The composite, haustellate mandibles of netwinged beetle and firefly larvae (Coleoptera: Lycidae, Lampyridae)


Material Information

The composite, haustellate mandibles of netwinged beetle and firefly larvae (Coleoptera: Lycidae, Lampyridae)
Physical Description:
xvi, 229 leaves : ill., photos. ; 29 cm.
Cicero, Joseph M., 1951-
Publication Date:


Subjects / Keywords:
Lycidae -- Anatomy   ( lcsh )
Lycidae -- Development   ( lcsh )
Fireflies -- Anatomy   ( lcsh )
Fireflies -- Development   ( lcsh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1991.
Includes bibliographical references (leaves 220-227).
Statement of Responsibility:
by Joseph M. Cicero.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001747377
notis - AJG0200
oclc - 26371908
System ID:

Full Text








As a strong advocate of the need for studies in basic,

practical entomology, committee chairman Howard Frank allowed

me full latitude for developing my own school of thought in

comparative developmental insect morphology. This latitude was

generously granted at the expense of his own research funds,

including an assistantship appointment, and his research

objectives in biological control. I anticipate that, in the

years to come, the philosophy that allowed for this will be

paid off many times over, as I plan to radically change the

way entomologists view insect ontogeny, phylogeny, and

ontophylogenetics. Howard was tremendously supportive during

my philosophical difficulty with qualifying exams. Further,

during national Society meetings, I detected him working very

persuasively behind the scenes to help me build a strong

reputation among scientific circles. I wish I could be certain

that he knows of my appreciation for these things.

I found Jim Lloyd to be so complex, captivating and

astonishingly right about virtually everything he expounds on

that I have no idea how to acknowledge him as a scientist,

friend and long-time firefly correspondent. So I turned to the

acknowledgement section of the dissertation submitted by his

former student John Sivinski (1982), who undoubtedly spent


considerable time deliberating the same thing. Indeed, John

did find right words, and my paraphrasing of them will be the

second dilemma that he has helped me out of, "...Jim is likely

to be the most interesting person I will ever meet. (page


The first dilemma came when, recently, Fred Bennett had

to respectfully ask that he be excused from my committee in

the twilight of his academic career. John gladly accepted this

vacancy and, in so short a time, sized up the ramifications of

my research and helped me tame many of my wild ideas about

evolution. Fred always showed good intentions as he watched me

gradually drift many miles away from the subject he mastered,

biological control.

Jonathan Reiskind served as the extradepartmental

committee member. As with Fred, my exploration of cell,

developmental and molecular biology took me many miles away

from his specialty subjects. I do appreciate that he stayed

with me through to completion. The perspective he wielded

during his critical review of my dissertation made the text

and figures considerably more user-friendly.

Henry Aldrich is a man of truly massive accomplishment.

I strongly urge students reading this to take his classes in

electron microscopy and cytochemistry, if for no other reason

than to profit from the mere meeting of a man with so

staggering an erudition. He and Chris West are the finest


teachers I know of, and so figured in decisively during my

adventure away from macrobiology.

Chris and I have become dear friends because we think a

lot alike in so many aspects. Frighteningly a year younger

than me, so charismatic and dynamic, he has made himself a top

flight cell, developmental and molecular biologist. A wizard

at inductive thinking, it is my great loss that circumstances

didn't allow me to fully apply myself to his teachings.

May I also extend sincere acknowledgement to all other

faculty and support personnel, especially Ross Brown, Brad

Coriell, Harvey Cromroy, Greg Erdos, Steve Lasley, Jim Nation,

and Tom Walker, of whom I'll carry memories of association.

Much of my field research was done in areas that required

permission, a base of operation, and access by day and night

without prior scheduling. I would like to thank stewards,

managers and hosts for obliging these needs. They include

David Marqua of the Davis Mountain Resort, Steve Prchal of the

Sonoran Arthropod Studies Inc. (SASI), Tucson, Arizona, Randy

Brown of the San Felasco Hummock, Gainesville, and Mark Deyrup

of the Archbold Research Station, Lake Placid, Florida.




ABSTRACT . . ... xv



METHODS . . ... 33
Overview . . .. 33
Collection . . .. 34
Lycidae . . 34
Lampyridae . . 36
Tenebrionidae . .. ... 36
Processing . . 38
Fixation . . 38
Embryos . . 39
Larvae . ... 42
Dehydration . .. 43
Mounting for SEM . .. 44
Embedment for TEM . .. 44
Mounting on a TEM chuck stub .. 45
Sectioning . . 47
Viewing . . 48
Records ... . ... 49
Film . . ... 50
Artifacts and complications . 50

RESULTS .. . . 76
The Larval Lampyrid Syntrophium .. .77
The mature syntrophium . .. .77
The younger syntrophium . .. .79
The Larval Lycid Syntrophium . .. .80
The mature syntrophium . 81
The younger syntrophium . .. .83
The Pharate Larval Lycid Syntrophium .. .84

The Pharate Tenebrio Larval Mandible .. 85

DISCUSSION ... . 157
The Structural Level .... ..... 158
The base ... . 160
The shank . ... 162
The apex .... 163
Development .... . 163
Analysis of Hypothesis ... 164
Anticipation of the Cellular Level . 168
Concluding Remarks .... .. 169

Parasitism and Saprophagy . 193
The Experiment ... ............. 196
The Birthing as a Normal Event .. 197
Larval Constitution of Lampyridae and
Lycidae . . 198
The Birthing as a Freak Event ..... 200
Review of Other Practitioners .. 201
Developmental .... 201
Genetic .. .... 204
Analysis of Review .......... 206
Developmental ...... .. 206
Genetic . . 207
Relevance of the Review to the Birthing 208
Literal atavism .*. *209
Adult Constitution of Lampyridae and Lycidae 212
Concluding Remarks .... .. . 214

REFERENCE LIST ... .. . 220



Figure 1. Bearer and one cluster of siblings. Note
cannibalism (arrows). . 4
Figure 2. The second of two clusters of siblings. 4
Figure 3. Left, one of the sibling larvae. Right, larva
of Photuris. These can be compared with a picture
of the bearer's head capsule in the appendix
(Figure 114, page 216) .. 6
Figure 4. Current conception of phylogeny of families
concerned (Crowson 1972). . .... 6
Figure 5. Apex of sibling abdomen, dorsal view. One of
the 4 other siblings bound to it at birth was cut
away for dissection. The other 3 fell away from the
cord (arrow) that tied them. .......... 8
Figure 6. Ventral view of the apex of a sibling abdomen
in strong transmitted light to show light organs
(arrow). Note the cord and filamentous pygopod. 8
Figure 7. Mouthparts of L. sanquineus larva. A) Basal
zone. B) Mesal zone. C) distal zone, clearly
identifiable as the gala. D and E) Maxillary
palpifer. The palpifer is scalloped between these
arrows to receive the A-B-C complex. F) Labial
palp. . . .. 21
,Figure 8. Exuvial halves. A) Antenna. B) Ecdysial line.
C) Membrane. D) Stylet. E) Labrum, corresponding to
(A) in Figure 7. F) Channel from which the stylet
can be withdrawn. ........ .. 23
Figure 9. SEM of Pyractomena mandible, ventral view. A)
Acetabulum. B) Basolateral opening to the interior.
Mandibles are connected to the head not only by the
acetabulum, but also by tissue that passes into
this opening. . . 25
Figure 10. Classical drawings of lampyrid larval
mandibles have identified these features. A) Canal.
B) Pore. C) Retinaculum. D) Acetabulum. The outer
aspect (E), ensheathing the canal, will be referred
to as the "cone" in lieu of its actual identity. 25
Figure 11. Models that might explain channelled
mandibles. A) Canal as an apophysis. Inset shows
the opposed epicuticles expected if lampyrids
conformed to De Marzo and Nilsson's (1986)
findings. B) Canal as an apodeme. ... 27


Figure 12. Mandible of Pyractomena larva longitudinally
shaved. A) Basolateral opening, corresponding to
Figure 9B. . .... 29
Figure 13. Mandible of Pyractomena transversely shaved.
Ventral view. .... ...... 29
Figure 14. Whole mount of a 1st instar P. collustrans,
freed from the embedment plastic after being shaved
to the basal half of its mandible. . 31
Figure 15. Close-up of mandible in Figure 14. 31
Figure 16. Cross-section of the head of an anatreptic P.
collustrans embryo showing the difficulty in
identifying which lobe corresponds to which anlage
without 3-D reconstruction. ...... 52
Figure 17. Materials used. A) Petri dish with window. B)
Dechorionation of egg with two-sided tape. C) Vial
rack. D) Eppendorf* vial with tip cut off. E) Beem*
capsule modified for minute specimens. ... 52
Figure 18. P. collustrans embryo embedded and
photographed through a compound microscope. See
Figure 26. The ideal orientation is shown, with the
eyes eclipsed to the viewer. ......... 54
Figure 19. Anatreptic P. collustrans embryo, embedded
and photographed through a compound microscope. See
Figure 26. A) Mandible anlage. B) Maxillary palp
anlage. . . ... 54
Figure 20. Cracking the eggshell. .... 56
Figure 21. Whole embryo affixed to an SEM stub in its
native, curled state. The curl is then broken so
that the mouthparts face upward. ... 56
Figure 22. Blebbing of cuticle in an improperly
processed specimen. .... 58
Figure 23. Pharate larval Lycus lecontei, dorsal view,
fixed, then removed from its exuviae. Central gap
is where stomodeum passes through mouth when the
head is pulled out of the exuviae. A)
Labromandibular rudiment. B) Gala. C) Maxillary
palp. . .... ... .. .58
Figure 24. Trial specimen, a mole cricket embryo,
dehydrated with HMDS. . 60
Figure 25. Well organized SEM stub with all embryos
facing the same approximate way along scored lines
so that magnification and goniometer adjustment is
not necessary for perusal from one specimen to the
next. . . ... ...... 62
Figure 26. Viewing embedded embryos through a compound
microscope. Whittling of the stub was necessary.
Figure 18 and Figure 19 were photographed by this
means. ................ 64
Figure 27. Sectioning sequence, ultrathin alternating
with semithin. ................ 66


Figure 28. SEM stub mounted in a microtome chuck so that
tip of specimen's mandibles can be shaved. Large
arrow points to fragments of the mandible tip. 68
Figure 29. Finder sheet for SEM stub specimens. 70
Figure 30. Blob of glue held between tweezer tines to
clean the stub mounted embryo. . ... 72
Figure 31. P. collustrans embryo obscured by adherent
yolk and other material. . ... 72
Figure 32. P. collustrans embryo cleaned by the method
in Figure 30. Such results are rarely so good as
most debris is firmly dried to the embryo. A)
Invagination. . . ... 74
Figure 33. Lycid mutants. A) Failure to undergo
katatrepsis. B) Failure to breach the serosa during
katatrepsis. a) Chorion. b) Serosa. ... 87
Figure 34. Photuris eggs dried to about an 80% loss of
spherical volume. Larvae within suffer extreme
contortion from strong, localized, pressures. 87
Figure 35. First in longitudinal series, from dorsum to
venter. Canal actually opens dorsally to the buccal
cavity. The section angle makes it appear to open
laterally. See Figure 42. A) Dorsal deflection. B)
Antenna. . . ... .. 89
,Figure 36. Second in long series. A) Amorphous areas. B)
Apparent seam is probably a setal socket as no such
seam is indicated in transverse sectioning. 89
Figure 37. Third in long series. A) Section grazes
retinaculum . . 91
Figure 38. Final in long series. A) Section grazes tip.
B) Stomodeal exuviae. C) Base is expanded
interolaterally. . .... .. 91
Figure 39. Close up of amorphous disc in Figure 36A. 93
Figure 40. First in transverse series of P. collustrans,
from base to apex. A) Arrows point to the two lobes
that are under observation. B) Antenna. C) Maxillar
tissue. . . 93
Figure 41. Second in transverse series. A) Continuity
from transfrontal region to syntrophium. 95
Figure 42. Third in transverse series. Dorsal closure is
approached. A) Cells back away from the apex of the
2 lobes to form caps of cuticle. ... 95
Figure 43. Fourth in transverse series. Dorsal closure
of the canal is reached. A) Canal lumen is closed
to the mouth at this point. . ... 97
Figure 44. Fifth in transverse series. A) Horseshoe
shape of cuticle. B) Area continuous with head
tissue is outside the embrace of the horseshoe. 97
Figure 45. Sixth in transverse section. A) Points to a
lateral area that deposits a welt of cuticle of its
own between B) and C), the ends of the horseshoe. 99

Figure 46. Final in transverse series. Section cuts
tangent to the pore. Below are labial palps, gala
and maxillary palpifer. . 99
Figure 47. SEM of Photuris 1st instar showing pore,
upwardly angled tips and retinaculum, the latter
partially hidden by setae. . .. 101
Figure 48. Pore of Pvractomena larval mandible. A)
Although partly obscured by foreign material, a
seam can be seen to extend apically for a short
distance. . .... ... 101
Figure 49. Long section through the tip of a P.
collustrans mandible. Cell diameter is extremely
attenuated. . . 103
Figure 50. Cross-section through extremely tough tissue
forming the apex of a P. collustrans mandible. A)
Origin of a seta at the end of a dendrite. .... 103
Figure 51. Innervation of the mandible tip in P.
collustrans. A) Dendrite. B) Canal lumen. 105
Figure 52. Tip of P. collustrans mandibles. A) A
branched seta at the pore. B) Lack of .a seam. C) A
campaniform sensilla is on the opposite side. 105
Figure 53. First in series of an anatreptic P.
collustrans head capsule. A) Elongate lobe. B)
Smaller lobes. C) More lobes occur above (A); these
are shown in Figure 54. There is no indication of
which lobe corresponds to which mouthpart. .. .. 107
Figure 54. Same section as Figure 53. A) Elongate lobe,
same as Figure 53A. B) Shorter lobes pointed to in
Figure 53C. There is no definitive landmark for
orientation ..... . 107
Figure 55. Second in series. A) Section grazes the
definitive gala. B) This elongate lobe is the
maxillary palp. C) This lobe is therefore the
mandible. D) The other lobes are cervical. ... 109
Figure 56. Final in series. A) The gala protrudes
distinctly. . . 109
Figure 57. Young mandible, prior to commencement of
curvature, showing the apical invagination. Also
shown by SEM in Figure 32. The 3-armed pattern at
the palp apex is a peculiar ripple in the
section. . . 111
Figure 58. Young lampyrid syntrophium at commencement of
elongation and curvature. Note the extremely high
nuclear: cytoplasmic volume ratio. ... 111
Figure 59. Young P. collustrans fixed and embedded at a
point when the mandibles are first undergoing
curvature. A) Stomodeal exuviae. .... 113
Figure 60. Long section through the dorsal deflection.
A) Incipience of the canal. . ... 113

Figure 61. Facsimile 3-D reconstruction of the P.
collustrans mandible. Reconstruction ascertains
which of two grazed areas, circles, is of interest.
Square shows where Figure 60 is located. A)
Incipience of canal. B) Dorsal deflection. ... .115
Figure 62. Montage of specimen in Figure 60. A) Canal
orifice. 117
Figure 63. First in longitudinal series of a P.
collustrans katatreptic syntrophium after curvature
has commenced. A) Section grazes midlength of the
mandible, showing that the canal has progressed at
least to this point. . .. 119
Figure 64. Second in long series. A) Origin of the
canal. B) Progress of formation is at least to near
apex. . . ... 119
Figure 65. High magnification of a region in Figure 64,
showing the microvilli that appear during canal
formation . . 121
Figure 66. Third in long series. Section passes beneath
the canal and the longest length. A) Conduits are
present that are probably haemocoelic. See
Figure 67 . . .. 121
Figure 67. High magnification of the apex in Figure 66,
showing presumed haemocoelic conduits. .. .. 123
Figure 68. Head of an unidentified lycid larva.
Syntrophia are hypognathous and their bases (A) are
contiguous . 125
Figure 69. Interior view of head exuvial halves opened
and laid flat as in Figure 8. A) Acetabulum. B)
Ridge. See Figure 90. C) Crop. ......... 125
Figure 70. First in an extensive cross-sectional series
of Calopteron first instar. A) Labrum is continuous
across front. B) Its venter infiltrates the arc-
shaped mandible. . . .127
Figure 71. Second in series. A) Section grazes the
anterior extent of the labrum. . 127
Figure 72. Third in series. A) Labral halves are
separate and, B) fully contained within the arc-
shaped mandible. . . 129
Figure 73. Closeup of the base in Figure 72. A) Nerve
bundle is separate from the labral wall at this
cross-section. Black spheres are post-stain
residue. B) Labromandibular junction. ... 129
Figure 74. Sectioning continues. A) Gala is reached. .131
Figure 75. Close-up of gala-labromandibular interface. 131
Figure 76. Extreme close-up of Figure 75. Gala is
independent . . 133
Figure 77. Closeup of nerve bundle in Figure 75. A)
Fibrils. ............. 133
Figure 78. Sectioning continues. A) Gala ensheathes. B)
Maxillary palpifer appears. . ... 135

Figure 79. Closeup of left syntrophium in Figure 78. A)
Center of the mandible. The canal tracks along this
longitudinal. . .. 135
Figure 80. Final in series. Closeup of right syntrophium
in Figure 78. . . 137
Figure 81. First in series of Lycus lecontei syntrophium
apex. Mandible (lower) and gala (upper) exceed the
labrum in length. The gala undergoes dramatic
histological change from a cell-filled to,
Figure 82, an alveolate interior. . 137
Figure 82. Second in series. The alveolate interior of
the galar apex is formed by dehiscent spheres and
probably functions as a sieve during feeding. 139
Figure 83. Final and distal-most cross-section of
syntrophium. Mandible exceeds the length of the
gala. . . 139
Figure 84. Exuviae of L. sanquineus stretched over a
pinhead so that the mouthparts protrude. Figure 85
through Figure 87 are micrographs of a specimen in
this orientation. . . 141
Figure 85. Apical view, that is, looking down the shank,
of the syntrophium. The labral stylet is removed. 141
Figure 86. Closeup of oral orifice. A) The mandibles
arise as extensions of the pharyngeal wall and,
(B), abruptly bend to hypognathy. ... 143
Figure 87. Close up of the base, diagonal to the mouth
orifice. A) The mandible narrow at attachment and,
(B), its inner margin rolls over to form a
crescent. . . 143
Figure 88. Anatreptic specimen of Calopteron. A)
Mandible anlage. .. . .145
Figure 89. Katatreptic embryo of Calopteron. A) Labrum
is a transverse fold on top of the mandible. B)
Central flap, perhaps referrable to as the
hypopharynx. A mote of foreign material is in the
mouth . . 145
Figure 90. Katatreptic embryo of Calopteron. A) Division
between labro-mandibular and maxillar regions. This
is the ridge shown in Figure 69. .. 147
Figure 91. Side view of Calopteron mouthparts showing
three layers. A) Labrum. B) Upper mandibular lip.
C) Lower mandibular lip. C) Transverse object at
apex is the tip broken back by the methods of
Figure 28. See Figure 92. .. .. 147
Figure 92. Close up of Figure 91. A) Labrum. B)
Mandible. Both are appressed to each other as they
coextend . ..... 149
Figure 93. Inner ventral view of the Calopteron
syntrophium in Figure 91. A) Gala is upwardly
concave to receive the labro-mandibular affair. 149
Figure 94. Esophageal rim of a pharate larval Lycus
lecontei. . .... .151


Figure 95. Extreme close-up of Figure 94. Mouthpart
lobes are positioned along the rim of the mouth. A)
Antenna. B) Labromandibular rudiment. C) Maxillar
rudiment. .............. .151
Figure 96. Apolysed larva of Tenebrio with exuviae of
mandible lifted off like a dental crown to reveal
developing mandible within. This can be seen more
clearly in the closeup in Figure 97. .. 153
Figure 97. Close up of Figure 96. A) Pharate mandible
with its exuviae removed. . ... 153
Figure 98. Transverse section of pharate larval mandible
of Tenebrio molitor showing crenate pattern of
cuticle .. ............ .. 155
Figure 99. Close-up of creation in Figure 98. 155
Figure 100. Attempt to homologize (A) the lycid and (B)
the lampyrid syntrophia. Trace of (B) over (A)
facilitates comparison, but arrows point only to
aspects of (A). a) Inner canal wall. b) canal
lumen. c) Outer canal wall. d) Double layers of
cuticle. . . ... 171
Figure 101. Minimum requirements of lampyrid syntrophial
cell identity. A) These must be mandibular. B)
These must be labral. C) Dotted line indicates that
mandibular cells retreated after laying down
cuticle in this area. . ....173
Figure 102. All cells given identities required to
correspond directly to the lycid affair. A) These
would be mandibular. B) These would be labral. 173
Figure 103. Close up of Pyractomena mandible shaved as
described in Methods. A) Area corresponding to the
dotted line in Figure 101 and Figure 102. .. 175
Figure 104. Summary of areas used in key interpretations
for both families. A) Narrow mandibular base. B)
Arrows show fluting. C) Lumen. D) Center of
mandible. E) Labrum is continuous with transfrontal
tissue. F) Labrum undergoes fluting. G) Welt of
cuticle. . . ... 177
Figure 105. Typical textbook illustration of a mandible.
No indications are made regarding where the cells
filling the interior (cut-away) originate from. 179
Figure 106. Tracking of the lampyrid canal. ... 181
Figure 107. A) Lycidae. Labrum (a) falls short of
mandible (b). (c) Alveolate gala. (d) Tip of
mandible, cut for viewing path of imbibition
through pore. B) Lampyridae. Pore is formed because
labral field (a) is shorter than mandibular field
(b) . . . 183
Figure 108. An interpretation of the invagination in
Figure 32A. A) Mandible anlage flanked with labral
cells. B) Labral cells infiltrate. C) Invagination
is cast off. D) Anlage is smooth again. E) No
apodeme was found. . ... 185


Figure 109. Cross-section of anlage in Figure 108 held
static while cell fields pass through during
development. A) Flank of labral cells. B)
Invagination. C) Infiltration. D) Fluting. E) and
F) Separation. . .. 187
Figure 110. Mesal section of a 1st instar P. collustrans
syntrophium. A, B and C show three directions from
which cuticle might be deposited. A) Labrum, B)
upper mandibular lip, C) lower mandibular lip. 189
Figure 111. Close up of the base of the lycid
syntrophium in Figure 73B. A) Mandible. B) Labrum.
C) Labromandibular junction. The junction need only
be displaced outward, in the direction of the
arrow, for the two anlagen to be made one. .189
Figure 112. Generalized beetle head capsule overlaid
with a compartmentation pattern that might be
expected from what we know of Drosophila cell
behavior. . . 191
Figure 113. One of numerous vesicles composing the
pygopod of lampyrid larvae. . ... 216
Figure 114. Head of L. sanquineus larva, the bearer in
figure 1. . . .216
Figure 115. Abdominal apex of L. sanquineus larva. The
pygopod is probably the protracted rectum, shown
here in its retracted position. ... 218


Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



Joseph M. Cicero

December 1991

Chairperson: Dr. J. Howard Frank
'Major Department: Entomology and Nematology

During netwinged beetle embryogenesis, the labrum and the

mandible develop as separate anlagen. The anterolateral

corners of the labrum grow out as stylets. They are positioned

on top of the mandibles during this elongation, and their

inner faces take on a crescentic fluting. The mandibles grow

out as stylets and undergo fluting too, but, in contrast, the

crescent is open to the lateral aspect, rather than to the

inside. These two then interlock, with the labral stylet

inside the flute of the mandibular stylet. This arrangement

creates a food canal: the inner face of the labral stylet

'becomes the canal's outer wall, while the outer face of the

mandibular stylet becomes the inner wall.

During firefly embryogenesis, only one anlage develops in

the position expected for labrum and mandible. This lobe

elongates, curves, attenuates to a sharp tip, and its axial


cells pull away from each other to form a canal lumen. The

entire sucking apparatus forms without cuticle-bound


By comparing developmental aspects of the two families it

is suggested that the labral and mandibular anlagen of

fireflies have somehow united, so that cells of each migrate

out to positions required of them to form the mature structure

first, then lay down cuticle as a finishing touch, bypassing

developmental intermediate stages.

This study was prompted by the discovery of a netwinged

beetle larva that had ostensibly given birth to firefly

larvae. It seemed that a logical first step would be to study

an aspect of basic biology that might lead to a better

understanding of the relationships between them. Mouthparts

were chosen because they serve as important diagnostic

characters that separate the two families taxonomically.

Possibilities on how such a birthing could be so fall

into two groups. If the sibling firefly larvae are not a

product of the bearer's genome, then parasitism and saprophagy

are likely causes. If, however, the siblings were born by

paedogenesis, then horizontal saltation can be invoked.

Horizontal saltation is discussed as a thought experiment that

explores such possibilities as hypermetamorphosis, chromosome

rearrangement, and change in the regulatory system that

controls development.



This investigation was prompted by an extraordinary

discovery made in May of 1981 by Robert J. Weich in Tucson,

Arizona. At Campbell and Table Mountain Roads, he collected a

30mm by 9mm larva of Lvcus (Neolycus) sanguineus (Gorham)

(Lycidae) from which, after being dropped in preservative,

issued 9 larvae through a rupture in abdominal intersegment 3-

4 (Figure 1, Figure 2). These larvae are of the related family

Lampyridae (Figure 3, Figure 4), but nothing in the known

biology of either family suggests how such a relationship

could come about. The host or mother, it is not absolutely

certain which, will be referred to as the "bearer."

The bearer's internal tissue had been consumed and some

sibling cannibalism occurred, accounting for the considerable

development, to 14mm by 3mm, of the survivors. Gonads of

undetermined sex are present. Tanning and pigmentation had

occurred. A cord of some sort, issuing from within the anal

end of each sibling (Figure 5, Figure 6), had knotted them

together by their anal prolegs into 2 groups of 5. Tissue

masses corresponding in shape, position, cellular structure

and tracheal attachment to light organs of larval Photuris are

present, as well as fenestrations which, in Photuris, allow


light to emanate (Peterson 1970). Neither ability nor

inability to luminesce could be observed.

I would not expect anyone to believe such a report until

it is somehow verified. Disappointingly, however, there are

several reasons to think that determination of the mechanism

involved will come with great difficulty, if ever at all. The

birthing is so extraordinary as to render for consideration

only the most imaginative of conceivable mechanisms.

Parasitism, or scavenging, by a species with or without

adults, paedogenesis resulting in horizontal saltation or

atavism, horizontal gene flow (Dawkins 1982:159), perhaps

others, are so far-fetched that even collection and

observation of another such birthing would constitute no

evidence in favor of one or the other. A further complication

is that larvae of L. sanguineus are very difficult to locate

in the field and rear in the laboratory for the controlled

experiments required to determine the source of the genes at

play, whether intrinsic or extrinsic.

On the practical side, this discovery has provoked two

lines of meaningful and productive research. The first is

appended to this dissertation. It is a thought experiment that

accepts as its initial conditions the possibility that the

siblings are a product of the bearer's genome, and invokes

horizontal saltation as the explanation. The second line

developed from the idea that the main focus of attention might

best be placed on developmental morphology. So little is known


about this aspect of the families concerned that a careful

,examination of key character ontogeny would be a logical first

step toward an understanding of how this birthing could be so.

The mouthparts, and in particular, the mandibles are of

extraordinary design and are used as main diagnostic features.

Because of these two aspects, I have decided to dedicate this

dissertation to their elucidation.

Figure 1. Bearer and one cluster of siblings. Note cannibalism

Figure 2. The second of two clusters of siblings.





.ai 4 X

Figure 3. Left, one of the sibling larvae. Right, larva of
Photuris. These can be compared with a picture of the bearer's
head capsule in the appendix (Figure 114, page 215)

Figure 4. Current conception of phylogeny of families
concerned (Crowson 1972).

r 4

:. n

Figure 5. Apex of sibling abdomen, dorsal view. One of the 4
other siblings bound to it at birth was cut away for
dissection. The other 3 fell away from the cord (arrow) that
tied them.

Figure 6. Ventral view of the apex of a sibling abdomen in
strong transmitted light to show light organs (arrow). Note
the cord and filamentous pygopod.



w ::,,


Withycombe (1924) showed a remarkable curiosity in

recognizing that lycid larvae possess a paired, tubular,

piercing and sucking organ composed of three components.

...the mandibles are rather short and sharply pointed,
grooved dorsally. Upon each, dorsally, lies a pointed
'prostheca' or 'lacinia mobilis', grooved ventrally,
so that the two appendages form conjointly a tubular
piercing and sucking organ very similar to that formed
in Neuroptera by the mandible and maxilla. The maxilla
lobe (? lacinia) is short and tapering, but blunt-
ended. It is also grooved, and is in life often
applied to the mandible. There is a short, four-
jointed maxillary palpus, and a shorter, three-jointed
labial palpus...(page 322)

Scanning electron microscopy (SEM) shows that three zones

are indeed involved (Figure 7). The basal and distal zones are

rugose, setose, and separated by a smooth, glabrous mesal zone

that appears to have a longitudinal seam. Simple probing with

a dissecting pin shows that the distal zone is actually the

gala that has undergone internal collapse so as to allow it to

ensheathe the mesal zone. Further, by pulling the dorsal and

ventral halves of the head exuviae apart (Figure 8) and laying

them flat, it is seen that the basal zone, the 'prostheca', is

separate from the mesal zone and might actually be the labrum.

Labrum and clypeus are considered absent in larvae of this

family (Crowson 1955). The modern and foremost interpretation,


that the mandible is not composite, but instead cleft to base

into two lobes (Lawrence 1991), is incorrect.

Scanning electron microscopy of the larval lampyrid

mandible shows nothing in common with this arrangement

(Figure 9). In fact, the only shared feature is symmetry--

there is no left- or right-handedness as in certain other

families (Hillerton 1986). No aspect of the head is associated

with it except two tissue-connecting sites (Figure 9B,

Figure 12).

Returning to the head exuviae of Lycidae, close

inspection shows that the mesal zone is actually hollow and a

thin stylet can be lifted out from inside (Figure 8D). Still

further, the basal segment of the maxillary palp (the

maxillary palpifer) abuts closely and takes on a scalloping

that ensheathes the 3-zone complex in the area of contact

(Figure 7D,E).

The larval lampyrid mandible was known to be channelled

early in this century (Haddon 1915, Vogel 1915) (Figure 10),

and one cross-section of the cuticular ultrastructure has been

published (De Marzo and Nilsson 1986). These authors were

studying the mandibular channel of larval dytiscids and looked

to the channel of Lampyris noctiluca (L.) for any possible

similarities. They found that it, and the channels of dytiscid

'genera considered highly evolved, have a seam. They refer to

the seam as a "fused area", and believe that it represents the


interface between upper and lower lips of a longitudinal

groove (Figure 11).

Their belief is based on the pattern by which cuticle is

laid down, and on an assay referable to as "intermediate

character state comparison." The pattern is of two cuticles

with their epicuticles closely opposed to each other

(Figure 11, inset) suggesting an invagination. The character

states are of other dytiscid larvae, which show a gradation

from an open groove to a closed one. An initial inspection of

the channel in lampyrid species available to me shows that

they are at least superficially consistent with De Marzo and

Nilsson's cross-section of L. noctiluca (Figure 12 through

Figure 15). The inspection involves embedding the mandible in

plastic and sectioning though half its length or diameter. The

remaining half is then freed from the plastic with a powerful

oxidizing agent (see Methods) and coated for SEM. It is clear

that cuticle does indeed line the lumen of the canal as well

as the outside as one would expect, but the fused area is not


De Marzo and Nilsson's presentation of concordance

between this gradation and gradations of other dytiscid

characters makes for a tight case that satisfies contemporary

phylogenetic systematists. However, the composite nature of

lycid mandibles, together with the relatedness of the two

families, as well as the lack of a fused area, are grounds for

investigating the lampyrid mandible further.


There are at least 2 models that satisfy the De Marzo

cross-section (Figure 11). In one, the canal is a groove and

in the other, it is an apodeme which starts either apically,

basally or perhaps even internally. Both are common

endoskeletal features, except that apodemes do not penetrate

the opposite end to produce a continuous path. A more

appropriate reference would be hollow tentorium. Either model

as well as others are possible, but verifying which may not

determine whether we are dealing with two structures, one (the

canal) internal to the other. Might the mandible be inside the

labrum, or the labrum inside the mandible? A terminological

problem arises immediately which can cause confusion in this

dissertation. Therefore, during the descriptive phase, the

aspect that the canal is housed in will be referred to as the

"cone" (Figure 10).

The gala of lampyrids is involved in a maxillo-labial

affair that functions to scoop up food (Cicero 1982:275,

fig.7E; unpublished observations) and is independent of

mouthparts dorsal to the buccal cavity. However, there still

could be a relationship between the mouthpart complexes in

these two families. An investigation of their actual

construction is in order, and will be the specific mission of

this dissertation.

To determine actual construction, it will be necessary to

track the mouthparts during embryogenesis and possibly also

during pharate larval development. Regarding the latter, the


.question of how mouthparts of larval instar X+1 are

reconstructed from larval instar X has never been considered.

The epidermal cells in question may retain the original shape

by using the exuviae as a template or they may retreat back to

the head cavity and then reconstruct the mandible anew. Such

rebuilding may be a recapitulation of embryonic processes, or

a different program altogether. The cells might also die back

to stem cells which, in turn, would regenerate daughter cells

to reassemble the mandible. In any case, the components

involved might be separate and distinct at that time, within

the hidden confines of the exuviae.

A small literature exists for mandibles of larvae in

other beetle families. It is primarily composed of economic

species, such as stored product pests, and it is focused on

ways to identify species whose fragments reach the market as

contamination (Kuenberg 1977, 1981). Because of the variety of

shapes and dentations, as well as their durability, mandibles

are usually isolated in an intact state and can be matched up

to their owner in an atlas.

Necrobia rufipes (Cleridae) is perhaps the most closely

related stored pest to cantharoids, but our knowledge of its

mandible offers no obvious correlations. It possesses a ridge

on the ventral surface of the mandible that extends from the

molar area to the tip of the incisor, giving the appearance

that the mandible has upper and lower halves. The labrum of

clerids is described as free and distinct (Lawrence 1991).


Mandibular morphology, asymmetry and differential wear

are popular ecological assays for determining feeding habits

(Smith and Sears 1982; Wallin 1988), and can be used in

conjunction with midgut content surveys. These papers use the

conventional repertoire of terminology (Peterson 1982),

recognizing the uni- bi- and tripartite incisor, mola,

prostheca, penicillus, retinaculum, condyle, acetabulum,

lateral and cutting edges. They certainly had no reason to

venture the slightest suspicion of composite construction.

My challenge of the view that the "mandible" is a

singular unit in serial homology of mouthparts is a totally

novel pursuit for Coleoptera. Major primary literature, such

as Snodgrass (1935, 1951), Matsuda (1965) and Manton (1977) as

well as secondary literature (i.e. Chapman 1971) are uniform

in their presentation of the mandible as a discrete,

homological unit throughout the phylogeny of Mandibulata.

The primary literature offers very little in the way of

mandible anatomy also. Zacharuk (1979) has ventured into the

histological diversity of larval mandibles, i.e., those of

tanned and general larvae as well as their exuviae, but from

standpoint of seeing how cuticular hardness and mechanical

advantage develops with each molt. Similarly, Hillerton's team

(1984) found that hardness was gained by inclusion of zinc and

manganese to the extent of several percent dry weight of the

mandible in some beetles.


No other relevant information was found in the literature

on Coleoptera. It seems that if any preparatory help is to

come from prior works, it should be found among classically

haustellate orders. There might at least be features unique to

composite mouthpart design.

The term syntrophium was coined by Jobling (1976) to

replace the imprecise term "fascicle" used by earlier authors

,for the piercing organ of Diptera. He does not state the

rationale by which he decides which stylet belongs to which

primitive structure. Aphid stylets were given designations in

1928, but to investigate any earlier than 1976 would be futile

as it predates the coming of age of cytology and developmental

biology. Such designations are carried forward to the present

and are assumed by their basal orientation to be correct. It

can be argued that these designations need to be rechecked

since the stylets are minute, closely appressed, and their

bases are protracted (i.e., Wiesenborn and Morse 1988).

In Drosophila, the mouthparts have been traced back to

the original compartment they arise from. Compartments are

perhaps the ultimate designation. They refer to the minimum

number of cells, called polyclones, which are generated from

the same parental cell and which minimally embrace the

structure in question (Garcia-Bellido et. al 1973).

A checklist of things likely to be found in haustellate

mouthparts can bring this chapter to an end.


Salivary canal. In aphids (Forbes 1969) it is unpaired

and formed by an internal ridge of the maxillary stylet that

rolls over onto itself. In psyllids (Forbes 1972), it is

paired and formed from hollows in the interlock that holds the

maxillary stylets together.

Food canal. This duct is formed by the space between the

interlocked maxillary stylets and open to the outside. Modern

investigations, with their conviction to demonstrating actual

function, have gone to the extent of fixing the beak while it

is embedded in plant tissue, and sectioning for the processes

that are involved in imbibition (Kimmins 1986; Spiller 1990).

Unity. Stylets must be held together somehow, so that

they may be deployed as a unit. Maxillary stylets interlock

mechanically, like jigsaw puzzle pieces, which keeps them

appressed against each other, combines their tensile strength,

and allows them to slide upon each other as the need for

bending the syntrophium arises. Mandibular stylets concavely

embrace them but do not interlock. There is no information on

how they are prevented from bowing out apically, where they

are distant to the labral or other sheaths. Forbes (1969)

mentions that coadaptation of contour between mandible and

maxilla holds them appressed mechanically and helps to keep

stylet bundle compact, but inspection of his own photos shows

only a weak co-sinuate contour that is not convincing in this

regard. When dissected free of each other, both mandibular and

maxillary stylets are bowed inwardly. Possibly the cuticle is


laid down such that the laminar infrastructure always

maintains an inward convexity. Air space occurs between

maxilla and mandible but this has not been traced to see if

interlocks occur periodically. The air space may normally be

filled with saliva, water or other liquid, which would provide

capillary action that keeps the stylets appressed. However, in

midges (Jobling 1976) mentions mandibular base levers and

adductor-abductor muscles with no specific address to their

possible involvement in alignment.

Innervation. The discovery of neural tubes in the stylets

(Forbes 1966, 1969, 1972, 1977 and references) fit well into

the climate of scientific investigations of the time, wherein

questions on the physiological mechanism of behaviors such as

extension, positioning, penetration, inoculation, taste and

extraoral digestion were forefront in the agricultural theater

(i.e. Spiller 1990). These tubes are referred to as dendrites

instead of axons, because branching can be detected by

counting their cross-sections at various levels in the length

of the syntrophium. Axons do not branch.

Dendrite branches can be found free in the sheath or

imbedded in cuticle where they presumably service external

sensory organs. Forbes was not able to locate dendrite-organ

junctions; however, Honomichl (1978) and Zacharuk (1979),

respectively, found proprioceptors and scolopophorous organs

in their study animals. The actual locations of the cell

bodies have not been established.


The tubes are filled with a clear fluid, contain

neurotubules, and are enclosed in an electron opaque cuticular

sheath. Pollard (1972) presumed the fluid to be haemolymph and

the absence of haemocytes to be due to the impassable

diameter. Fibrillar tubular bodies occur outside the sheath

(Forbes 1972:565). Glia are not mentioned by the authors and

not apparent in their micrographs. So-called multivesicular

bodies have been found, and these are now well-known as

neurohemal vesicles.

Cibarium. This is a food chamber that is variously formed

depending on which structures wall off the mouth during

imbibition of food. It is pulsatile in some insects (i.e.

Hemiptera) but not in others (i.e. Orthoptera) (Snodgrass


Sitophore. The sitophore is defined by Snodgrass (1951)

as an expansion of the hypopharynx that makes up the floor of

the cibarium. Troughlike in shape, it serves to hold

masticated food before delivery to the mouth. He illustrates

it for larval Dytiscus (loc. cit., p. 103, fig. 37G) as a

transverse rod that spans the gap between the mandibles, but

doesn't indicate that it is involved in their anatomy or


Miscellaneous. An amorphous zone, circular in cross-

section, appears at the base of the aphid maxillary stylet in

Forbes' (1966) micrograph, but he doesn't identify or comment

on it.


Adelgid stylets are curious in that they are studied by

embedding and sectioning the whole body (Forbes and Mullick

1970). Sections pass through the stylet bundle as many as 8


Fine pictures of nematode stylet ultrastructure are

available but not applicable to insects. Stylets are

regenerated from arcade-like progenitor cells during molting

(Endo 1985).

Figure 7. Mouthparts of L. sanguineus larva. A) Basal zone. B)
Mesal zone. C) distal zone, clearly identifiable as the gala.
D and E) Maxillary palpifer. The palpifer is scalloped between
these arrows to receive the A-B-C complex. F) Labial palp.


Figure 8. Exuvial halves. A) Antenna. B) Ecdysial line. C)
Membrane. D) Stylet. E) Labrum, corresponding to (A) in
Figure 7. F) Channel from which the stylet can be withdrawn.



Figure 9. SEM of Pyractomena mandible, ventral view. A)
Acetabulum. B) Basolateral opening to the interior. Mandibles
are connected to the head not only by the acetabulum, but also
by tissue that passes into this opening.

Figure 10. Classical drawings of lampyrid larval mandibles
have identified these features. A) Acetabulum. B) Canal. C)
Pore. D) Retinaculum. The outer aspect (E), ensheathing the
canal, will be referred to as the "cone" in lieu of its actual

C ";

Figure 11. Models that might explain channelled mandibles. A)
Canal as an apophysis. Inset shows the opposed epicuticles
expected if lampyrids conformed to De Marzo and Nilsson's
(1986) findings. B) Canal as an apodeme.




Figure 12. Mandible of Pyractomena larva longitudinally
shaved. A) Basolateral opening, corresponding to Figure 9B.

Figure 13. Mandible of Pyractomena transversely shaved.
Ventral view.


Figure 14. Whole mount of a 1st instar P. collustrans, freed
from the embedment plastic after being shaved to the basal
half of its mandible.

Figure 15. Close-up of mandible in Figure 14.

wP .

;- ,!'"
iJ~~ "





Specimen processing, from collection to photograph in

hand, is a long, laborious task, likened by Hayat (1986) to

the bottom of an inverted pyramid. Eggs were deshelled in

batches and the embryos were fixed, rinsed, dehydrated, and

mounted for SEM or embedded for TEM/LM. In some cases

extensive serial sectioning and staining was required of

several specimens in order to compare the developmental status

of the whole mandibular structure for each. In the youngest

embryos, for instance, mouthpart anlagen are so small and

amorphous that in order to identify which contour curvature of

any given photograph belongs to which mouthpart, a three-

dimensional reconstruction needed to be assembled from

photographs taken of all cross-sections (Figure 16). Larvae

were studied for external indications of the moment of

apolysis so that the youngest pharate larvae ("apolytics")

could be secured.

The number of specimens used for SEM and TEM was about

150, and the number of useful cross-sections generated reached

about 200, a number so large as to require that, whether for


3-D reconstruction or not, at least half be photographed at

different magnifications for later collation.

The need for photography decreased as it was gradually

determined what is significant and what is not. Negatives and

positives were developed by hand.


Availability of collection and ease of rearing dictated

which species could be used. Tenebrio molitor (Tenebrionidae)

was selected to fulfill the convention that the same assay be

required of an outgroup for comparison.


Calopteron terminal (Archer and San Felasco Hammock,

Alachua Co., Florida, January) provided pupae and adults.

Their larvae feed underground and are rarely found in numbers.

Pre-pupal larvae come to the surface and congregate on the

underside of a rock or fallen branch to overwinter as pupae

and were collected in this dormant state (McCabe and Johnson

1980; Miller 1988; Young and Fischer 1972). Pupae were housed

in a screen cage with a thin layer of moist sand and a rock at

the bottom while awaiting eclosion. Eggs were transferred from

under the rock to moist filter paper in plastic petri dishes

fitted with a screen ventilation window (Figure 17A). After

the females died, the sand was washed into a vial rack


(Figure 17C) half-filled with water to winnow for any stray


Larvae of Lycus lecontei Green (Arthropod Discovery

Center, Tucson, Arizona) are available in large numbers

throughout the year under fallen saguaro cactus carcasses.

Brought to the laboratory, they were provided with "cookies"

of dried saguaro flesh (chunks whittled down into discs).

Apolytics were identified by their color. Freshly tanned

individuals have a high contrast to their light-brown to dark-

brown mottling, but as a molt approaches, they take on a dull

gray, greasy cast. The latter readily commence larval and

pupal molting processes when sprayed with water, and become

sluggish and unresponsive to pin-pricking. These individuals

can be decapitated and verified for separation. Apolysis is

long over by the time they cement their anal proleg down.

Pupae likewise readily molt to adults and eggs can then be

secured from them in the same way as with C. terminal. No

external indication of separation, such as retraction of

ocular pigment, could be found.

Adults, larvae and exuviae of Lycus sanquineus (Pena

Blanca, Santa Cruz Co., Arizona; Davis Mts. Resort, Jeff Davis

Co., Texas, July-August) were collected in small numbers.

Adult females readily laid eggs. Larvae were found under

rocks. Attempts to induce larvae to molt failed.

Three unknown lycid larvae were collected under dead pine

bark in Gainesville in February. Two were reared to Plateros


sp., and the third was preserved in alcohol for low

magnification study.

Archbold Biological Research Station (Lake Placid,

Highland Co., Florida, May) was scouted for Lycus lateralis

(Melsh.), but the population could not be located. It was

either inactive, or extinct in the area at the time.


Adults and larvae of Photuris A (Lloyd 1965) (Gainesville,

July) were collected by their lights at night. Housed as

above, they provided eggs and one pharate pupa. It seemed

interesting to process this specimen also, for a glimpse of

the change from larval mouthparts to adult mouthparts.

Housed as above, males and females of Photinus

collustrans (Gainesville, March) produced eggs. Larvae are too

difficult to find in the wild (Wing 1988), and attempts to

rear them from egg to second instar failed.

Exuviae of Pyractomena lucifera and P. borealis

(Gainesville, January) were collected off tree trunks

(Buschmann 1984).


Tenebrio molitor was auditioned as a non-cantharoid

subject because of the ease with which it can be cultured.

Stock was obtained from Rainbow Mealworms Inc. (Compton,

California) and reared in plastic shoeboxes with 2 inches of


flour, bran and yeast as a nutritious support matrix. Potato

'slices were added thrice weekly for food and moisture. Larval

and pharate larval mandibles were examined for any indications

of composite structure.

Various approaches were tried to develop a method for

determining whether any external indications of apolysis

exist. As a first attempt, these slices were turned over and

the writhing masses of larvae were scanned for individuals

that might show indications of onset of molting. Sufficient

time was spent to reach the realization that such an approach

is futile.

As an alternative, groups of 3 mid-instar larvae were

isolated in media vials, the bottoms of which were tightly

fitted with a wafer of fresh potato. The vials were held in

stacking trays, capacity 30 (6 rows x 5 rows). Two such trays

were prepared. With this arrangement, larvae on top of their

wafer floor scraping for food, a total of 180 could be scanned

rapidly for any indications of onset. After several hours, the

earliest post-apolytic stage that could be obtained with this

approach was a point half-way through eclosion. This indicated

that the apolysis-ecdysis transition is rapid and indiscrete.

Groups were increased to 4 and 5 per vial, but results did not

improve. Other difficulties with this approach were the drying

out of wafers and the accumulation of excrement in the plane

of feeding.


A final approach, barely satisfactory for the needs of

this study, was to place several dozen larvae in a 3" petri

dish fitted with two '" thick wafers that together covered

about 80% of its floor. By turning them over in alternation,

one was eventually spotted that was slightly curled and lying

on its side. Removed to the microscope stage, the stemma,

spiracles and other key areas were monitored for indications

of separation.

This larva set up anteriorly-directed, undulative

spasmodic movements within 3 minutes of the above cessation of

feeding, and the outer cuticle split mere seconds later. No

anatomical indications of separation were detected. It was

decapitated immediately and processed for microscopy.

A second larva was obtained in the same way for

additional microscopy technique. Because of the rapidity of

the molting process, it is suggested that some electronic

means of detecting onset, such as tapping into the nervous

system, might be the best way to study the pharate larval

stage of this species.


The summary of microscopy technique below is designed to

convey the special considerations that arose in this study and

assumes the reader has an understanding of the basic regimen.


In my hands, a modified Karnovsky's consisting of 2%

gluteraldehyde, 1.5% formaldehyde, 1.5% sucrose and a few

drops of acrolein in 0.1M sodium cacodylate buffer yielded

satisfactory results for embryos and tenerals. Specimens were

left in cold (40C) fixative overnight, then rinsed with buffer

for 1-2 hours. Carnoy's, Bouin's and FAA (formal acetic acid)

were not used because they are coagulant fixatives that create

serious artifacts. Although still widely used today, they are

strongly discouraged by leading histologists (e.g., Hayat

1981, 1986).

Specimens are easiest to divide up for SEM or TEM/LM

during the buffer rinse. Those selected for SEM can then be

osmicated as it does enhance electron opacity. Osmication is

not recommended for TEM and LM because the resulting light

opacity makes it too difficult to orient the specimen relative

to the block face (Figure 18). Also, osmium reacts unfavorably

with LM dyes.

Specimens usually float in fixative because of cuticular

waterproofing and air in the tracheae, and attempts were made

to make them sink. The surfactants known as Aerosol O.T.*

(Sargent) and Photoflo* (Kodak), vacuum, and gentle, downward

prodding with a rod were tried with variable success.


Staging. Insects such as Drosophila have features that

allowed investigators to divide the entire embryonic process

up into numerous periods according to the appearance of

conspicuous anatomical landmarks (Campos-Ortega 1985). They

are available in large numbers, have a low yolk content and

will complete maturation while submerged in refraction-

enhancing oils (paraffin oil, Voltalef oil, immersion oil). I

examined and tested my animals for the feasibility of staging

as a preliminary to the main thesis, but they possess none of

these features.

However, incipient segmentation is visible in anatreptic

specimens without dechorionation, and the mouthparts are mere

bulges at this time (Figure 19). Therefore, only stages

following this point were required and, given enough

specimens, a sufficient spectrum of intermediates could be

harvested by processing a large number of batches

indiscriminately. Age differences were then inferred by the

length and curvature of the mandibles.

Dechorionation. Dechorionation was needed only initially

to learn key features that allowed for a rough estimation of

age. Thereafter, transmitted light sufficed. It was

advantageous to leave some rings of chorion intact for

strength, oxygen and waterproofing. Fully or partially

dechorionated eggs desiccate quickly and require that a moist

atmosphere be maintained. Using upward motions, the egg is


rolled on double stick tape within a corral that will catch

the egg should it flick off the adhesive (Figure 17B). After

the desired amount of chorion has been removed, the egg is

rolled off the tape and lifted with a camel's hair brush.

Cracking the eggshell. The egg is compressed with

tweezers that also embrace a pin ca. 0.3mm in diameter

(Figure 20). This protects the embryo from being squashed.

Cracking is best done under buffer or fixative. Dechorionated

eggs are flaccid and cannot be cracked in this manner. Also,

they do not fix with the phase-partition techniques of Zalokar

(1971, 1977) either, probably because the vitelline membrane

is impervious to heptane. This being the case, the only method

available for accessing the internal milieu of a dechorionated

egg is to poke holes with a pin and try to rip the shell

without damaging the embryo within.

Rinsing. Agitating the opened egg in buffer will rinse it

of yolk that otherwise adheres to the embryo during the

fixation process. This is especially desirable for SEM


Fixing. Opened eggs were soaked in fixative for at least

an hour before deshelling. This gave the embryo additional

strength to withstand any distortion that may occur while

peeling. Once freed from the eggshell, the embryo was left in

cold fixative overnight. Only a few specimens were kept in

each test tube so that those on the bottom were not fixed in

a compressed state by specimens on top of them.


DeshellinQ. Deshelling is required for all microscopy

techniques, and should be done under fixative. The most

difficult aspect is freeing the shell from the mesothoracic

spiracles which adhere tightly to the air pores. Careful

ripping and shredding of the shell with one pair of tweezers

while holding with another is the usual method.

Fracmentizing the embryos. It seemed advisable to cut or

lacerate the embryo to allow rapid entry of fixative. Controls

were not run to determine its actual benefits, but when the

mouthparts are hidden in the curl of the body, then the body

must be broken in half anyway. Therefore, a decision must be

made as to whether to fragmentize before, during or after

fixation. The uninsulted body probably stands up best to the

harshness of the dehydration process. Further, whole embryos

are easier to position on a stub, and their brittle nature

allows for clean fracturing of unwanted body parts

(Figure 21). Some of the SEM preparations showed blebbing,

perhaps due to gasses entering between folds of cuticle during

critical point drying (Figure 22). For embedment preparation,

it is sometimes useful to remove unwanted body parts so that

the cross-sections are less confusing, and so that the knife

need not endure the cutting of other body parts before it

reaches the mandibles.


Tweezers were sanded to ultra-sharpness for the delicate

task of teasing the apolysed head out of its exuviae. The

,major obstruction was pulling the crop through the narrower

pharynx (Figure 23). Their mouthparts and those of tenerals

were cut off at the frontal region and taken directly into


Fixation was bypassed for mandibles of alcoholed larvae

and exuviae; these were taken directly into dehydration, which

helps to clean their surfaces and avoids taking water into the

EM column. The hard exoskeleton of these and live, post-

teneral larvae make TEM prohibitive. Fixative cannot penetrate

into the mandible interior, and knives, whether glass or

diamond, cannot stand up to it.


A full ethanol series (25-50-75-95-100%) was required

after rinsing of fixative. Specimens were left in 95%

overnight. Critical point drying was used instead of

hexamethyldisializane (HMDS, Nation 1983), although it can

give satisfactory preservation also (Figure 24). HMDS is

convenient when dehydrating hard mandibles, where, as

mentioned above, the concern is to avoid bringing water into

the EM column. Whole embryos can be bagged with fine mesh


netting for critical point drying. Fragments can be held in a

Beem* capsule modified to hold Milliporee filters (Figure 17D).

Mounting for SEM

The specimen holder used in dehydration was placed under

the microscope in the same focal plane as the gummed SEM stub

so that transfer could be done quickly and smoothly. Dried

.specimens were arranged so that the area to be viewed faced

upward. Lifted with one tweezer tine made tacky by scraping on

skin of fingers, they were then touched down on the stub in

even orientation. After sputter coating, each row was

underscored with a razor blade. This organization minimized

goniometer movements and greatly facilitated the systematic

perusal of each specimen (Figure 25).

Embedment for TEM

Quetol or hard-formulation Spurr's is recommended for so

hard a structure as a mandible. A plastic-acetone gradation of

25%-75%-100% was carried out at room temperature. Infiltration

was protracted to several days for each step since whether

tanned or not, the narrow, cuticularized passages are very

difficult to infiltrate.

It is usual that the specimen sinks to the bottom of the

mold during polymerization. This complicates trimming for two

reasons. Firstly, the bottom is a coarse surface, leaving an

opaque, frosted imprint on the hardened block. The frost, and


the meniscus at the top, do not permit viewing to estimate the

orientation of the head. Secondly, trimming requires that

several millimeters of plastic surround the specimen, so that

there is room for a block face to be constructed around the

desired area of approach, and for a base where glue is

applied. Blocks can be turned upside down in their molds and

a fresh, thin layer of plastic added. This layer will be flat,

transparent, and an interface will not develop which might

otherwise complicate sectioning.

Mounting on a microtome chuck stub

Because of the expected complexity of mandible

embryogenesis, it seemed likely that accurate knowledge of the

angle of approach in the microtome would greatly facilitate

micrograph interpretation. The best references for this are

the ocular pigment masses, which can be eclipsed through any

of the five sides of the block apex. Eclipsing through the

block face is probably best, yielding a lateral approach to

long sections of the outer mandible (Figure 18). The inner

mandible is then reserved in case remounting at a different

orientation is desired. LKB 8800 Ultrotome microtomes, used in

this study, are designed so that the chuck can receive a

cylindrical stub, to which the block is glued. Since it is

best that the block face be perpendicular to the axis of the

chuck stub and well supported at its base, the base needs to

be perpendicular also. Because of refraction, varying


translucency of body parts, prismatic reflections and other

difficulties with viewing specimens en bloc, this requirement

is very difficult to satisfy.

To accomplish it, the block face plane is estimated and

marked with a razor blade under the dissecting microscope. A

coping saw is used to trim very conservatively using this

guideline. Specimen-to-base distance should be about 10mm, to

face, 5mm. The block is then mounted in a flat-jawed chuck and

polished with the microtome. The meniscus is removed at this

time. Trimming continues on the microtome with occasional

recourse to another dissecting microscope until the specimen

is close enough to the face that it can be viewed with minimal

refraction, yet deep enough that the face angle can be

corrected when, inevitably, the compound microscope shows that

it is incorrect. To view under a compound microscope sometimes

requires whittling the stub (Figure 26). Now that the eclipse

is virtually true, the block can be glued to the stub. The

stub is stood on a slide and the block is propped with

fragments of plastic such that the eclipse can be viewed

through a compound microscope. Cyanoacrylate (Crazy Glue')

penetrates between the fragments with a low capillary tension

and will not disturb the prop. Now that the inter-ocular axis

is parallel to the stub axis, the stub can be mounted in the

cylindrical chuck and the face can be honed so that it is

perpendicular to the microtome arm advance. The sides can now

be trimmed to a rectangle so that the specimen can be viewed


through them under the compound microscope without prismatic

reflections. Older compound microscopes generally have enough

clearance to give at least 400X to the apex of a stub lying on

its side. The final trimming strokes to a trapezoid shape can

be made after the shape of the mandibles is noted or,

preferably, photographed. Tedious recourse to and from

microscope and microtome can yield a ca. 97% accuracy in

angle of approach. Trimmed blocks are very fragile and should

be housed in a cropped Eppendorf' tube (Figure 17E).


The late embryonic P. collustrans mandible is about 40l

in basal diameter allowing for a maximum of 800 500A sections,

far beyond what is required for this study. It was determined

during development of methods that the best sectioning pattern

is semithin (lg) for LM and ultrathin (silver, 500A) for TEM

in alternation (Figure 27). Semithin sections of cuticle-

coated specimens can be fragile enough that at least two

should be taken at a time so that at least one might reach the

microscope intact. If more than two are taken, the risk is run

that a critical area of the specimen will be passed over

without securing an ultrathin section for it. Plastic adheres

to the cuticle very poorly and parts of semi-thin sections can

fold over or fall out if the section is fished from the boat

with a camel's hair. Use of a platinum wire loop solves this

problem. Sections are floated onto a drop of water on a slide,


dried, stained with Toluidine Blue in 1% sodium borate buffer,

rinsed, dried, and mounted with Pro-texx.

The block face is best prepared for the silver sectioning

to follow by advancing short of the second full micron so that

a third pass gives ca. 0.1A. The only consideration involved

in deciding on the number of ultrathin sections to take at a

time is abrasion to the cutting edge. At least two are

necessary since the one adhering to the cutting edge is liable

to be damaged when teased. 1x2 slot grids are required for TEM

since there is only one locus of interest per section. Uranyl

acetate-lead citrate are used as post-stain. In most cases all

need not be post-stained. LM slide libraries, as well as SEM

photographs, serve as the primary focus of research in being

well suited for the gross orientation required of this study.

TEM grids can be recoursed to for a more detailed look where



Exterior and interior views of post-embryonic mandibles

can be obtained in a variety of ways. Straight-forward viewing

of mandibles under a dissecting microscope requires at least

,80X. A Zeiss with 80X, another with 96X, and Leitz with 110X

were used in this study. Reflected light can be used with a

compound microscope to go beyond these magnifications at the

expense of depth of field. This is helpful for cursory

examination as it bypasses the processing required by the SEM.


Transmitted light is useful for preliminary examination of

lycid mandibles and this view is accentuated with any of the

refraction-enhancing oils.

Interior views of post-teneral mandibles can be had by

embedding in resin, shaving to the desired cross-section,

dissolving away the resin, then mounting for SEM (Figure 12

through Figure 15). The critical point dried embryonic

mandible is brittle enough to allow chipping without

embedment- the SEM stub to which the embryo is affixed can be

clamped in an ultramicrotome even though it is designed for

resin blocks. The magnification of the dissecting microscope

fitted on the ultramicrotome is too low to observe progress

but the edge of the knife can be watched for fragments (Figure

21). It is advisable that photos be taken of specimens before

and after this fracturing.


An efficient method of cross-referencing notes and

micrographs to specimens was devised in consideration of the

fact that specimens are too small to be remembered by some

idiosyncracy, as is done with, for example, pinned insects.

Scanning electron microscopy. A plastic petri dish served

as a desiccator jar platform for SEM stubs. It was poked with

holes and the holes numbered in reference to a logbook.

Specimens were given a number according to their rank and file


on the stub which was entered on a finder sheet (Figure 29)

and dialed into the micrograph.

Transmission electron microscopy. Numbers were written on

the bottom of block stubs for reference to photographs and

drawings that indicate angle of approach and progress. Grid

box number was referenced to micrograph number. LM slides were

'referenced according to the box number of the two grids they

are between.


The cost of Polaroid 55 sheet film is currently more than

$2.00 per sheet, while Tri-X Pan 35mm film can be purchased in

100 foot rolls at about 11i per shot. The difference in

quality is negligible.

Artifacts and complications

Several annoying complications appear in one step or

,another of the extensive processing required of specimens.

These are mostly tolerated as they do not detract sufficiently

from the quality needed for this study.

Shrinkage and collapse of tissue occurs quite often but

usually in areas other than the mouthparts. Distortion of this

sort is easy to recognize and does not cause


So-called "post-stain garbage", probably Pb(CO3)2

(Figure 73), appears as opaque circles that are easily


recognized from actual specimen structure. It is best avoided

by testing the lead citrate preparation on a blank grid before

using it on those with specimens. Extensive rinsing should

also be employed.

Knife marks result from the use of glass knives for tough

material such as cuticle. Even diamond knives would not stand

up to this material for long, and are far too expensive to be

subjected to such abuse anyway. Knife marks are a cosmetic

problem, and to avoid them would increase the consumption of

knives by perhaps tenfold.

Shrinkage of semithin sections is common and causes

annoying transverse ripples (Figure 57, Figure 63). In most

cases these too are only cosmetic and for the level of

interpretation required in this thesis, can be ignored.

Scanning electron microscope stub mounted embryos can be

cleaned of adhering yolk and other debris (Figure 30,

Figure 31, Figure 32) by dabbing with a tiny blob of adhesive

held in the tweezer tines. This is a delicate operation, but

may be necessary if an important area is obscured. Specimens

to be cleaned should be carefully lifted to another stub so

that others are not sputter-coated twice. Sonication might

also work during the buffer rinse, this was not attempted.

Microscopy is now a highly sophisticated technique.

However, the results they produce have interpretative limits.

For cytodynamics, cytochemistry and functional morphology,

other techniques must one day be employed.

Figure 16. Cross-section of the head of an anatreptic P.
collustrans embryo showing the difficulty in identifying which
lobe corresponds to which anlage without 3-D reconstruction.

,Figure 17. Materials used. A) Petri dish with window. B)
Dechorionation of egg with two-sided tape. C) Vial rack. D)
Eppendorf" vial with tip cut off. E) Beem" capsule modified for
minute specimens.

A r .. B


I L'kk

Figure 18. P. collustrans embryo embedded and photographed
through a compound microscope. See Figure 26. The ideal
orientation is shown, with the eyes eclipsed to the viewer.

Figure 19. Anatreptic P. collustrans embryo, embedded and
photographed through a compound microscope. See Figure 26. A)
Mandible anlage. B) Maxillary palp anlage.



Figure 20. Cracking the eggshell.

Figure 21. Whole embryo affixed to an SEM stub in its native,
curled state. The curl is then broken so that the mouthparts
face upward.


Figure 22. Blebbing of cuticle in an improperly processed

Figure 23. Pharate larval Lycus lecontei, dorsal view, fixed,
then removed from its exuviae. Central gap is where stomodeum
passes through mouth when the head is pulled out of the
exuviae. A) Labromandibular rudiment. B) Gala. C) Maxillary



Figure 24. Trial specimen, a mole cricket embryo, dehydrated
with HMDS.


Figure 25. Well organized SEM stub with all embryos facing the
same approximate way along scored lines so that magnification
and goniometer adjustment is not necessary for perusal from
one specimen to the next.

_]_ *. 'J ^ ^


'''~ ?; "

/ 1

i i~t~'S:

Figure 26. Viewing embedded embryos through a compound
microscope. Whittling of the stub was necessary. Figure 18 and
Figure 19 were photographed by this means.



Figure 27. Sectioning sequence, ultrathin alternating with

Q] /
' /


Q flTX
4 4

Figure 28. SEM stub mounted in a microtome chuck so that tip
of specimen's mandibles can be shaved. Large arrow points to
fragments of the mandible tip.


.. .. .. .. .. .. .

- -

Figure 29. Finder sheet for SEM stub specimens.

1^ q 3 4 5 6 72

8 9 10 11 12 13

1A 15 16 17 1819 20
21 22 2324 25 26 97
2829 3 39 33 34
S35 36 37 .'8 30 4 0
41 42 43 44
45 4647
1 11 2 1 3 1
2 12 22 32
3 13 23 33
4 14 24 34
5 15 25 35
6 16 26 36
7 17 27 37
8 18 28 38
9 19 29 39
10 20 30 40

Figure 30. Blob of glue held between tweezer tines to clean
the stub mounted embryo.

Figure 31. P. collustrans embryo obscured by adherent yolk and
other material.


A ._.___ .i,,, V


4% .' K

Figure 32. P. collustrans embryo cleaned by the method in
Figure 30. Such results are rarely so good as most debris is
firmly dried to the embryo. A) Invagination.


-~--;--- ----"LYY~c'


Normal lampyrid and lycid embryos undergo the typical

blastokinetic movements, anatrepsis and katatrepsis.

Anatrepsis occurs when the body elongates to the extent that

the abdomen curves over the dorsum of the body, as in

Figure 33A. Katatrepsis follows, and is characterized by a

convulsion that reverses this curve so that the abdomen is

positioned ventrally, as in Figure 33B. During this

convulsion, the embryo breaches the serosa (Anderson 1973).

Throughout this dissertation, I shall refer to embryos that

had not undergone katatrepsis prior to fixation as

"anatreptic" embryos. Similarly, "katatreptic" embryos are

those that have undergone katatrepsis.

Two blastokinetic mutants were discovered among the lot

of Calopteron eggs harvested, one in which katatrepsis is

omitted and another in which katatrepsis is accomplished but

the embryo fails to breach the serosa and develops within it

(Figure 33).

Mature Photuris eggs possess extreme tolerance to

inanition. Several of a batch were mistakenly allowed to dry.

They suffered as much as 80% loss in spherical volume when the

shell collapsed (Figure 34). The embryos could be seen in


severely contorted postures by the coloration of the eyes,

setae and mouthparts that had been achieved. Amazingly enough,

when rehydrated, the embryos slowly expanded and closed to

healthy, ambulating, luminescing 1st instars.

It will be easiest to begin this exposition of mouthpart

ontogenesis with a description of the finished products, then

follow with the earlier stages leading up to it. The term

"syntrophium" will herein replace the structure classically

called "mandible."

The Larval Lampyrid Syntrophium

On encountering a prey item, such as a snail, lampyrid

larvae will attack by impaling the foot with a persistent bite

that causes the foot tissue to lacerate when the snail

withdraws (Cicero, unpublished observations). Imbibition of

the slime is presumably through the canal, but no experiments

have been undertaken to prove this. Their syntrophia

articulate laterally only; they cannot perform dorsoventral

grinding movements.

The mature syntrophium

This structure is very long, ~4X the basal diameter of

the antenna. It is a symmetrically paired structure with a

dorsally deflexed (Figure 35A), interolaterally expanded base

(Figure 38B), a ventrolateral acetabulum/condyle articulation,

a shank that curves inward 900 and upward ~100 to an acuminate


tip, a minute retinaculum in the center of this curvature, and

an internal canal. The tips angle further upward because the

whole affair is slightly rotated about its long axis (Figure


Transverse and longitudinal series can be compared for

most of the particulars required to understand its internal

structure (Figure 35 through Figure 46). The canal is married

to the inner wall of the cone for its entire length. It opens

to the buccal cavity dorsally, extends through the cone's

shank, and ends in a slit-like pore. The pore is located at an

outer, anteapical site (Figure 46, Figure 47, Figure 48). A

seam can be detected near the pore in mid-instar Pyractomena

and Photuris.

At the base, where it opens to the buccal cavity

(Figure 42), the canal does assume the motif of an evagination

as modelled earlier (Figure 11A). The tip, distal to the

pore, is entirely cuticle, suggesting that cells back off as

they lay the cuticle down.

At the cytological level, most striking is the high-

nucleus-to-cytoplasm ratio in this area. These cells,

continuous with those of the head epidermis, run up into the

shank where they undergo extreme attenuation to accommodate

the apical diameter of the tip (Figure 49). A large,

apparently disc-shaped, amorphous area occurs near the base

which, oddly enough, appears to be open to the cytoplasm of

the cells surrounding it (Figure 38), as is a similar region


noticed in the antenna. There may be a partition present that

is too fine for the processing to detect. This region

resembles that found in the aphid stylet, mentioned in Chapter

2 (Forbes 1969:560).

Innervation is indicated through to the pore by

dendrites. These probably service the campaniform sensilla and

branched seta that occur at the base of the pore (Figure 50

through Figure 52).

The younger syntrophium

For the youngest specimens, extensive cross-sectioning

was necessary to determine which lobes were facial and which

were cervical. The syntrophium is the lobe just above the gala

and below the front (Figure 53 through Figure 56). It is

produced by a compound epidermal cell field that grows outward

in length as a short, globose structure, rounded at tip and

.preserving a spacious haemocoelic lumen. Its apex begins to

invaginate (Figure 32, Figure 57, Figure 59). The lumen fills

with cells as curvature commences so that only narrow conduits

are present which, presumably, are for haemolymph transport

(Figure 67). An extraordinarily high nucleus:cytoplasm ratio

is noted at this age also (Figure 58).

Sections were scrutinized for the mechanism by which

curvature is accomplished. Possibilities include a greater

density of cuticle along the outer face than the inner face,

or perhaps a ligature that ties the tip to the front. No


indications were found. There is also no indication of any

association between the invagination and curvature or canal

formation. No developmental consequences of this invagination

were detected.

A molt occurs at about this time of curvature initiation,

and the exuviae of the stomodeum can be seen in front of the

mouth (Figure 57, Figure 59). At commencement of curvature,

the incipient canal appears at the bottom of the dorsal

deflection and progresses in a distal direction through the

center of the shaft (Figure 60 through Figure 67). The dorsal

deflection is far too subtle to identify at first appearance.

Most striking at this time is the generation of highly

electron dense material in cell membranes that are in the

proximity of the presumptive canal path. The canal is fluid

filled and results from a roughly axial zone of cells that

present microvilli toward the center and pull away from each

other. The lengthening of the canal continues during curvature

until, presumably, it breaches to the exterior anteapically

and then lays down cuticle that forms the lining of the canal.

This last moment in attainment of continuity was not detected

in the cross-sections collected.

The Larval Lycid Syntrophium

In contrast to lampyrids, lycid larvae are sedentary

feeders, poking the syntrophium into decaying organic


material, slimemolds, fungi and other sources of food

,depending on the species (Lawrence 1991).

The mature syntrophium

This composite structure is very short, only ca. 1 to

1.5X the diameter of the basal antennomere and retractable

between prognathous- opisthognathous positions. When in the

posterior position, the shank proceeds ventrally for a short

distance, then curves like an elephant tusk directed caudad-

gradually diverging outward, then curving downward to an

acuminate tip (Figure 68). Figure 69B shows a strong apodeme

that marks the ventral limit of the epicranial half, uniting

it with the labium (Figure 69, Figure 90).

Four paired structures can now be identified as

functional components- the labrum, mandible, gala and the

maxillary palpifer. In Plateros the latter two are

independent. In the other genera studied, Calopteron and

Lycus, the gala is crescent-shaped in cross-section

(Figure 7C), and this crescent is upwardly open to receive the

mandible and labrum as their sheath. The palpifer assumes a

concavity where it rests against the tripartite complex

(Figure 7D and E). Both the gala and palpifer are rugose,

setose and easy to recognize as separate from the mandible.

The association between labrum and mandible can be

followed in cross-sectional series (Figure 70 through


Figure 80). As noted by initial inspection with LM

(Figure 8E), the labrum is longitudinally parted into two

rugose, setose, spindle shaped sclerites which are separated

from the front and the mandibles by a membrane. It emerges

from the anterior edge of the head as a transverse fold as

does any other appendage, such as a wing. Its ventrolateral

surface infiltrates the lumen of the mandibular arc and

coextends with it as an internal stylet. When the dorsal and

ventral exuvial halves of the head are pulled apart

(Figure 8), this stylet breaks off at its base and appears to

be a separate structure.

Nested in the stylet is a rectangular, cuticular sheath

that contains two neural tubes. These tubes are wrapped in

glia and contain a meshwork of microfibrils. The sheath

marries and unmarries itself from the stylet as it travels the

stylet's length (Figure 73, Figure 75).

The labrum and mandible emerge from the head separately

and remain separate for their entire length. That is to say,

no cross-bridges were detected in any cross-section, and the

two can be teased away from each other suggesting that there

is no adherence. The gala, too, remains discrete (Figure 76).

All three are cell-filled. In the L. lecontei gala, the cells

end before the tip, which then becomes an alveolate matrix

created by dehiscence of spheres (Figure 81 through

Figure 83).


Figure 86 through Figure 93 show different profiles of a

pair of exuvial halves stretched over a pinhead (Figure 84) so

that the mouthparts are exserted. These supplement the above

cross-sectional series to give a full understanding of the

arrangement at this age. The mandible issues from the

pharyngeal walls as a transverse process, but the inner margin

abruptly rolls over as the structure undergoes a bend to

hypognathy. As can be seen in the close-up (Figure 87), the

roll is to the interior. Following the roll distally to

Figure 85, it is seen that the mandible is twisting so that

the roll takes this originally inner margin all the way under

itself. The originally outer margin rolls also, although only

slightly so as to curl into the flute. Figure 93 is an inner

ventral picture showing acceptance of this pair by the gala.

The younger syntrophium

The youngest embryos processed were anatreptic Lycus

sanguineus and Calopteron terminal. At this time the labrum

is a shallow, bilobed swelling, higher than wide (Figure 88).

A minute flap, presumed to be the hypopharynx by comparison to

older embryos, appears between the mandibular lobes

(Figure 89).

Scanning electron microscope processing of older C.

terminal embryos yielded

clear views of mouthpart composition (Figure 89 through

Figure 93). As with the post-embryonic mouthpart arrangement,


both the dorsal and ventral surfaces of the labrum are layered

with cuticle; the labrum actually emerges as a transverse

fold. Its lateral margins autonomously take on an inward

fluting and grow out on top of the mandible. The tip of this

flute coextends with the tip of the mandible.

To be certain of the association of these two minute

characters, the SEM stub was mounted on a microtome and the

tips were chipped (as in Figure 28). Figure 91 shows 3 layers,

now identifiable as labrum, dorsal mandibular lip and ventral

mandibular lip. In Figure 92, the fractured tip shows clearly

that the labrum is filled with cellular material and the

mandible retains a small amount also. This figure compares

well with that of Figure 80, which shows a trapezoidal, cell

filled labrum inside the arcuate mandible. No observations

suggest how the stylet inserts itself into the mandible.

The Pharate Larval Lycid Syntrophium

The operation of decapitating and fixing an apolytic,

then extracting the head from its exuviae was exasperatingly

difficult for several reasons. The tissue is very soft, yet

snug in the surrounding exuviae. Considerable tugging and

wiggling is necessary, especially to pull the pharynx over the

crop. Ecdysing larva must also "unswallow" the whole stomodeal

exuviae, but the crop probably collapses to allow smooth

passage. In the case of my specimens, the fixed crop retains

its girth. Also, there was no way to monitor development to