1. The eye grows by the addition of new ommatidia rather than by an increase in size of existing ones.

  2. New ommatidia develop as a result of induction of epidermal cells by functional ommatidia.

  3. The eye has great regenerative ability, but following removal of the entire eye and the growth zone that surround it regeneration does not occur.

  4. Transplanted eye cells or epidermal cells from normal-eyed or from lavender-eyed mutants resulted in the development of normal eye pigmentation in a mutant that lacked any pigmentation in the eye. Only the eye to which the graft had been made developed pigments. The development of pigments was a result of tissue contact, not of a blood-borne factor.

  5. When epidermal tissues from the prothorax of normal-eyed roaches were placed above the developing eye of a mutant with no pigment in the eye they were eventually incorporated into the eye as it grew. When these cells became eye cells, normal eye pigments were synthesized.

  6. Growth of the eye is by a constant recruitment of epidermal cells that express their own genetic capability when they become eye cells.

Periplaneta americana L., the American cockroach, has been studied by many investigators, but phenomena related to normal development of the compound eye and its capacity for regeneration have apparently not been the subject of careful investigation. These phenomena and the related one of the genetics of pigmentation in the eye are the subject of this paper.

In the exopterygotes investigated in embryonic and nymphal life (Aphis-Witlaczil, 1884; Rhodnius prolixusMellanby, 1937; Notonecta glauca-Lüdtke, 1940) the compound eye begins to differentiate at the posterior margin and differentiation progresses toward the future anterior margin of the eyes (Lew, 1933).

There are apparently only three previous experimental studies related to regeneration of the eye of P. americana. Bodenstein (1962) transplanted imaginal eyes to the prothoracic shields of nymphs and found that they molted in synchrony with the host and that a new cornea was formed. In a single paragraph Penzlin (1963) says that the eye is capable of some regeneration, but does not go into detail. Parabiotic experiments by Bodenstein (1959) in which the white-eyed mutant (also known as pearlRoss, Cochran & Smyth, 1964) was joined to a normal-eyed nymph failed to change the eye color of either partner. A similar experiment in which the lavender-eyed mutant was joined to the normal (Ross et al. 1964) failed to change the eye color of either partner.

Ross et al. found that in crosses between the white-eyed (pearl) and lavender that the hybrids had normally pigmented eyes as did hybrids between either lavender or pearl and normal-eyed roaches. They concluded that both pearl and lavender are autosomal recessives and are inherited independently.

Mutants of the strains studied by Bodenstein and Ross et al. were used in this study.

The experimental insects were kept at 24 °C, were fed commercial dog food and fresh apple and provided with water.

Since the two mutations had been shown to be autosomal by Ross et al. (1964), no selections were made on the basis of sex. In the first experiments, surgery was performed 4 days after an ecdysis, in order to be certain that the insect was well past the extreme sensitivity to the anesthetic that is characteristic just before and after ecdysis. In later experiments familiarity with the appearance of the eye during the intermolt made it possible to select animals that were far from an ecdysis.

In making the transplants, great care was exercised so that the graft and the tissue of the host fitted closely together. If the edge of the graft protrudes beneath the tissue of the host the epidermis of the latter tends to cover the graft and the grafted tissue heals independently.

Melted soft paraffin was used as a bandage. This bandage covered the area of the operation completely so that the grafts would be held firmly in position.

Tissues for histological examination were prepared by standard techniques.

P. americana L. is an exopterygote insect that passes through a variable (10–12) number of instars. During the course of this investigation the nymphal instars varied in duration from about 2 weeks (first one) to 2 months (last instars). One year passed from hatching to adulthood. This slow development makes the gradual changes taking place in the eye comparatively easy to follow. More than 200 nymphs, including several individuals of each instar, were studied in this experiment.

At the time of hatching the eye is triangular in shape. This triangle is nearly equilateral, with its apex at a point near the dorsal edge of the base of the antenna and the base of the triangle extending from the posterodorsal corner of the eye to the ventral tip of the eye which is slightly below the ventral edge of the border of the antenna. An unpigmented zone that is approximately equilateral in its dimensions completely surrounds the pigmented, functional ommatidia of the eye. The posterior edge of the eye has large ommatidia and the dorsal, anterior’ and ventral tip edges of the pigmented area show a pattern of smaller and more numerous ommatidia.

As the nymph grows, the growth of the eye is more extensive dorsally, anteriorly and ventrally than it is along the posterior edge. The eye of the adult more nearly approximates an isosceles triangle, the base of which is the dorsal edge of the eye with the apex of the triangle at the ventral tip. The eye of the adult (Fig. 4) is strongly curved anteriorly and extends almost half-way around the base of the antenna both dorsally and ventrally.

Fig. 1.

Shape and dimensions of the area of mature ommatidia and the maturation zone of the compound eye of representative nymphal stages and the newly molted imago (I). (Camera lucida drawings.) O, Ommatidia; Mz, maturation zone. Line indicates 1 mm.

Fig. 1.

Shape and dimensions of the area of mature ommatidia and the maturation zone of the compound eye of representative nymphal stages and the newly molted imago (I). (Camera lucida drawings.) O, Ommatidia; Mz, maturation zone. Line indicates 1 mm.

Fig. 2.

Regeneration in the compound eyes of nymphs from which a part of the mature area of ommatidia and the growth and maturation zones has been removed. (Semidiagrammatic.) M1, First post-operative molt; M2, second post-operative molt. Stippled area represents regenerated area.

Fig. 2.

Regeneration in the compound eyes of nymphs from which a part of the mature area of ommatidia and the growth and maturation zones has been removed. (Semidiagrammatic.) M1, First post-operative molt; M2, second post-operative molt. Stippled area represents regenerated area.

Fig. 3.

Transplantations between the wild-type and the lavender- and pearl-eyed mutants. (Semidiagrammatic.) Column 1, donor and source of graft (in rectangle). Column 2, host and location of graft (in rectangle). Column 3, result. Black eye represents wild type. Vertical bars represent lavender eye. No shading represents pearl eye. Stipples indicate development of wild-type pigmentation in the pearl grafts or eye in epidermis that was induced to become a part of the eye. In row A the graft is to the head of a wild-type nymph from which the eye had been removed.

Fig. 3.

Transplantations between the wild-type and the lavender- and pearl-eyed mutants. (Semidiagrammatic.) Column 1, donor and source of graft (in rectangle). Column 2, host and location of graft (in rectangle). Column 3, result. Black eye represents wild type. Vertical bars represent lavender eye. No shading represents pearl eye. Stipples indicate development of wild-type pigmentation in the pearl grafts or eye in epidermis that was induced to become a part of the eye. In row A the graft is to the head of a wild-type nymph from which the eye had been removed.

Fig. 4.

Normal imaginal eye of Periplaneta americana showing the anatomic relationships between the compound eye, the unpigmented ocellus, and the base of the antenna. A maturation zone surrounds the eye of this recently molted specimen,×10.

Fig. 4.

Normal imaginal eye of Periplaneta americana showing the anatomic relationships between the compound eye, the unpigmented ocellus, and the base of the antenna. A maturation zone surrounds the eye of this recently molted specimen,×10.

The eye of the adult covers a much greater proportion of the head than does that of the young nymph. In the first instar the eye extends dorsally and ventrally only slightly above and below the dorsal and ventral margins of the base of the antenna, but in the adult both the gena and vertex are greatly reduced in relation to the size of the eye.

The width of the unpigmented zone of the eyes of nymphs of various stages varies and there is a maturation zone around the eye of the newly ecdysed adult (Fig. 1). The insects used for Fig. 1 were selected at random before the exoskeleton had hardened and darkened following an ecdysis. The drawings were made 24 h later, after hardening and darkening were complete.

The maturation of the ommatidia is a continuous process in this insect. Observation of the eye of a newly ecdysed nymph of any instar (Fig. 5) as well as the newly molted adult reveals that the margin of the area of pigmented ommatidia appears to be serrate with that half of an ommatidium immediately adjacent to the mature area pigmented and its distal half unpigmented. As growth continues, these ommatidia become progressively pigmented and the margin of the mature area becomes smooth in outline. This process is first noticeable along the posterior margin of the eye at the point where the eye is narrowest – that is, a little below the mid-posterior edge. This is apparent by about 10 days following the ecdysis. The progressive pigmentation of the other edges of the eye continues slowly until the next ecdysis.

Fig. 5.

A part of the maturation zone at the dorsal edge of the eye of a newly molted third- or fourth-stage wild-type nymph showing the spread of pigment granules into the maturation zone. From a whole mount,×200.

Fig. 5.

A part of the maturation zone at the dorsal edge of the eye of a newly molted third- or fourth-stage wild-type nymph showing the spread of pigment granules into the maturation zone. From a whole mount,×200.

With the approach of the next molt the dorsal, anterior, and ventral tip sections of the area that was unpigmented previously are darkly pigmented. There remains a thin white band along the posterior edge. Growth is less extensive there at every instar than in the other areas.

The dimensions of the unpigmented zone just after a molt added to those of the mature ommatidia can be taken as a rough estimate of the dimensions of the mature area of ommatidia of the next instar. This cycle of maturation is repeated from instar to instar and includes the imago. The imaginal eye is not completely pigmented at the time of the larva–adult ecdysis, but the unpigmented zone surrounding the eye of the newly molted adult becomes pigmented over a period of not less than 2 weeks following ecdysis (Fig. 1).

Granular intracellular pigments can be seen beyond the original dimensions of the unpigmented band along the dorsal edge of the eye in nymphs of fifth and later instars. Appearance of these pigments indicates that the molt will take place within approximately 2 days and can be used for predicting the next molt.

As mentioned, the dimensions of the unpigmented zone of a given area change from one instar to the next. This area is widest in those areas which are growing most rapidly. The changes in the dimensions of the unpigmented zone are reflected in the progressive changes of the shape of the mature area of the eye.

The pattern of the maturation of ommatidia of different sizes, as outlined above, is maintained throughout the growth stages. Smaller and more numerous ommatidia are formed along the dorsal, anterior, and ventral tip margins ; larger and fewer ommatidia are formed along the mid-posterior edge.

Studies of frontal sections cut across the center of the eyes of newly molted sixth- or seventh-instar nymphs (Fig. 6) show that the functional ommatidia of the wild-type eye are densely packed with large pigmented granules that are particularly numerous along the length of the rhabdom and just underneath the cornea. Pigmented granules are located, as well, along the nerve fibers leading to the optic ganglion. The pigmented exocuticle of the head becomes gradually thinner as it approaches the cornea.

Fig. 6.

Frontal section of eye of newly molted wild-type nymph showing the arrangement of pigment granules along the rhabdom and underneath the cornea, (a) Growth zone, (b) maturation zone, (c) mature ommatidia.

Fig. 6.

Frontal section of eye of newly molted wild-type nymph showing the arrangement of pigment granules along the rhabdom and underneath the cornea, (a) Growth zone, (b) maturation zone, (c) mature ommatidia.

The margin of the eye shows ommatidia in various stages of maturation. Those nearest the densely pigmented and presumably functional ommatidia have lenses and are pigmented on one side only, the side nearest the mature ommatidia. Farther from the mature ommatidia can be seen tightly packed elongated cells, the long axes of which are oriented at right angles to the body surface. These cells have large elongate nuclei. The proximal ends of the cells curve beneath the functional part of the eye.

Farther away from the eye, occasional mitotic figures which are oriented so that the division products will be at right angles to the body surface, can be seen. These divisions can be seen in newly molted nymphs as well as in those later in the instar, but have not been observed in areas far from the eye in newly molted nymphs (Fig. 6).

The mitotic region surrounding the eye will be called the growth zone, and the unpigmented zone seen in the living animal will be designated the zone of maturation.

In order to test the capacity of the eye to regenerate and in order to see whether the pattern of growth as outlined above could be interrupted, a series of experiments involving removal of one-half of the mature ommatidia of the eye was performed. In every case, whether or not the maturation zone was left intact, extirpation was followed in the succeeding molt by a rapid intrusion of epidermis, followed, in succeeding ecdyses, by a gradual return of the eye to its normal shape and size, with ommatidia of normal sizes being formed as outlined above (See Figs. 2 and 7).

Fig. 7.

Eye of an imago from which the dorsal half of the mature ommatidia was removed at the third or fourth instar,×10.

Fig. 7.

Eye of an imago from which the dorsal half of the mature ommatidia was removed at the third or fourth instar,×10.

These extirpations were made from the dorsal one-half, the ventral one-half and the anterior and posterior halves. The only major distinction in the pattern of returning to normal was that at the first post-operative ecdysis there was a wider maturation zone at the anterior edge than obtained as a result of the other operations.

In the event that a few intact pigmented cells were left after an operation, a new maturation zone formed completely around these cells and an additional eye developed as a result. (See Fig. 8.)

Fig. 8.

Eye of an imago from which the dorsal half of the mature ommatidia was removed at the third or fourth instar and that developed two small dorsal areas of ommatidia. One of these was posterodorsal, and the other anterodorsal. The antero-dorsal one became incorporated into the eye.×10.

Fig. 8.

Eye of an imago from which the dorsal half of the mature ommatidia was removed at the third or fourth instar and that developed two small dorsal areas of ommatidia. One of these was posterodorsal, and the other anterodorsal. The antero-dorsal one became incorporated into the eye.×10.

The eye therefore has great capacity for regeneration, but in an additional experiment the removal of the entire eye, mature ommatidia, maturation zone, and growth zone was followed by the side of the head being covered completely with epidermis and exoskeleton typical of the head and there was no regeneration of the eye (See Fig. 9).

Fig. 9.

Right side of the head of an imago from which the entire eye was removed at the third or fourth instar,×10.

Fig. 9.

Right side of the head of an imago from which the entire eye was removed at the third or fourth instar,×10.

Fig. 10.

Grafted eye of a wild-type imago from which the eye was completely removed at the third or fourth instar. After the nymph had molted twice a rectangular graft from the center of the eye of a late instar pearl imago was placed into the area from which the eye had been removed. The graft grew, changed shape, and developed wild-type pigmentation,×50. (See also Fig. 3 A.)

Fig. 10.

Grafted eye of a wild-type imago from which the eye was completely removed at the third or fourth instar. After the nymph had molted twice a rectangular graft from the center of the eye of a late instar pearl imago was placed into the area from which the eye had been removed. The graft grew, changed shape, and developed wild-type pigmentation,×50. (See also Fig. 3 A.)

The experiments are summarized diagrammatically in Fig. 3. In the first experiment the eye was entirely removed from animals with normally pigmented eyes. They were allowed to ecdyse twice in order to be certain that the eye was entirely absent and then a rectangle of tissue from the center of the eye of a pearl nymph was placed into the location of the original eye. These grafts grew, gradually changed shape to approach the normal and became pigmented normally. The pigment appeared first near the edge of the graft and gradually spread inward. (See Figs. 3 A and 10.)

As a further test, tissues from the eye were placed on the prothorax. These grafts grew, gradually widening toward the anterior end of the host, and in every case where a pearl graft was placed into either a lavender or a normal-eyed host the graft developed pigments. The eyes of pearl hosts receiving grafts of either lavender or normal tissues did not develop pigments, and lavender eye tissues did not develop normal pigments, regardless of the host.

An additional test was the exchange of one-half of the eye, including the growth and maturation zone, between similarly sized individuals of the different genetic combinations. In every case the pearl half of the eye, whether host or graft, developed normal pigments. The other eye of a pearl host was never affected. See Figs. 11–13.

Fig. 11.

The dorsal one-half of the right eye is from a graft from a wild-type nymph. The wild-type pigment has begun to develop in the pearl half of the eye. The left eye is unpigmented,×10. (See also Fig. 3E.)

Fig. 11.

The dorsal one-half of the right eye is from a graft from a wild-type nymph. The wild-type pigment has begun to develop in the pearl half of the eye. The left eye is unpigmented,×10. (See also Fig. 3E.)

Fig. 12.

Lateral view of a lavender-eyed imago. The dorsal half of the eye was a graft from a pearl-eyed nymph. The pearl portion of the reconstructed eye has developed wild-type pigmentation which is spreading dorsally. Neither the lavender portion of the host’s eye nor the other eye of the host developed wild-type pigmentation,×20. (See also Fig. 3 G.)

Fig. 12.

Lateral view of a lavender-eyed imago. The dorsal half of the eye was a graft from a pearl-eyed nymph. The pearl portion of the reconstructed eye has developed wild-type pigmentation which is spreading dorsally. Neither the lavender portion of the host’s eye nor the other eye of the host developed wild-type pigmentation,×20. (See also Fig. 3 G.)

The final experiments were the placement of tissues from the prothorax far above the eye and into the center of the eye. As the eye grew, these tissues were eventually incorporated into the eye, and when a pearl eye grew into lavender or normal tissue and this tissue became part of the eye, normal pigments developed. (See Fig. 14.)

Fig. 13.

Lateral view of the right eye of a pearl imago. The dorsal half of the eye was a graft from a lavender-eyed nymph when both insects were fifth or sixth instar. The pearl half of the eye has developed wild-type pigmentation. The lavender portion of the eye has remained lavender. The other eye of the pearl host did not develop pigment,×20. (See also Fig. 3F.)

Fig. 13.

Lateral view of the right eye of a pearl imago. The dorsal half of the eye was a graft from a lavender-eyed nymph when both insects were fifth or sixth instar. The pearl half of the eye has developed wild-type pigmentation. The lavender portion of the eye has remained lavender. The other eye of the pearl host did not develop pigment,×20. (See also Fig. 3F.)

Fig. 14.

Late instar pearl nymph that had received a graft of epidermis from the prothorax of a wild-type nymph. The graft was placed far above the eye. Following the third post-operative molt pigment has begun to develop in the eye of the host and the line of the graft can be seen,×20. (See also Fig. 3 M.)

Fig. 14.

Late instar pearl nymph that had received a graft of epidermis from the prothorax of a wild-type nymph. The graft was placed far above the eye. Following the third post-operative molt pigment has begun to develop in the eye of the host and the line of the graft can be seen,×20. (See also Fig. 3 M.)

It can be readily observed that the eye of P. americana is constantly growing and maturing since there is a constant maturation of ommatidia all around the periphery of existing ommatidia. Further indication of this can be seen following the removal of tissue anywhere around the periphery. New growth and maturation zones appear by the time of the ecdysis succeeding the operation with no evidence of a scar.

Following complete removal of the eye, no eye regenerates from the remaining head epidermis, much of which would eventually have been replaced by eye tissue had the eye remained in place. This epidermis is not determined in its developmental fate at the time of hatching.

That there is progressive recruitment of cells from the epidermis by the mature ommatidia is indicated by the fact that epidermal tissue from a normal eyed or lavender-eyed mutant placed far above the eye ofa pearl-eyed mutant is eventually incorporated into the eye and synthesizes pigments that are typical pigments characteristic of the normal-eyed insect. There is no synthesis of these pigments until the cells become eye cells. Epidermis that is placed directly within the eye will heal the wound and is also induced to become eye tissue after which it synthesizes pigments.

The work of Ross et al. (1964) established the fact that pearl and lavender are inherited independently as autosomal recessives. The F1 hybrids between them or with either one crossed with the normal eyed roach had normally pigmented eyes. It is of interest here that putting eye tissues of these two mutants in conjunction results in the synthesis of normal eye pigments in the pearl mutant tissue. This plainly indicates the co-operative action of genes in different cells. That this is a result of direct contact is indicated by the fact that the other eye of a pearleyed mutant is never affected by these grafts.

Two substances, each acting only over short distances by direct tissue contact, are indicated here : one that induces the epidermal cells to develop as eye tissue, and a second one that results in the synthesis of the pigments of the eye.

Part of a dissertation presented to the faculty of the University of Virginia. I would like to thank my major professor, Dr Dietrich Bodenstein, and Drs James Dent and Howard Hamilton of the University and Dr Horton H. Hobbs of the U.S. National Museum.

I would also like to thank Drs Donald Cochran and Mary Ross, Virginia Polytechnic Institute, who provided the mutants used in this study.

Bodenstein
,
D.
(
1959
).
Contributions to the problem of eye pigmentation in insects: studied by means of intergeneric organ transplantation in Diptera
.
Smithsonian Mise. Coll
.
137
,
23
41
.
Bodenstein
,
D.
(
1962
).
Humoral Conditions and Cellular Interactions in the Development of the Insect Eye
.
Proc, of the Twenty-third Biology Symposium
,
Oregon State University Press
.
Lew
,
G. T.-W.
(
1933
).
Head characters of the Odonata with special reference to the development of the compound eye
.
Entomologia am
.
14
,
41
97
.
Lüdtke
,
H.
(
1940
).
Die Embryonale und Postembryonale Entwicklung des Auges bei Notonecta glauca (Hemiptera-Heteroptera)
.
Z. Morph. Ökol. Tiere
37
,
1
37
.
Mellanby
,
H.
(
1937
).
The later embryology of Rhodniusprolixus
.
Q. JI microsc. Sci
.
79
,
1
42
.
Penzlin
,
H.
(
1963
).
Uber die Regeneration bei Schaben (Blattaria). I. Das Regenerations-vermogen und die Genese des Regenerat
.
Wilhelm Roux Arch. EntwMech
.
154
,
434
465
.
Ross
,
M. H.
,
Cochran
,
D. G.
&
Smyth
,
T.
, Jr
. (
1964
).
Eye-color mutation in the American cockroach, Periplaneta americana
.
Ann. ent. Soc. Am
.
57
,
790
792
.
Ward
,
C. L.
&
Hammen
,
C. S.
(
1957
).
New mutations affecting tryptophan-derived eye pigments in three species of insects
.
Evolution
11
,
60
64
.
Witlaczil
,
E.
(
1884
).
Entwicklungschichte der Aphiden
.
Z. wiss. Zool
.
40
,
559
596
.