THE PIGMENTARY EFFECTOR SYSTEM IV.—AFURTHER CONTRIBUTION TO THE ROLEOF PITUITARY SECRETION IN AMPHIBIANCOLOUR RESPONSE.*

BY LANCELOT T. HOGBEN, M.A., D.Sc., F.R.S.E., Assistant Directorof the Animal Breeding Research Department.

{From the Animal Breeding Research Department, The University, Edinburgh.}

CONTENTS

1. Introduction 2492. The Bionomic Aspect of Amphib-

ian Colour Response . . .251(a) The Pigmentary Effector

System of Amphibia . 251(i) Reactions of Melano-

phores to Natural Stimuli 252(r) Post-mortem Condition . 254(d) Colour Change in the

Breeding Season . . 255(«) Time Relations in Colour

Response . . . 2553. Further Experiments on the

Specificity and Localisation ofthe Melanophore Stimulant ofPituitary Extract . . .256

(a) Specificity — Experimentson T o a d s . . . . 256

(i) Localisation . . . 2574. The Effect of Pituitary Removal

and the Subsequent Adminis-tration of Pituitary Extract toUrodele Larvae . . . . 260

4. Effect of Pituitary Removal—contd.

(a) Technique of Hypophy-sectomy in Urodela . 260

(6) Effects of Pituitary Re-moval . . . . 261

(c) Subsequent Injection ofPituitary Extract . .261

{d) Regulation of the Chro-matic Function in Uro-dela 262

5. Further Experiments on the Re-lation of Nerves to AmphibianMelanophores . . . . 263

(a) Effect of Transection ofthe C.N.S. . . . 264

(6) Effect of Severing WholeNerve Supply to theHind" Limb . . . 264

6. The Chromatic Function in OtherVertebrates . . . . 266

7. Summary 2688. References 269

1. Introduction.

IN a series of papers (Hogben and Winton, 1922-23) alreadypublished, evidence has been presented in favour of an inter-pretation of the regulation of colour response in Amphibia bythe secretory activity of the pituitary gland. Previous con-

* Received July 1st, 1923.VOL. 1.—NO. 2. 249 R

Lancelot T. Hogbentributions were based exclusively on the study of the commonfrog (Rana temporarid): the salient conclusions arrived at wereas follows :—

(1) The synchronous pigmentary changes displayed by thecommon frog in consequence of the reciprocal " expansion"and " contraction " of the melanophores and xantholeucophoresare conditioned by a variety of natural stimuli of whichtemperature, humidity, and illumination are the most signifi-cant : warmth, dryness, and bright light promoting the pallorassociated with melanophore contraction, while cold, moisture,and shade induce the darkening of the skin which results frommelanophore "expansion."

(2) Extracts of the pituitary gland of Mammals, Birds,Reptiles, Amphibia, and Fishes have a specific local action onthe melanophores of the frog inducing the expanded condition ;the reaction is an extremely delicate indicator of pituitaryextract.

(3) While removal of the anterior lobe of the pituitarygland in the frog does not interrupt the normal sequence ofpigmentary response to natural factors, frogs remain afterremoval of the whole gland uniformly pale with maximalcontraction of the melanophores and complete expansion ofthe xantholeucophores in natural conditions, which invariablyinduce darkening of the skin in controls. After injection ofpituitary extracts the characteristic melanophore expansionis evoked in hypophysectomised frogs as in normal paleindividuals; but the former when subjected to such treat-ment regain pallor even when exposed to surroundings whichrepresent the optimum conditions for pallor in normal frogs.

(4) The colour responses of the frog are gradual, requiringperiods from an hour to several days to reach completion, afact which converges with the negative effects of nerve sectionand stimulation, and with the action of drugs on the hypophy-sectomised animal to the conclusion that the rhythm ofpigmentary change is regulated by endocrine rather thannervous influences.

Observations essentially confirmatory of the above but notyet published in full are mentioned by Krogh in his recentmonograph on the Capillaries (1922). In this paper the

250

The Pigmentary Effector System IVattempt will be made to extend such conclusions to otherAmphibia, and to supplement previous observations on minorissues which seemed to merit further investigation.

Acknowledgment is made to the trustees of the DixonFund for a grant in aid of the expenses incurred in thisresearch.

2. The Bionomic Aspect of Amphibian Colour Response.In setting forth an interpretation of colour response in the

Amphibia as a class, brief attention is due to the bionomicaspect of the problem. The subject has been reviewed sothoroughly by Fuchs (1914) that there is, however, only needto refer to more recent evidence and such data as have emergedfrom the author's own observations.

(a) The Pigmentary Effector System of Amphibia.'—As far asis known the cutaneous pigmentary effectors of the Amphibiaconsist of three practically ubiquitous types. These are (a)epidermal melanophores, (6) dermal melanophores, and (c)dermal xanthophores (xantholeucophores). In various speciesone or other of the above have been described as having lostthe power of reacting by visible change; but this fixity maybe due to the nature of the internal environment. Thus themelanophores of old black axolotls are always expanded innature; but, as will be seen, the effects of hypophysectomyshow that this state of affairs is not due to any inability tocontract. In addition to the cutaneous pigmentary organsenumerated above there are other pigment effectors of moredeep-seated situation, namely (a) the retinal pigment cells, and(b) the internal melanophores of many connective tissues,lining the peritoneum and lymph spaces. Except whenotherwise stated subsequent remarks apply only to thecutaneous, more especially the dermal (or corial) melanophores,whose activity constitutes the dominant factor in bodily colourchanges among Amphibia.

The exact manner in which the melanophores " contract"and " expand" has been the subject of much controversy.Some workers following v. Wittich (1854), Bimmermann (1878),Carnot (1896) hold that the melanophores are amoeboid,contracting bodily. This view has been advocated recently

251

Lancelot T. Hogbenby Hooker (1914). Others following Lister (1858), Miiller(i860), Kahn and Lieben (1907), Dawson (1920), affirm thatthe cell processes are not withdrawn in the contracted phase,which results purely from a migration of pigment granules fromthe cell processes into the central mass. The latter viewclaims more support from recent investigators. I haveexamined preparations of frog's skin with melanophorescontracted by the methods of Cajal, da Fano, and Golgi, andhave never obtained clear evidence that the " cell processes "are not in reality preformed intracellular canaliculae into whichpseudopodia are protruded, as Hooker maintains, duringexpansion. But in all living and fixed preparations of theskin of the frog and axolotl, I have repeatedly found isolatedclumps of granules left behind in the "cell processes" after themajority are seen to be aggregated in the central mass. Suchappearances are not readily explicable on any supposition, butthat contraction is due to a streaming of granules within cellprocesses which themselves remain static. Nevertheless, whileinclining to the view that melanophores do not contract andexpand sensu stricto, one may employ these terms as sanctionedby usage, without prejudice to the question which has beendiscussed, to describe the appearances respectively associatedwith pallor and darkening of the skin in Amphibia.

(&) Reactions of Melanophores to Natural Stimuli.—One ofthe most important factors involved in colour response in adultAmphibia is moisture. This of course does not enter into theresponses of larval and aquatic forms as a significant elementin the normal rhythm of pigmentary change. It seems to begenerally agreed that humidity promotes darkening of the skinin adult Anura, and my own observations indicate that, innature, moisture is the predominant factor which determinescolour change, since, except at ordinary temperatures, light anddarkness have no effect on frogs placed in dry surroundings.With the exception of Dawson's observations on Necturus(1920), it is universally affirmed that warmth promotes thecontracted phase and cold tends to induce darkening of theskin both in adult Anura (v. Wittich, Biedermann, Lieben,Hargitt, et al.), larval Anura (Cole, 1922), and adult Urodela(Flemming, Laurens). Another factor worthy of mention is

252

The Pigmentary Effector System IVthe oxygen content of the surrounding atmosphere. It isgenerally agreed that oxygen deficiency induces darkening,and my own observations on the frog agree in this respectwith those of others. In most adult Amphibia, a notableexception being Necturtis (Dawson), light, if it has any affect,promotes the contracted phase, darkness or dull backgroundthe expanded. I am inclined to believe that the distinctionbetween background and shade is greatly overstressed by somewriters. The response appears to be under visual controlboth in adult (Lister, Rogers) and larval (Babak, Laurens)Amphibia. Anuran tadpoles with the exception of Ranapipiens (Hooker) react in the same way as the adult bycontraction in strong light and expansion in darkness. InUrodele larvae (Laurens) there is a primary response to lightby expansion and to darkness by contraction, followed afterprolonged exposure by the more usual reaction of pallor in thepresence and darkening in the absence of light. It is thissecondary response which is abolished by removal of the eyeand section of the optic nerves.

In order to make clear the precautions necessary in con-trolling experiments on colour response, reference may be madeto the table given in the preceding paper of this series. Myexperiments on frogs and toads were performed by means ofglass containers suitable to contain a single individual, andcovered at the top with muslin held by a rubber band so thatfree interchange with the surrounding atmosphere was available,and any danger of vitiating the experiment by alteration inthe gaseous content of the inspired air was avoided. Suchcontainers were placed in a cold room and fitted with envelopesof black paper and moistened with addition of about 2 cm.depth of water to reproduce optimum conditions for darkening ;or they were placed in a warm room, thoroughly dried andwell illumined on a white background, when it was desiredto reproduce optimum conditions for pallor. From 6° to io° C.is a good "low" temperature, and 200 to 24° C. a suitable"warmth."

I have carried out experiments with normal and blindedfrogs and find that, as would be expected, blind frogs reactlike the seeing animal independently of light in warm dry

VOL. I.—NO. 2. 253 R 2

Lancelot T. Hogbenor cold wet situations. However, placed in water against apure white background under bright illumination at moderatetemperature 12° to 15° C, the mode of response in the blindedand seeing frogs is sharply contrasted : for the dermal melano-phores of the latter expand, while those of the normal animalcontract. My own observations on the adult frog are inagreement with those of Rogers (1906) on the adult Salamander,namely, that in the presence of moisture, within the limitingrange of temperature for which light is a significant factor,the dermal melanophores of normal frogs contract in brightlight, whereas those of the blinded frog remain permanently

FlG. I.—Melanophores of the web of the foot of two frogs (JZana temporaria) kept in waterunder strong artificial illumination on a pure white background at 18° C. : on the leftnormal seeing individual, on the right blinded animal.

expanded in light or shade alike. "Adaptation to back-ground " is thus under control of the organs of vision inAmphibia as in Fishes.

(c) Post-mortem Condition. — It is here necessary to add afurther comment on the post-mortem condition of the melano-phore, which is assumed by many other authors, followingLister, to be that of contraction. This is not an accuratestatement of the case. It is true that in general melanophoresare found contracted in dead frogs, the reason being that inlive frogs the melanophores always tend, unless the illuminationis very bright and the surroundings very dry, to be a littleexpanded. After observing the post-mortem condition ofseveral hundred frogs exposed to optimum conditions fordarkening or pallor, I can confidently affirm that melanophores

254

The Pigmentary Effector System IValways tend to respond in the opposite sense to the conditionprevailing at the time of death. If very pale animals diethe melanophores invariably display partial expansion as themoribund state becomes advanced. I mention this fact becauseit seems to invalidate the method recently adopted by Uyeno(1922) for studying the action of COS and oxygen upon themelanophores of the isolated limb.

(d) Colour Change in the Breeding Season.—It has long beennoticed that an intense darkening of the skin is characteristicof the frog in the breeding season. I have pointed out else-where that there is no need to postulate any direct connectionbetween this phenomenon and the physiological conditionincident to reproduction, since the breeding season, whenpairing takes place round the shady margins of muddy pondsat a comparatively low temperature, is pre-eminently the timewhen large numbers of frogs can be seen in nature subjectto the optimum stimuli for melanophore expansion—moisture-cold, shade. This season I separated four pairs of copulatingfrogs in a white tank in bright light from the stock which werekept in shade on a black background, with the result thatcomplete contraction of the dermal melanophores of the formerensued within forty-eight hours. It is a fact of commonobservation that during copulation the female is paler thanthe male. Examination of the skin microscopically shows,however, that the melanophores are equally expanded in normalconditions in both sexes, so that there would appear to be asecondary sexual distinction in the number of melanophoresper unit area in the two sexes, an observation which I have notseen recorded elsewhere.

(e) Time Relations in Colour Response among Amphibia.—Itis a remarkable fact that the extremely protracted latent periodwhich intervenes between the application of the appropriatestimuli and the production of a particular colour change inconsequence, has been studied by few investigators with greatcare or precision. Had more attention been paid to thisaspect of the problem, it can hardly be doubted that the searchfor an alternative to nervous control would have been madeat an earlier stage. In my experience darkening and pallorin response to the optimum conditions is never complete in

255

Lancelot T. Hogbenfrogs before at least one and a half hours, and more usuallysix to twenty-four hours have elapsed. The only exact observa-tions I have been able to find are those of Laurens (1915)on Urodele larvae. These are worth quoting as well toemphasise an important aspect of the problem which confrontsus at a later stage. The following quotation is taken fromLaurens's paper on Amblystoma larvae :—

" 1. Expansion of the melanophores of seeing larvse inthe light one and a half to two hours.

" 2. Expansion of the melanophores of eyeless larvse inthe light two to three hours, and contraction of themelanophores of seeing larvae in darkness two tothree hours.

" 3 . Contraction of the melanophores of eyeless larvae indarkness four to five hours.

"4. 'Secondary' contraction of the melanophores of seeinglarvae in the light three to five days, and ' secondary'expansion of the melanophores of seeing larvae indarkness five days or more."

Laurens does not give the temperature conditions to whichsuch periods are subject.

3. Further Experiments on the Specificity and Localisation ofthe Melanophore Stimulant of Pituitary Extract

(a) Specificity.—As the excitatory action of pituitary extractupon the melanophores of the frog (or toad) is an eminentlysuitable subject for class demonstration, it is perhaps advisableto take this opportunity of insisting upon the elementarycondition that to demonstrate the expansion of the melano-phores it is first necessary that they should be in the contractedphase; i.e. experiments should only be performed on frogskept in separate aerated containers (p. 253) in bright, warm,dry surroundings for several days before experiment, and, inany case, the web should always be examined under the lowpower of a microscope to ascertain whether the pigment cellsare fully contracted initially.

The two following experiments were carried out to test thereactivity of the toad's skin to pituitary extract. Toads areperhaps more erratic in their pigmentary reactions than frogs,

256

The Pigmentary Effector System IVbut prolonged exposure—a week, for instance—to optimumconditions for pallor (bright illumination, warmth, and dryness)or darkening of the skin (shade, cold, moisture) displays assharp a contrast between the extreme conditions which, as inthe frog, are respectively buff or lemon yellow at one end of thescale and a fuscous grey or coal-black at the other.

(i) Four pale toads (Bu/o vulgaris) previously kept fortwenty-four hours in dry well-illumined jars at 22° C. wereinjected into the peritoneum with 0.5 c.c. frog's Ringer solution.Four pale toads similarly treated received by intra-peritonealinjection 0.5 c.c. of a 1 per cent, boiled extract of ox pituitary(posterior lobe). At the end of an hour the melanophores werecompletely expanded in the latter series, the toads being darkgrey to black in appearance. The others remained pale. Thedarkening of the pituitary series lasted twelve hours.

(ii) Ten male toads of yellowish aspect were injected intothe peritoneum as follows : Nos. 1 to 5 with 0.5 c.c. 1 per cent,(ox) pituitary extract; No. 6 with 0.5 c.c. saline; and Nos. 7,8, 9, 10 respectively with 0.5 c.c. 1 per cent, extract of brain,spleen, liver, and salivary glands in Ringer (boiled). At theend of an hour Nos. 1 to 5 were darker than the rest whichremained pale, while the pituitary series assumed maximumdarkening of the skin within three hours from injection.

The last experiment supplements previous observations onthe tissue specificity of the pituitary melanophore response.The following additional experiment with frogs may also beincluded in this connection.

(iii) A male rabbit was decapitated, boiled extracts asbefore being made of brain, spleen, liver, testis, and pituitaryground up with sterile sand in Ringer's solution. Eight pale frogswere respectively injected in pairs with 0.5 c.c. of a 1 per cent,and a o. 1 per cent, extract of brain, spleen, liver, and testis byintraperitoneal admission. Another pair were injected respec-tively with 0.5 c.c. of a o. 1 per cent, and 0.01 per cent, extractof the pituitary. After half an hour the characteristic darkeningoccurred in the last named pair, the others remaining pale.

{b) Localisation of the Melanophore Stimulant.—Many of theexperiments in earlier contributions were carried out withMessrs Burroughs Wellcome's standardised liquid sterile extract

257

Lancelot T. Hogbenof ox posterior lobe (" infundin "). In an experiment previouslyrecorded no appreciable melanophore activity of anterior ascompared with posterior lobe extracts of the mammalian glandwas found. Since then I have made several tests, thoughdifficulties in securing the glands within three hours of killingmakes it difficult to interpret the results with confidence. Inthe following experiment the glands were removed from theircapsules and immersed in acetone within twenty minutes ofkilling. By this means they are reduced to a leatheryconsistency within half an hour of immersion in the dehydratingfluid, so that the pars anterior, intermedia, and nervosa can bereadily dissected without the contamination inevitably resultingfrom their viscous texture in the fresh gland. Portions of thethree regions (pars anterior, intermedia, and nervosa) wereseparated from half a dozen pituitaries so treated, dried andweighed, boiled with a known volume of Ringer's fluid fortesting. Three series each consisting of six pale frogs

Amount ofdried tissuein gram3.

O O I

0005

O-OOI

00005

00001

000005

0000025

OOOOOI

0000005

00000025

Series I.(Pars anterior).

BlackMelanophores reticulate

...

BlackReticulate

GreyStellate

GreyStellate

PaleContracted

PaleContracted

...

Series II.(Pars intermedia).

...

BlackReticulate

...

BlackReticulate

...

BlackReticulate

BlackReticulate

PaleContracted

PaleSlightly stellate

Series III.(Pars nervosa).

...

BlackReticulate

BlackReticulate

...

GreyStellate

GreyStellate

...

GreyStellate

PaleContracted

258

The Pigmentary Effector System IVweighing 17 to 20 gms. were used for injection of dosescorresponding to different quantities of the three extracts.The melanophores of the web were in each case completelycontracted at the beginning of the experiment. The gross andmicroscopic appearances exhibited three hours after injectionare set forth in the table on preceding page.

A second experiment with different samples on the samethree series of frogs gave comparable results to those set forthin this table. From this it would seem that while there ismore of the melanophore stimulant in extracts of mammalianpars intermedia, it is by no means lacking from preparations ofthe anterior lobe. The melanophore test could not thereforebe legitimately used, as it was at first hoped, to detect thepresence of posterior lobe secretions in anterior lobe extracts.

The bearing of such experiments on the question of thelocalisation of the pituitary secretions is one which cannot bediscussed profitably until comparative tests of the pressor,oxytocic, and melanophore activities of different samples havebeen carried out, as is intended, simultaneously. It must beremembered that the secretory products may not improbablydiffuse with comparative facility from one region to another,and a relatively higher activity of extracts prepared from agiven region does not necessarily imply that such a portionrepresents the seat of origin of the active substance. It isby implantation and extirpation rather than the administrationof extracts that the localisation of the active constituents ofpituitary preparations will be understood. Removal of theanterior lobe of the frog as recorded in the third paper of thisseries (Hogben and Winton, 1923), in agreement with theunpublished experiments of Krogh and Rehberg {pp. cit.) doesnot result in the condition of permanent pallor consequent uponcomplete hypophysectomy. The melanophore stimulant is nottherefore uniquely confined to the anterior lobe. It may thenbe manufactured either by the posterior lobe, diffusing into theanterior portion, or, indifferently, by both parts of the gland.The latter supposition is not welcome from the standpoint ofhistology; and Swingle's (1921) experiments on the implanta-tion of the various portions of the amphibian pituitary intoAnuran larvae favours the former alternative, viz., that the

259

Lancelot T. Hogbenmelanophore stimulant originates in the pars intermedia anddiffuses thence into the other portions of the gland.

One other point is of sufficient importance to merit brieftreatment in this place. If the melanophore stimulant isnormally present in relatively large quantities in the anteriorlobe, its relation to the pressor and oxytocic substancesemerges in a new form. It has already been pointed outthat there are several indications which militate againstidentifying the melanophore with the pressor principle. Inthe light of Herring's work these considerations are enhanced;and the identity of the melanophore and oxytocic substanceswhich remained a more open question becomes less likely.However, this question requires simultaneous examination ofthe relative potencies of extracts from different regions byall three tests, and on the evidence, so far available, it isinadvisable to state more than a suggestion.

4. The Effect of Pituitary Removal and the SubsequentAdministration of Pituitary Extract to Urodele Larvae.

In experiments upon the effects of pituitary injections onaxolotls {Amblystoma tigrinuni), it was noticed that such pro-cedure is followed by a very noticeable intensification of thenormal black body colour of the melanic variety, in which themelanophores of older individuals (one to four years) in normalconditions display a condition of stellate expansion giving placeto the reticulate condition after treatment with pituitary extract.It seemed therefore likely that pituitary secretion might beinvolved in the regulation of colour response in Urodeles aswell as in Anura. To test this possibility the effects ofpituitary removal on the pigmentary responses of axolotlswere investigated.

(a) Technique of Hypophysectomy in Urodela.—I have notbeen able to find any reference to previous attempts to removethe pituitary in Urodela, despite the relative facility of theoperation on these animals. The gape of the mouth in axolotlsis extremely propitious for this purpose. When it is distendedby spring forceps with pointed extremities bent outwards, thepituitary is readily visible as a prominent white spot beneaththe thin integument and translucent cartilage of the roof of

260

The Pigmentary Effector System IVthe mouth. The morphological characteristics of the organhave been made the subject of a recent study by Atwell, towhose paper the reader may refer for further information.

Ether was found to be the most satisfactory anaesthetic.The axolotl is placed on its back and its mouth distended inthe manner indicated. To penetrate the thin parasphenoidabove the gland by means of a dental drill fitted with a rose-burr of suitable size is a practically instantaneous proceeding.The pituitary now exposed is loosened by means of a fine cataractneedle, and a pipette attached to a filter pump applied to theaperture sucks up the gland leaving a clear view of the brainuninjured below. The whole operation does not require morethan two minutes, when anaesthesia is accomplished. Axolotlsrecover from the anaesthetic best, if placed in running water ofinsufficient depth to cover the body. The wound heals rapidly.

{b) Effects of Pituitary Removal.—Twenty axolotls of themelanic variety ranging from nine months to four years in agewere treated by the above procedure. The results of completeremoval are uniform. There was no operative mortality, andsince several are alive six months after treatment, one mayinfer that viability is not impaired. Removal of the pituitaryin Amphibia as a class is not attended by the characteristicallyfatal disturbances incidental to the operation in Mammalia.

Twenty-four hours after operation the animals are visiblymore pale. When several days had elapsed, they assumed anaspect never encountered to the writer's knowledge on nature—certainly not in aquaria. They now presented a pale greyappearance reminiscent of Smith's, Allen's, and Atwell'sdescriptions of the hypophysectomised tadpole. In thiscondition, which on microscopic examination was seen to resultfrom complete contraction of the melanophores (fig. 2), theaxolotls have remained for several months in the laboratory.

(c) Subsequent Injection of Pituitary Extract.—Introductionof a very small trace of pituitary extract into the peritoneumwas followed in every case by a complete general expansionof the cutaneous melanophores, so that the operated animalsregained the appearance of normal black axolotls. The effectis manifest two to four hours after injection, and may last forseveral days with sufficiently large doses. To make the

261

Lancelot T. Hogbenexperiment more rigorous, the entire series of experimentalanimals were injected tri-weekly for a month with amounts(0.2 c.c. of a 1 per cent, saline extract) large enough tomaintain continuously the black colour associated with melano-phore expansion throughout the period of injections. Withina few days after cessation of the treatment, however, the formerpallor was re-established.

As recorded elsewhere (Hogben and Crew, 1923—thisJournal), hypophysectomised axolotls were employed to test forthe presence of physiologically active substance in the foetuses ofsheep and oxen at different ages from fecundation. Since then Ihave had opportunity of testing—with positive results-—-for the

FIG. 2.—Dermal melanophores of the black Axolotl, age two years : on the left normalindividual, on the right hypophysectomiaed animal.

presence of the melanophore stimulant in an extract made fromthe pituitary of an abortive human foetus in the fourth month.

(d) Regulation of the Chromatic Function in Urodela.—Fromthe foregoing observations it may be inferred that pituitarysecretion in Urodela as well as Anura constitutes an essentialfactor in the regulation of colour response. From the surveyof the bionomics of amphibian response at the beginningof this paper, it seems legitimate to state that there are noestablished phenomena relative to pigmentary response in Anurathat necessarily presuppose the existence of any auxiliarymechanism. One may provisionally suggest that in Urodelesas in Anura pituitary secretion is controlled by variousagencies, including thermo-receptors, in the skin, and that itis reflexly inhibited by light acting on the retina. This wouldexplain why in the Salamander Diemyctilus (Rogers)—withinthe range of external conditions for which light is a significant

262

The Pigmentary Effector System IVfactor—section of the optic nerves resulted in permanentexpansion of the melanophores, though transection of the corddid not interrupt the normal rhythm of colour response.

However, the phenomena described by Laurens and sum-marised in para. 2, seems to indicate a more complex problemin the case of young salamander larvae. For in these con-tinued illumination operates in the reverse way on blindedindividuals, which, instead of remaining permanently dark,react to light by melanophore expansion and to darkness bypallor. Laurens was not aware of the significance of pituitarysecretion, and there is no need, therefore, to criticise hishypothesis of nervous control {vide infra, 5). The reversal ofresponse implies (a) a regulating mechanism subordinate tothe organs of vision, which we are now justified in identifying—predominantly at least—with pituitary activity ; (6) a modeof reaction depending on the effect of light on the skin.Laurens explains (6) as due to the direct reactivity of themelanophores to light. This, however, does not seem toharmonise with the permanent pallor resulting from pituitaryremoval. Moreover, there is another possibility. Parker(1904) has indicated considerations in favour of the view thatAmphibia possess photo-receptors in the skin. A possibleexplanation of the paradoxical results of blinding in youngaxolotls might therefore be attempted by supposing that colourresponse in Urodeles depends on a balance between pituitaryactivity and some other mechanism (possibly adrenal secretion)which promotes melanophore contraction, adjusted in such away that light acting prepotently on the retina inhibits theformer and activates the latter, while light acting on photo-receptors on the skin activates the former and inhibits the latter.

5. Further Experiments on the Relation of Nerves toAmphibian Melanophores.

In an earlier contribution (Hogben and Winton, 1922) thenegative results through stimulation and section of peripheralnerves were recorded in connection with the cutaneousmelanophores of the frog. No attempt was then made toreview the state of the evidence for the existence of nervousregulation in amphibian colour response. After referring to

263

Lancelot T. Hogbentwo experiments which throw some additional light on thequestion, further attention may be devoted to this possibility.

(a) Effect of Transection of the C.N.S.—Concerning theconsequence of spinal transection as with the results ofperipheral stimulation, the testimony of different observers isdiscordant. Needless to say any manipulations involvingtransection of the C. N.S. must avoid injury or disturbanceto the pituitary. In two pale frogs the C.N.S. was dividedrespectively immediately behind the optic lobes and themedulla, with utmost precaution to leave the pituitary untouched.For three days they were kept under observation in warmdry surroundings, the skin retaining its pallor consistently.The melanophores remained contracted till the animals werereplaced in cold moist surroundings, when the body uniformlyassumed the darkening characteristic of melanophore expansion.The accompanying text figure (fig. 3) shows the state of themelanophores twenty hours after operation.

(6) Effect of Severing Whole Nerve Supply to the Hind Limb.—In agreement with Hooker it was found (op. cit.) that sectionof the sciatic did not produce local melanophore expansion inthe limb on which the operation was performed. Hooker(1912) however, states, "Severing the lumbo-sacral plexus orremoval of the lower sympathetic ganglia does not affect thenormal and synchronous change of colour in Rana fusca.When, however, both these processes are carried out together,the lower part is at first slower in reacting (italics inserted). . . .These results demonstrate that there are two nerve centresfor the control of the melanophores. . . . " It is not easy tosee why these results demonstrate any such thing; but itseemed desirable to repeat the double procedure. When thiswas done on a pale frog, kept under warm dry conditions, nodifference between the two limbs was observed during a periodof two days, after which the animal was placed in coldwet surroundings. The pallor of the operated limb wasevident on the following day in contrast to the rest of thebody. The contrast disappeared after three days. It wasevident also that the leg was much swollen. Microscopicexamination showed almost complete stasis of the blood inthe much dilated capillaries of the web. This might be

364

The Pigmentary Effector System IVanticipated from Krogh's work on the sympathetic controlof capillary tone. Krogh goes so far as to state all " agencieswhich I have enumerated as giving rise to capillary dilatationwill also, when applied in sufficient strength, cause oedema,or stasis, or both." It is little cause for perplexity that theendocrine mechanism is prevented from exercising its normalefficiency when the channels through which it operates aresuspended by the indirect action of nervous decontrol of thecirculation.

It is not easy to prove a negative, and to deny thatamphibian melanophores are subject to direct innervation would

FIG. 3.—Melanophores of the web of a frog twenty hours after transectionof the C.N.S. immediately behind the mid-brain without damageto the pituitary: optimum conditions for pallor.

not be justified ; but in the light of the new evidence onecan approach the question without a strong predispositionin favour of the positive attitude. The problem has beeninvestigated in the past by methods which fail to discriminatebetween direct nervous control of the melanophores andconcomitant vasomotor consequences of the experimentalprocedure. And Spaeth, than whom no modern worker hasinsisted more strongly on the nervous control of colour responsein fishes, admits freely that there exist "no satisfactorydemonstrations of the nerve endings in melanophores of reptilesand amphibians." In no case have those who record positiveresults from nerve section and stimulation carried out theirexperiments with rigid controls under optimum external

VOL. I.—NO. 2. 265 S

Lancelot T. Hogbenconditions operating in the opposite sense to the recordedeffects. In no case have photographs of such observationsbeen published for the enlightenment of other investigators.Moreover, on no relevant points relating to nerve section andstimulation is there unanimity in the testimony of differentobservers. The only argument that can be adduced in favourof the positive view with general assent is that derived fromthe opposite effects produced on amphibian melanophoresrespectively by those drugs which stimulate or prevent nervousimpulses reaching plain muscle supplied by sympathetic nerves.And judging from the effect of drugs on the hypophysectomisedfrog (The Pigmentary Effector System III.), it would appearthat even if a nervous mechanism exists it plays no significantrdle in the rhythm of normal colour response in Amphibia.But the cogency of applying an analogy based purely onexperiment with plain muscle may be seriously questioned.

6. The Chromatic Function in Other Vertebrates.

Apart from Amphibia two other classes of vertebratesexhibit colour response. Of these there seems to be accreditedexperimental evidence with indisputable histological supportfor the view that in Fishes at least nervous impulses play asignificant r61e in regulating the chromatic function. However,v. Frisch (1912), one of the most recent investigators in thisfield, points out that the consequences of nerve section andstimulation are not such as to suggest that nervous control canadequately interpret the entire range of normal responses, butseem to imply some auxiliary mechanism, the nature of whichfor the present remains obscure. If, as Spaeth (1917) claims,pituitary extract induces contraction of fish melanophores, itis clear that the term comprises structures which displayimportant differences in their physiological properties; andunless Spaeth's observations are discredited, one can onlyconclude that the regulatory mechanism of colour response inFishes and Amphibia is fundamentally dissimilar.

It only remains to be asked, therefore, whether the interpreta-tion advocated for the Amphibia in this series of contributionscan be legitimately extended to the Reptilia. As in the caseof Amphibia there are no data derived from microscopic

266

The Pigmentary Effector System IVexamination to suggest that the melanophores are directlyinnervated. Unfortunately there has been little recent physio-logical enquiry into the chromatic function of Reptiles in spiteof the fact that the proverbial (and grossly exaggerated)reactions of the chameleon prompted many earlier investiga-tions. From the important work of Redfield (1918) on theAmerican lizard Phrynosoma, it would seem that the responseto light depends on the direct reactivity of the melanophoresthemselves and is a local phenomenon, while the response tonocuous stimulation ("excitement pallor") is due to adrenalsecretion. Redfield's experiments do not, however, provideany justification for the view that nervous influence plays adirect part in these proceedings, though the action of adrenalinappears to be local. Neither of the conclusions stated abovesuggest a close resemblance between the chromatic functionin Reptiles and Amphibia. In order to test the action ofpituitary extract on Reptilian chromatophores, eight chameleonswere placed by the author in a warm darkened chamber fortwo hours, by which time maximum pallor ensued. Four werethen injected each with 0.5 c.c. of 1 per cent. " infundin"(tested simultaneously on the frog). After half an hour, onehour, two and four hours had elapsed no change was visible.The injected chameleons remained as pale as the controls.The experiment was repeated with similar results. It doesnot appear, therefore, that Reptilian melanophores respond likethose of Amphibia to pituitary extract.

That there should be a different mechanism of colourresponse in the three groups of vertebrates which display thechromatic function is not so surprising as it might appear atfirst sight. For while in all vertebrate groups the predominantrdle in effecting changes in bodily colouration is discharged bythe black pigment cells, the number of cutaneous pigmentaryeffector organs (xanthophores, erythrophores, etc.) is fairlynumerous. In the same animal, e.g. the frog, there are twotypes of pigmentary effector organs—xantholeucophores andmelanophores—whose reactions are diametrically opposite inmost respects. It is not more remarkable that the " melano-phores " of Reptiles or Fishes should respond differently fromthose of Amphibia, than that the xantholeucophores of the frog

267

Lancelot T. Hogbenshould respond in a manner different from the reaction of themelanophores of the same species.

7. Summary.1. The contracted melanophores of both Anura (frogs and

toads) and Urodela (salamanders) react by maximal expansionto pituitary extract; the active substance in the latter does notappear to be confined invariably to the posterior lobe of themammalian gland. A positive melanophore response wasobtained from the gland of a human abortus in the fourth month.

2. The removal of the whole pituitary gland may beaccomplished in Urodele larvae (Amdfystoma tigrinum) at anyage (nine months to four years) without impairing theirviability. After complete removal of the pituitary in Urodelelarvae, as in adult and larval Anura, the melanophores remaincontracted and a state of permanent pallor ensues. Thenormal colour resulting from melanophore expansion can bere-established by injection of pituitary extract; but suchanimals regain pallor, although exposed to conditions in whichmelanophore expansion invariably occurs in normal individuals.

3. It is legitimate to conclude that pituitary secretion isthe main factor in regulating the chromatic function throughoutthe Amphibia as a class. Fluctuating pituitary secretion incorrelation with those conditions that evoke colour responsein the frog {cf. the third paper of this series) provides asatisfactory basis for the interpretation of all the accreditedbionomic data concerning colour response in adult Amphibia.Possibly adrenal secretion or some auxiliary mechanism playsa subsidiary part; but there are no satisfactory grounds forbelieving that nervous agencies directly influence amphibianmelanophores; and there is reason to believe that even ifamphibian melanophores are directly innervated, nervouscontrol is not significant to the normal rhythm of colour change.

4. The study of amphibian colour response provides evidencenot only of the presence of physiologically active substancesin the pituitary, but functional activity of the gland in theintact animal. It does not appear, however, that the interpreta-tion of colour response here put forward for Amphibia can beextended to Reptiles and Fishes.

268

The Pigmentary Effector System IV

8. References.Allen (1917), "Effects of the Extirpation of the Anterior Lobe of the Hypophysis of

Rana pipiens," Biol. Bull., 82.Atwell (1919), "On the Nature of the Pigmentation Changes following Hypophy-

sectomy in the Frog Larva," Science, 49.Atwell (1919), "The Development of the Hypophysis in the Anura," Anal. Rec, 16.Atwell (1921), "The Morphogenesis of the Hypophysis in the Tailed Amphibia,"

Anat. Rec, 22.Babak (1910), "Zur chromatischen Hautfunktion der Amphibien," Arch. ges.

Physiol., 181.Babak (1912), "Ueber den Einfluss des Nervensystems auf die Pigmentbildung,"

Centr. d Physiol., 26.Babak (1913), "Ueber den Einfluss des Lichtes auf die Vermehrung der Haut-

chromatophoren," Arch. ges. Physiol., 148.Ballowitz (1893), "Die Nervendigungen der Pigmentzellen, u.s.w.," Zeit wiss.

Zool., 66.Biedermann (1892), "Ueber den Farbenwechsel der FrSsche," Arch. ges. Physiol., 61.Cole (1922), "The Transplantation of Skin in Frog Tadpoles, etc," Journ. Exp.

Zool., 86.Dawson (1920), "The Integument of Necturus tnaculosus," Journ. of Morph., 84.Dutartre (1890), "Sur les changements de couleur chez la grenouille commune,''

C. R. Acad. Sci., 181, Paris.Ehrmann (1892), "Zur Kenntniss von der Entwickelung und Wanderung des

Pigments bei den Amphibien," Arch. Dermat. u. Syphilis, Jahrg. 24.Flemming (1897), "Weitere Bemerkungen iiber den Einfluss der Licht und

Temperatur auf die Farbung der Salamanderlarvae," Arch. f. mikr. Anat., 48.v. Frisch (1911-12), " BeitrSge zur Physiologie der Pigmentiellen in der Fischhaut,"

Pflugcfs Arch., 18av. Frisch (1911-12), "Ueber den Einfluss der Temperatur auf die schwarzen Pigment-

zellen der Fischerhaut," Biol. Centr., 81.v. Frisch (1911-12), "Ueber fSrbige Anpassung bei Fischen," Zool. Jahrb., 82.Fuchs (1906), "Physiologie der Pigmentzellen," Biol. Centr., 26.Fuchs (1914), " Der Farbenwechsel und die chromatische Hautfunktion der Tiere,"

Handb. d. vergl. Physiol., 8, 1548-53, Erste Halfte, Jena.Gamble and Keeble (1900), " Hippolyte variant, A Study in Colour Change,"

Q.J.M.S., 68.Hargitt (1912), "Behaviour and Colour Change in Tree Frogs," Journ. Anim.

Behav., 2.Hering and Hoyer (1869), "Ueber die Bewegungen der sternformigen Pigment-

zellen, u.s.w.," Centr. med Wissen., T.Hogben (1923), "A Method for the Removal of the Pituitary in Frogs and Toads,"

Quart. Journ. Exp. Physiol., l aHogben and Winton (1922), "Studies on the Pituitary—I. The Melanophore

Stimulant in Posterior Lobe Extracts," Biochem. Journ., 16.Hogben and Winton (1922), "The Pigmentary Effector System I.," Proc. Roy. Soc, 9aHogben and Winton (1922), "The Pigmentary Effector System II.," Proc. Roy.

Soc., 94.Hogben and Winton (1923), "The Pigmentary Effector System III.," Proc. Roy.

Soc., 96.Winton and Hogben (1923), "Studies on the Pituitary—II. Basal Metabolism in the

Hypophysectomised Frog," Quart. Journ. Exp. Physiol., 18.VOL. 1.—NO. 2. 269 S 2

Lancelot T. HogbenHogben and Winton (1923), "The Mechanism of Amphibian Colour Response,"

Abstr. Xlth Internat PhysioL Congress.Hooker (1912), "The Reactions of the Melanophores of Rana fusca in the Absence

of Nervous Control," Zeit. allg. PhysioL, 14.Kahn (1922), "Studien iiber die Innervation der Chromatophoren auf Grund gegen-

satzlicher Giftwirkungen," Pfinger's Arch., 106.Keeble and Gamble (1903-4), "The Colour Physiology of the Higher Crustacea,"

Phil. Trans. Roy. Soc. B., 196 and 19aKeller (1895), " Ueber den Farbenwechsel des Chamaleons," Pfluger's Arch., 61.Krogh (1922), "The Anatomy and Physiology of Capillaries," Yale University Press.Langley (1921), "The Autonomic Nervous System," Part 1, Cambridge.Laurens (1915-16), "The Reactions of the Melanophores of Amblystoma Larvas,"

Journ. Exp. Zool., 20 and 21.Lieben (1906), "Ueber die Wirkung von Extrakten chromafinnen Gewebes auf die

Pigmentzellen," Zentr. PhysioL, 20.Lister (1858), "On the Cutaneous Pigmentary System of the Frog," Phil. Trans.

Roy. Soc., 14aM'Cord and Allen (1917), "Evidence associating Pineal Gland Function with

Alteration in Pigmentation," Journ. Exp. Z00L, 2aMilne-Edwards (1834), "Ueber die Farbenveranderungen des Chamaleon," Arch. f.

Ana/., PhysioL, u. tviss. Med.Parker (1903), "The Skin and the Eyes as Receptive Organs to the Reactions

of Frogs to Light," Amer. Journ. PhysioL, 10.Parker (1906), "The Influence of Light and Heat on the Movement of the

Melanophore Pigment especially in Lizards," Journ. Exp. Zool., 8.Pouchet (1848), "Note sur la mutability de la coloration des Rainettes et sur la

structure microscopique de leur peau," C. R. Acad Set'., 26, Paris.Pouchet (1872), "Du r61e des nerfs dans les changements de colouration" (and other

memoirs, ibid.), Journ. d. PAnat et d. I. PhysioL, 8.Redfield (1918) "The Physiology of the Melanophores of the Horned Toad

Phrynosoma," Journ. Exp. Zool., 26.Rogers (1906), "A Chameleon-like Change in Diemyctilus," Biol. Bull., 10.van Rynberk (1906), Ueber den durch Chromatophoren bedingten Farbenwechsel

der Tiere," Ergeb. d. PhysioL, 5.Smith (1916), "The Effect of Hypophysectomy in the Early Embryo upon the

Growth and Development of the Frog," Ana/. Rec, 9.Smith (1920), "The Pigmentary and Endocrine Disturbances induced in the Anuran

Tadpole by the Early Ablation of the pars buccalis of the Hypophysis," Amer.Anat. Mem., No. 11.

Spaeth (1916), "Evidence proving the Melanophore to be a Disguised Form ofPlain Muscle," Journ. Exp. Zool., 20.

Spaeth (1918), "The Standardisation of Pituitary Extracts," Journ. Pharm. and Exp.Ther., 11.

Spaeth and Barbour (1917), "The Action of Epinephrin on Single physiologicallyIsolated Cells," Ibid., 9.

Sumner (1911), "The Adjustment of Flatfishes to Various Backgrounds," Journ. Exp.Zool., 10.

Swingle (1921), "The Relation of the pars intermedia of the Hypophysis toPigmentation Changes in Anuran Laws," Journ. Exp. Zool., 84.

v. Wittich (1854), "Die griine Farbe der Haut unserer Fro'sche," Arch. Anat.u. PhysioL

270