolated skin offers toward solving the problem of colour changes in the intact animal is of problematical value. In the one case we are dealing with the animal as a whole, the response being the result of a co-ordinating system, whereas with isolated skin the melanophores have been removed from the environment which is normal to their behaviour, and have been placed in an entirely artificial situation. It is obvious that factors which influence colour changes in isolated skin may not necessarily hold for the intact animal.

A study of the behaviour of isolated skin was undertaken for comparison with similar studies reported by Hadley (1928, 1931). This investigator found that isolated skin from four species of Anolis, A. equestris, A. iodurus, A. carolinensis and A. watsoni, showed responses to light, becoming brown in direct sunlight; in ordinary diffuse illumination, however, the skin remained green. These results were confirmed by Smith (1929) for the isolated skin of A. equestris. Sand (1935) has been reluctant to accept these direct observations “on purely theoretical grounds”. The implications of these observations are at once evident. If intact skin in A. carolinensis is directly responsive to light, then the mechanism for melanophore dispersion has been inadequately presented in the above sections.

The skin in normal intact lizards was tested for a local response to light. Three animals were placed on an illuminated white background until they had attained the light state. A light-opaque screen, fashioned from modelling clay, was placed on their backs, covering most of the trunk region, after which the lizards were transferred to an illuminated black background. The light screen was removed after the exposed portions of the body had turned brown. The areas of skin which had been covered and which, if the melanophores were directly responsive to light, should have remained green, were as brown as the uncovered regions of the body. Such results are in agreement with a hormonal control of melanophore dispersion. As an additional test along these lines, three hypophysectomized Anolis were placed in direct sunlight; none of them darkened. The unavoidable conclusion is that in intact skin of A. carolinensis the melanophores are not directly responsive to light.

The ability of blinded animals to darken in the light must consequently be due to the presence of dermal photoreceptors which reflexly transmit the stimulus to the pituitary, because, as has been described in an earlier section, blinded lizards which are hypophysectomized no longer darken in the light; similarly, hypophysectomized lizards which are blinded remain green in the light.

B. Direct response of isolated skin to light

In studying the responses of isolated skin of this lizard, however, results were obtained which tend to confirm Hadley’s observations. Isolated skin turned green and remained green when floated on the surface of cold-blooded Ringer’s solution. When a portion of excised skin was placed in direct sunlight (in front of a window) it darkened slightly. In such experiments the skin never became brown, indicative of maximum dispersion of the melanophore pigment, as was reported by Hadley (1928) and by Smith (1929) for A. equestris; instead it became a dark, muddy green after being for 1 min. in bright sunlight. When such darkened skin was moved from bright into diffuse light it became bright green again, usually within 30 sec. after the transfer. A normal lizard which was used as a control to check the reactions of the isolated skin remained dark both in the bright and in the diffuse sunlight. The local nature of this response was excellently demonstrated when one half of a strip of isolated skin in Ringer’s solution was shaded and the second half exposed to bright sunlight; the shaded portion remained bright green while the brightly illuminated region became darker. By focusing sunlight with a hand-lens (similar to the method used by Hadley) on an isolated strip of skin, the precise illuminated spot became dark brown. This response, however, proved to be irreversible, the area remaining brown for more than 7 hr. subsequent to its removal to diffuse light.

When isolated skin was similarly tested under bright artificial light (a photographer’s “photoflood” lamp), under a source of ultra-violet light (a mercury-vapour lamp), and under rays combined from both these sources, the skin remained bright green, not showing the change that occurred typically in bright sunlight. These responses of isolated skin to bright sunlight are unintelligible without further investigation; they may be due to a slight dispersion of the melanophore pigment, or even to changes in the xanthophore or oil-droplet layer in the skin. The above experiments show that while isolated skin may respond directly to light by a local darkening, there is no evidence favouring a similar conclusion for the intact skin of this lizard.

C. Responses of isolated skin to hormones

Another peculiarity in the behaviour of isolated skin from A. iodurus reported by Hadley (1931) was the response to adrenalin. He found that various concentrations of adrenalin when injected into intact lizards always brought about the green coloration. When, however, isolated skin was floated on solutions of adrenalin, maximum expansion of the melanophores resulted.

Hadley says nothing of a mottling pattern in the species he used. It seems possible for such a pattern to be present in one species and absent in another; there may even be variation in these markings in individuals of a species, as is the case with A. carolinensis, the pattern being very pronounced and broadly distributed in some lizards while in others it is reduced to a few relatively scattered scales. Specimens of A. iodurus which Hadley used for injection of adrenalin may have lacked the mottling pattern, while the isolated skin may have been from individuals or from regions of the body in which such markings were present.

When isolated skin from A. carolinensis was tested for its responses to adrenalin, the anticipated dispersion of melanophores in the mottling pattern occurred (Pl. I). Isolated skin remained green in Ringer’s solution. If placed in diluted “pituitrin” or in extract prepared from fresh intermediate lobes, the skin became uniformly brown; when placed in diluted adrenalin the mottling pattern was evoked against the green background-colour of the rest of the skin. Adrenalin and “pituitrin” did not cause the same generalized darkening of isolated skin from A. carolinensis which they effected in skin from A. iodurus. With “pituitrin” the melanophore dispersion is general, whereas with adrenalin only the melanophores of the mottling pattern responded. I am inclined to believe that Hadley’s results might have been due to chance selection of skin from regions that were part of the mottling pattern, although such a statement is really unjustified without actual reinvestigation of the same species of lizard.

These experiments on the behaviour of isolated skin show that there is a direct reactivity to bright sunlight, due either to a slight dispersion of melanophore pigment or to possible changes in the xanthophore or oil-droplet layer. The response of isolated skin to light is slight. There is no evidence for a local response to light in intact skin of A. carolinensis. Isolated skin responds to “pituitrin” and to adrenalin in a fashion identical with that resulting from the injection of these hormones into intact lizards.

The formulation of a comprehensive theory for metachrosis in vertebrates must still be postponed. The controversy which exists in this field is indicative of the confusion that can be caused by generalizations based on studies of melanophore activity in a few species. Such generalizations have failed to be substantiated by subsequent investigations within the class. In reptiles the present data for the African chameleon indicate nervous control, but the activity of the melanophores of Anolis appears to be regulated exclusively by endocrines.

Carlton
,
F. C.
(
1903
).
Proc. Amer. Acad. Arts Sci
.
39
,
259
.
De Beer
,
G. R.
(
1926
).
The Comparative Anatomy, Histology and Development of the Pituitary Body
.
London
.
Hadley
,
C. E.
(
1928
).
Proc. Nat. Acad. Sci., Wash
.,
14
,
822
.
Hadley
,
C. E.
(
1931
).
J. exp. Zool
.
58
,
321
.
Hogben
,
L. T.
(
1936
).
Proc. roy. Soc. B
,
120
,
142
.
Hogben
,
L. T.
&
Mirvish
,
L.
(
1928
).
J. exp. Biol
.
5
,
295
.
Hogben
,
L. T.
&
Slome
,
D.
(
1931
).
Proc. roy. Soc. B
,
108
,
10
.
Hogben
,
L. T.
&
Winton
,
F. R.
(
1922a
).
Proc. roy. Soc. B
,
93
,
318
.
Hogben
,
L. T.
&
Winton
,
F. R.
(
1922b
).
Proc. roy. Soc. B
,
94
,
151
.
Hogben
,
L. T.
&
Winton
,
F. R.
(
1923
).
Proc. roy. Soc. B
,
95
,
15
.
Lundstrom
,
H. M.
&
Bard
,
P.
(
1932
).
Biol. Bull. Wood’s Hole
,
62
,
1
.
Noble
,
G. K.
&
Bradley
,
H. T.
(
1933
).
Biol. Bull. Wood’s Hole
,
64
,
289
.
Parker
,
G. H.
&
Porter
,
H.
(
1934
).
Biol. Bull. Wood’s Hole
,
66
,
30
.
Redfield
,
A. C.
(
1918
).
J. exp. Zool
.
26
,
275
.
Smith
,
D. C.
(
1929
).
Proc. Nat. Acad. Sci., Wash
.,
15
,
48
.
Wykes
,
U.
(
1936
).
J. exp. Biol
.
13
,
460
.
Young
,
J. Z.
(
1933
).
Quart. J. micr. Sci
.
75
,
571
.
Young
,
J. Z.
(
1935
).
J. Biol
.
12
,
254
.
Zoond
,
A.
&
Eyre
,
J.
(
1934
).
Philos. Trans. B
,
223
,
27
.

Behaviour of isolated skin. Figs. 1, 3 and 4 are photographs of the same strip of skin from one hind leg; figs. 2 and 5 are the same skin from the second hind leg, serving as control.

Fig. 1. The mottling pattern showing against the generalized green colour of the rest of the skin; mottling was effected by floating the skin on the surface of a solution of 1:5000 adrenalin.

Fig. 2. The control skin from the opposite leg, kept in Ringer’s solution, has remained green.

Fig. 3. The skin shown in fig. 1 was transferred to Ringer’s solution. The mottling pattern began to fade in 8 min. and at the end of 40 min. had faded completely, the skin being like the control shown in fig. 2.

Fig. 4. The same skin, 17 min. after transfer to 1:10 solution of “pituitrin “, has turned dark brown, a distinctly different colour from the mottling pattern.

Fig. 5. The control skin shown in fig. 2, 26 min. after transfer to 0 · 3 c.c. of Ringer extract of 1 pars intermedia from Anolis. The generalized brown colour is identical with that shown in fig. 4.

Fig. 1. The mottling pattern showing against the generalized green colour of the rest of the skin; mottling was effected by floating the skin on the surface of a solution of 1:5000 adrenalin.

Fig. 2. The control skin from the opposite leg, kept in Ringer’s solution, has remained green.

Fig. 3. The skin shown in fig. 1 was transferred to Ringer’s solution. The mottling pattern began to fade in 8 min. and at the end of 40 min. had faded completely, the skin being like the control shown in fig. 2.

Fig. 4. The same skin, 17 min. after transfer to 1:10 solution of “pituitrin “, has turned dark brown, a distinctly different colour from the mottling pattern.

Fig. 5. The control skin shown in fig. 2, 26 min. after transfer to 0 · 3 c.c. of Ringer extract of 1 pars intermedia from Anolis. The generalized brown colour is identical with that shown in fig. 4.