Web contact: pietsch@indiana.edu
ABSTRACT
Salamander larvae typically adapt their dermal melanophores to achieve camouflage, and it has been known for some time that removal of the eyes abolishes what is now known to be a neuroendocrine response. Here we survey the contribution of the optic system to the bright and dark camouflage reactions and report that:
- the stimulus depends on an interaction between the direct and the reflected light;
- an eye mounted atop the head and oriented vertically tended not to support camouflage, even though the animal responded to visual cues and was able to learn a vision- dependent task;
- deviating the transplanted eye off the vertical axis significantly enhanced the recovery of the camouflage reactions (bright and dark);
- amputating or reorienting the following parts of the brain did not abolish either the bright or the dark camouflage reaction: telencephalon, epithalamus, pretectum or tectum; whereas lesions of the ventral optic pathway blocked brightening;
- transection near the midbrain-hindbrain junction, well posterior to known optic terminals, retarded the dark reaction;
- when the latter lesion was combined with disconnection of the telencephalon and the epithalamus -- contrary to predictions from the lesions executed separately -- the animals lost the bright reaction;
- the hypophysis is necessary for darkening but supported darkening even though detached, displaced or reoriented;
- the pineal gland was not essential for the grosser aspects of camouflage in Ambystoma larvae but this gland may play an adjunctive role in fine-tuning.
In the species familiar to us, the camouflage reactions emerge in development with the onset of free feeding (as the animal makes the transition from embryo to larva, per se) remain vigorous through mid-larval stages, then gradually diminish and disappear with metamorphosis. Changes exhibited by the dermal melanophores, Laurens' principal focus, are very readily judged under the dissecting microscope. Indeed, his original descriptions, expanded upon in subsequent years (14) remain valid today. What he meant, and we shall mean, by 'bright' and 'dark' reactions may be appreciated in Fig. 1.
Most wild species of Ambystoma blanch after about an hour in total darkness (photographic darkroom conditions). When illuminated, even at intensities corresponding to moonlight (Laurens' characterization), the animals darken if the receptacle is black, brighten if it is white or assume various tawny hues in a transparent bowl. Eyeless animals will only darken when illuminated, and then irrespective of the reflectance of the background, a phenomenon some investigators attribute to the effects of non-specifically absorbed radiation (24). We tested the latter hypothesis in the course of this investigation.
As is the case with amphibians in general (23), darkening appears to depend on the pituitary gland: hypophysectomy cancels the dark reaction (4), as do pharmacological blockers of the melanophore-stimulating hormone (MSH) of the pars intermedia hypophysis (29). Some workers present evidence suggesting that the pineal body antagonizes the release of MSH (1, 2, 24) while others report that pinealectomy has no effect on pigmentation, explicitly of Ambystoma larvae (3). Our investigation included experiments on both the epithalamus and the hypothalamus.
The Ambystoma brain consists of a richly fasciculated medullary substance surrounding a confluent, ependyma-like central gray matter. Except for the giant Mauthner cell near the roots of cranial nerves VII and VIII (see ref. 22 for pictures and literature), the neurons of the central nervous system of even the adult salamander are difficult to distinguish, as such. Nuclei are indistinct fields of cells that can only be rigorously identified by tracing the origin or termination of fiber pathways, an undertaking beyond our resources and, obviously, outside the scope of the present investigation. We preceded our experiments with morphological analysis and foresaw little opportunity for minutely correlating structural damage with camouflage deficits. However, adequate landmarks exist for surgically isolating the region of the brain containing, if not all, certainly the vast majority of known optic terminal fields. Given attachments to the eyes, can this zone provide the necessary and sufficient neural conditions for camouflage? In addition, as is known from Herrick's classic analyses and more recently the application of various tracing techniques (6, 8, 9, 12, 13, 15,), the several optic tracts group into diverging formations, some pathways proceeding dorsally to the thalamus, tectum and pretectum, others bending posteriorly and remaining ventral to the cerebral aqueduct en route principally, although not exclusively, to terminals in the peduncle. Does the salamander without a tectum or other dorsally situated terminals exhibit camouflage reactions? What effects follow the interdiction of the territory of the ventral pathways'?
The animal quarters were a windowless, air-conditioned room equipped with an automatic timer light switch set to deliver alternating 12 h cycles of light and darkness. (Cycles of light and dark enhance the long-term viability of a colony in our experiences.) During some acute experiments, when prolonged survival was not a factor, animals were illuminated continuously throughout the observation period. Stock animals were kept in 1/20th Holtfreter's solution, in transparent receptacles, 30 cm or more from opaque objects and were exposed to 400-500 Lux (lumens/m square) during the light cycle. All animals were fed fresh suspensions of newly hatched, vigorously swimming brine shrimp embryos, the ration controlled by volume.
Transplanted eyes were donated from the subject itself or, when a third eye was involved, from a sibling (same egg clutch). To manipulate and maintain the orientation of a prospective eye transplant, an asymmetrical flap of periorbital skin was left attached to the dorsal quadrant of the excised globe. The sticky deep surface of the flap proved useful for initially holding the eye to the transplant site. Once in place, the eye was secured with a modified Stultz Brucke (30) or straddle bridge. The latter device was made with two unbent safety pins and a section of Tygon tubing. Immersed, Tygon has a refractive index close to that of the anesthetic fluid, thus permitting a clear view of the eye with the Brucke set in place. The smoothness, flexibility and slight convexity of the wetted material permitted final adjustments of the transplant's visual axis. In transplants to the orbit, the optic nerve stump was aimed at the optic foramen; i. e., an attempt made to reestablish the original visual axis in these operations. Prior to transplanting an eye atop the head, the membranous neurocranium was excised to expose the epithalamus and mesencephalic tectum; the donor eye was then removed, floated into position and its optic nerve stump abutted against the pretectum. Sham operations, craniotomies identical to the preparative stages of eye transplanting, were introduced into several experiments as controls.
Brain lesions were inflicted with the aid of the map shown in Fig. 2. Except where specified otherwise, the lesions were inflicted after opening the brain case and involved complete transection through the indicated plane. In some instances a tracing of the region, projected onto the operating field with a drawing tube, served as guide for placing the incision. The details and rationale of the experiments will be presented with the results.
Fig. 2. Surgical Map. For a more detailed legend go here.
Except for special tests, the camouflage reactions were stimulated by placing the subjects in a light chamber consisting of a bank of conventional radiological film viewers with circular fluorescent lamps and the diffusers removed. Mounted on the underside of a laboratory table, the viewers delivered 1400 (+/- 98 S.D..) Lux to a platform 53 cm below the inferior surface of the lamps. The floor of the animal receptacle (where precise Lux readings could not be made) was 50 cm from the light source. No detectable differences in the reactions of animals could be attributed to the specific location on the platform. Many experiments involved observation periods of several months. To provide light- dark cycles, and thus maintain the health of these subjects, the wiring of the light chamber was spliced into the automatically controlled circuit of the room.
To test individual bright reactions, we placed animals in bride's white Styrofoam 'Dixie' cups (6SJ12, Dart Company, Mason, Ml) containing 60 ml of 1/20th Holfreter's solution. In the light chamber, as measured below the fluid line, the luminance of the white cups was 360 (+/- 10 S.D.) Nits. Cups were set into recessed trays so that an experimental group could, when necessary, be transported as a whole. In some experiments, bright tests were conducted using the standard light chamber but with several animals pooled in a white plastic utility pan of luminance similar to the Styrofoam cups.
The dark phase of the reaction was tested by placing the subjects in black plastic utility pans whose maximum reflection was at a wavelength of 575 nm and whose luminance in the standard light chamber, below the fluid line, was 4.95 (+/- 1.02 S.D.) Nits. In addition, adjunct tests were conducted using brown polypropylene cups with maximum reflection at 600 nm and a luminance in the light chamber of 10 Nits.
Experiments were performed to evaluate horizontally directed illumination. A viewer of the type used in the standard light chamber was inverted on a table top and the receptacle was centered within the area circumscribed by the fluorescent lamp. In some of these experiments the subjects were exposed individually, in which case a clear polycarbonate tumbler served to contain the animal. During exposure, the wall of the tumbler was 9 cm from the surface of the lamp; the animal was at the horizontal level of the lamp with no illumination from above. Under these conditions, the illuminance at the surface of the clear tumbler was 2000 Lux (as compared with 1400 Lux in the standard light chamber). To assess the elevated intensity some subjects were put into white cups and placed in a special light chamber where the illumination was vertically directed but of 2800 Lux. Additional tests of lateral illumination were conducted in the inverted viewer but with groups of 7 animals en masse in a clear plastic canister.
Investigations were made into the following aspects of the camouflage reactions:
Concerning response times, animals reached a bright or dark plateau within an hour after transfer to a white or black receptacle, irrespective prior conditioning. However, the maximal bright or dark reaction was reached only after several days of exposure to a given background, and prior conditioning may affect this long-term response. Therefore, to control for the aforementioned variables, the following precautions were taken:
The critical experiments were conducted in volleys consisting of 3 normal larvae each, the small number so that photographing and checking could be carried out almost simultaneously. The three prospective test subjects of a volley, drawn from the same stock and exhibiting identical pigment patterns, were placed in clear glass finger bowls and arbitrarily designated A, B and C. Subject A was decanted into a clear polycarbonate tumbler, and B and C into standard white cups. Subject A was set at the center of a circular fluorescent tube; B was placed in the standard light chamber for vertical illumination (1400 Lux); C was exposed to vertical illumination but at 2800 Lux. In each experiment, C reacted as B; i.e., the elevated illumination in the special chamber did not influence pigmentation. Other comparisions are most succinctly illustrated by focusing on the details of one specific experiment.
Animal A retained its stock melanophore index during a 7-day period of observation. B, reacting typically, had become bright during the first hour of exposure (Figs. 4 and 5). On day 7, A and B traded places; we closely monitored them for an additional two days. In the clear cup, receiving lateral illumination, B reverted to its tawny stock coloration. A, now in a white cup, brightened within the hour and maintained its bright coloration throughout the observation period (see Figs. 6 and 7).
Results of the type of experiment described in the last paragraph were replicated in several seasons for A. opacum and A. punctatum. Less rigorously monitored experiments were conducted but with batches of 7 larvae per group in a common container. Again, laterally directed illumination failed to stimulate brightening in any animal of any group. When the animals were transferred to white containers and placed in the standard light chamber they brightened in typical fashion within the hour.
| Operation | Number | POSITIVE SUBJECTS | |
|---|---|---|---|
| Cases | Percent | ||
| Unoperated | 33 | 33 | 100 |
| Sham-operated | 6 | 6 | 100 |
| One-eyed* | 27 | 27 | 100 |
| Eyeless** | 31 | 0 | 0 |
| Orthoclops | 12 | 10 | 83 |
| Contraclops | 13 | 9 | 69 |
| Triclops | 18 | 18 | 100 |
| Biclops | 10 | 10 | 100 |
| Cyclops-I | 49 | 4 | 8 |
| Cyclops-II | 33 | 10 | 30 |
*one natural eye removed, the other intact
**bilaterally enucleated (not to be confused with eyeless mutant axolotls)
Of 27 animals with one intact natural eye (one-eyed), all showed camouflage
reactions indistinguishable from the unoperated subjects. {Subsequent image analysis of pigment
spots reveal a subtle but statistically significant difference in the extent of
blanching between one-eyed and two-
eyed
animals.}
None of 31 bilaterally enucleated animals (eyeless) showed any sign of a bright
reaction, postoperatively.
Of 13 animals with one eye transplanted to the opposite orbit, and the other discarded (contraclops), 9 regained the bright reaction.; the balance never did so.
It seemed important to control for occluding the epiphysis with an eye situated atop the head, a de facto condition in the cyclops preparations. Two kinds of operations were performed to test for potential variables of this sort: (a) triclops (32), subjects with both natural eyes intact and a sibling's eye transplanted to the dorsum of the head; (b) biclops, with the dorsally mounted eye but one natural eye removed. Biclops subjects served as controls for the possibility in triclops that intact visual pathways might override any deleterious or negative consequences of the ectopic eye.
All triclops and biclops subjects showed normal camouflage reactions.
In cyclops-II, an attempt was made to tip the eye off the vertical axis. Among 33 cyclops-II subjects, 10 eventually exhibited the bright reaction (see Figs. 9 and 10); 23 did not.
The bright-competent orthoclops, the contraclops and the cyclops also showed normal dark reactions when tested in black receptacles. Eye transplant recipients that failed to recover the bright reaction by the sixth week postoperatively never again exhibited brightening.
| SUBJECTS | Number | Trial 1 | Trial 2 | Both Trials |
|---|---|---|---|---|
| Unoperated | 6 | 6 | 5 | 5 |
| One-eyed | 4 | 4 | 4 | 4 |
| Eyeless | 8 | 0 | 0 | 0 |
| Orthoclops | 12 | 11 | 11 | 11 |
| Cyclops-I | 13 | 7 | 5 | 3 |
| Cyclops-II | 15 | 7 | 7 | 4 |
Of 6 unoperated animals, all responded positively on at least one Fresnel trial; 5 did so on both. The 4 one-eyed subjects were positive to the Fresnel test on both trials. None of 8 eyeless animals made a positive response. Among orthoclops, 11 of 12 reacted positively on at least one trial and 10 on both. Of the 13 cyclops-I subjects, 7 responded positively on the first trial, 5 on the second and 3 on both. Of 15 cyclops-II animals, 7 responded positively on the first trial, 7 on the second and 4 on both. The Fresnel test data were pooled for each subgroup, and, for purposes of comparison with the camouflage reactions, were converted to percentages. The comparison is depicted in fig. 11. Among unoperated, one-eyed, eyeless and orthoclops subjects, the values for the Fresnel test approximated those of the camouflage response. With cyclops-I, however, where less than 10% of the subjects showed a positive camouflage reaction, their collective Fresnel test scores reached the 50% level. Among the cyclops-II subjects, 30% exhibited positive camouflage reactions, and their Fresnel test values were 51%.
One camouflage-positive cyclops-II had been among this series; it was included to check against the possibility that, in cyclops, camouflage competency might prevent light/shock avoidance learning.
Table 3 summarizes the light/shock avoidance results:
| SUBJECTS | N | AVOIDANCES Mean (+/- S.D.) | Significance Level* (versus controls) |
|---|---|---|---|
| Eyeless | 3 | 0 | |
| Controls | 3 | 47 (12.2) | |
| Orthoclops | 3 | 40.0 (8.0) | not significant |
| Camouflage-Negative Cyclops | 3 | 93.3 (9.2) | 0.01 |
| Camouflage-Positive Cyclops | 1 | 60 |
Subjects were A. opacum larvae from the same egg clutch. Testing was conducted 3 months postoperatively with 40 trials per day for 4 consecutive days; the data represent avoidances in 160 trials per subject.
*t-test
In larval Ambystoma, the optic nerves enter the ventrolateral wall of the diencephalon. The optic chiasm is external, as in higher vertebrates. Uncrossed projections have been detected with tracer techniques (see especially ref. 6) but most optic nerve fibers decussate upon entering the brain stem (see Fig. 12). Reaching the contralateral side, the fibers in question bend sharply out of the horizontal plane and, as separate pathways with numerous minor branches, diverge to five major regions:
From preliminary analyses of our own slides, we concluded that reproducible operations, in quantity, would not be possible with a map based strictly upon the conventionally recognized primitive regions of the brain. However, sufficient landmarks existed for dividing the brain into the regions A, B and C shown in Fig. 2, regions which approximate but are not coextensive with the fore-, mid- and hindbrain. Plane I passed from a point just posterior to the fundus of the pineal body through the zone between the anterior commissure and the optic chiasm. Plane II extended dorsoventrally from just in front of the cerebellum through the anterior extreme of the metencephalon, well posterior to the peduncle. Incisions I and II are depicted in Fig. 13. Region C, as a whole, was inoperable with the brain in situ and manipulations of it were deferred. {C can be transplanted either to the head of an animal with an intact medulla or to another larva's dorsal fin.} Other lesions are represented by Arabic numerals in Fig. 2. Their limitations and rationale will be taken up in context.
Fortuitously, plane I fell on an imaginary line which, if extended to the posterior poles of the eyes, would form the base of a low isosceles triangle whose sides were occupied by the intracranial segments of the optic nerves. In the standard operation (with the cranium opened), it was possible to identify -- and then avoid -- the optic nerves (see Fig. 13).
The experiments explicitly focused on region A included:
The exceptional cases were from a group of five A. punctatum larvae that had been used to evaluate a nonstandard means of inflicting incision I directly through the skin without first opening the cranium. The two subjects in question began darkening immediately after surgery. Suspecting damage to the optic chiasm of the latter subjects, we fixed the group. Slides revealed that the lesion had passed anterior to the optic chiasm in animals with their camouflage reactions intact but posterior to it in the subjects that had begun darkened immediately after surgery (Figs. 18 and 19).
Of over 100 subjects with incision II alone, all but three runted A. punctatum larvae (25 mm) assumed bright coloration when tested in white cups. (The exceptions were from non-standard operations; they permanently lost the bright reaction, and we eventually concluded that their optic pathways were damaged.) No A. opacum larvae with incision II were capable of a dark response (Figs. 22 and 23). But A. punctatum and A. tigrinum larvae with this lesion exhibited a spectrum of deficits ranging from no dark reaction at all to the barest perceptible difference from the controls.
A. punctatum larvae, 25 mm, were chosen, their small size because the endolymphatic sacs were readily visible; the species they seemed to exhibit more viability to lesion II than A. trigrinum; their egg clutches are typically large, thus ensuring, at the planning stage, adequate numbers of siblings. Prior to the main experiments, and for control purposes, skin incisions alone were inflicted in animals from the same stock as the prospective test group; this operation had no effect on the camouflage reactions. The definitive experiment was conducted twice.
An entire stock group raised in the same canister was subjected en masse to preoperative bright and dark testing with the intent of culling any animals lacking normal camouflage reactions (none had to be discarded). The group was seined, rinsed with fresh Holtfreter's solution and transferred to MS 222. Approximately half their number, selected at random, received incision II (quick version), and the others were left unoperated. Simultaneously, the operated and unoperated subgroups were rinsed, transferred to white pans and placed in the standard light chamber for 24 h to ensure that bright reaction remained intact. Then both subgroups were simultaneously poured into black pans of equal dimensions and returned to the light chamber, side by side, for 3 days of continuous illumination. Animals of each subgroup then were examined individually and quickly sorted into one of 5 lots according to the Hogben-Slome indices of their first three dermatomes. Lots were counted after sorting the subgroup.
In the first series, 30 unoperated subjects showed a mean Hogben-Slome index of 4.0 (+/- 0.83 S.D). The corresponding 26 subjects with incision II showed a Hogben-Slome index of 3.3 (+/- 0.49 S.D). In the second experiment, in addition to the latter mentioned data, the group was analyzed preoperatively as well, wherin the subjects exhibited a mean preoperative Hogben-Slome index of 4.7 (+/-1.98 S.D). The unoperated subgroup of 16 subjects showed a mean Hogben-Slome index of 4.7 (+/- 0.48 S.D). Twelve animals with incision II (2 died prior to evaluation) showed a Hogben-Slome index of 3.1 (+/- 1.33 S.D). F and t tests on the data indicated no significant differences between the preoperative and unoperative sets of values. In both experiments, the differences between operated and unoperated subjects was significant beyond the 99% level. Therefore, although the effects of incision II were less devastating than with A. opacum, the lesion tended to retard the dark reaction even of A. punctatum.
Still, the outcome of some pinealectomies deserves special note. A chance observation had been made that not all pinealectomized subjects seemed to darken in brown cups. We decided to conduct the 'brown test' more carefully. The experiment was carried out with 6 matched A. opacum larvae, three each with a simple pinealectomy or a sham operations. The animals showed typical reactions when tested in white cups or black pans, two days postoperatively. The brown test was initiated 4 days postoperatively after 24 h of brightening, and the animals were then monitored for two days. By 3 h, the controls had attained a Hogben-Slome index of about 3.5, as did one pinealectomized subject. Two pinealectomized animals failed to darken at all during the observation period. When transferred to black receptacles, the latter pinealectomized animals, which had failed the brown test, did darken within a few hours.
Now, although the fundus of the pineal body was conspicuous, and is easily plucked out, the stalk was invisible during an operation. Also uncertain was just how much of the epithalamus and the nearby pretectum had been spared or lost, and the histological data were inconclusive. Therefore, we conducted the experiments collectively denoted by 1 in Fig. 2. Often performed in conjunction with the tectectomies (lesion 3) to be reported on below, these experiments included:
Represented as 3 in Fig. 2, the experiments on the tectum, per se, spared the pineal and adjacent epiphysis; the IIIrd ventricle remained covered; but the operation involved excision of the roof of the cerebral aqueduct, the exicised piece being either discarded or reimplantated in some instances orthotopically, in others with the anterior-posterior poles reversed. All subjects exhibited bright and dark reactions indistinguishable from the unoperated control animals, both acutely and for periods of up to three months (Figs. 26-29).
Lesions 4 was inflicted from the dorsal approach, after craniotomy, with the aid of a drawing tube map and on an imaginary line between the anterior borders of the endolymphatic sacs. The failure of tectectomy to block the camouflage reaction justified injuring the tectum. Seven 30 mm A. punctatum larvae received lesion 4, were placed in white cups after surgery and were observed for 2 months. Five immediately and permanently lost the bright reaction whereas 2 remained indistinguishable from the normal controls. An acute back-up series was performed for histological analysis in which 4 of 4 subjects lost the bright reaction. Fixed three days postoperatively, these specimens showed lesions passing through the caudal peduncular region and the corpus mamillare area of the hypothalamus;the optic nerves and chiasm were intact.
The postchiasmal commissure appears as an opalescent transverse strand when the underside of the neurocranium is appropriately illuminated. Comparisons between slides and dissections indicated that the latter structure could serve as a landmark for placing a lesion in the ventral optic funiculi without cutting the optic chiasm per se. With the aid of an eyepiece reticle, lesion 5 was inflicted through the roof of the mouth, 200 um behind the postchiasmal commissure. Placed in white cups after surgery, the animals began darkening immediately. Examination of sagittal sections showed that lesion 5 passed through the root of the infundibulum, transected the medullary substance between the optic chiasm and the peduncle and disrupted the peduncular gray matter (Fig. 30,); the optic chiasm itself was intact.
Removing region either B or AB as a unit (including the hypophysis but disconnecting the optic nerve) caused brightening of animals reared in black pans. Returning either B or AB reinstated darkening; replacing A alone did not.
Series of experiments were conducted with genetically eyeless axolotls, mutant animals that, unlike the principle species, reportedly (and confirmed pari passu by us) do not blanch in total darkness (4). With region B removed, the eyeless axolotls became bright. Region B was reimplanted into eyeless axolotls, in some instances orthotopically and in others with the ventral surface facing up. Both kinds of subjects were able to darken.
Additional experiments were performed on eyeless mutants with region AB reimplanted either orthotopically or with the ventral surface facing up. These animals were placed in clear tumblers and atop the diffuser of an X-ray viewer in order that illumination would come from below. The subjects darkened equally well whether AB faced up or down. Transferring the subjects to the standard light chamber, and thus reversing the direction of illumination, had no effects on their coloration.
Further comparisons were made between region B and the hypophysis with a group of 28 mm A. tigrinum from a single, fortuitously large clutch of eggs. The animals were kept in light during the entire (acute) course of the experiment. Controls for lighting included groups of normal animals maintained throughout in either black or white pans. The test subjects were initially divided into three lots:
The B-less controls remained bright in black pans. But within a few hours postoperatively the recipients of either B or the hypophysis had darkened (Figs. 33 and 34).
The latter experiments were extended as follows. On the next morning, region B was transferred back to the original donors, the latter animals having become bright in the interim. By evening, the original donors were again dark, but now the hosts were bright. The exchange was repeated among half the animals in question. By morning, the pigment patterns had reversed again. The animals in Fig. 1 were the subjects of the re- reversal procedure.
The cyclops subjects that lacked the bright reaction nevertheless out-performed the control and the orthoclops animals by 2-fold. Just such unpredictable high scores were obtained with cyclops in previous studies, which included the testing of triclops (32). In the latter investigations, the triclops acquisition rates were elevated above normal by precisely the increment attributable to the increased visual input, and the data seemed predictable by the Weber-Fechner law. Paradoxically, however, the cyclops, instead of learning more slowly than the triclops (or normal) animals outscored triclopes by an enormous amount (173 versus 117, on a scale with normal = 100). We concluded then that much more is involved in the visual perception of even a simple salamander larva than can be ascertained strictly from overt behavior; that our tests had failed to reveal what must have been the dampening effects of active inhibition (the active-negative mode) mediated by the natural eyes of triclops but absent in cyclops. We suggested then that a vertically oriented eye lacks the geometry to perceive a full range of visual cues. Poor camouflage reactions among cyclops-I and the appreciably higher recovery rate of brightening among cyclops-II are consistent with the latter hypothesis. Simply put, for the camouflage reaction, the visual fields of cyclops-II worked better than those of cyclops-I.
Our findings also support an hypothesis advanced by others(4, 24) to explain darkening among eyeless animals; namely that the effect is a net consequence of nonspecifically absorbed radiation.
{Non-specific absorption would appear to induce the release of MSH. In total darkness, with, for practical purposes, no radiation being absorbed, the MSH-releasing mechanisms would relax, and brightening could opccur. In the bright reaction, the MSH-releasing mechanisms, whatever their source, presumably would be inhibited. See text, below.}
After manipulating regions B or AB, in which the optic nerves had been cut, we found that darkening occurred equally well whether B or AB faced toward or away from the light source. The activation threshold of the MSH system must been exceedingly low, and given the capacity to manufacture melanin, it would appear that a normal Ambystoma larva would darken except for the intervention of signals evoked at the retina. We suggest that the 'strategy' the camouflage network is the selective and progressive inhibition of the mechanisms associated with the release of MSH. The hypothalamus appears to be involved in the latter process (see literature and discussion in refs. 4 and 29). It is also well known that the pars intermedia of the hypophysis is important in metamorphosis (23). Thus camouflage may provide a model for studies with implications beyond vision as such.
By itself, incision I canceled neither the bright nor the dark reaction. Alone, incision II exerted its effects on the dark reaction, and these subjects brightened. If the dorsally situated optic terminals play no role in camouflage, then I+II should have had the same effects as II alone: the animals should have brightened. Instead, they invariably lost the bright reaction. Casually considered, this finding seemed illogical and we unsuccessfully sought to explain it by assuming that the optic nerves or optic chiasm had been inadvertently damaged. Microscopic inspection failed to support this assumption. Moreover, in each operating session the controls included some incision I alone, all of which subjects retained the bright reaction. Proceeding from the evidence, we were forced, first to admit that I+II had unmasked otherwise hidden contributions to brightening from region A and, secondly, to treat formally the logic in I, II and I+II. The key considerations of the latter analyses are useful to this discussion and will be presented in the next paragraph.
The data following incision II alone seem noteworthy on several counts.