Click here for the original journal page (in Acrobat pdf format).

The text below is grayed out because it is not intended to be read. It is a necessarily imperfect OCR of the original and is only used by a search engine.

Journal of The Lepidopterists^ Society

Volume 14                                           1960                                           Number 2

A STUDY OF FIRST INSTAR LARVAE OF THE SATURNIIDiT, WITH SPECIAL REFERENCE TO NEARCTIC GENERA

by Roger W. Pease, Jr.

Introduction

First instars of many species of Lepidoptera have been described, usually with non-comparative and non-technical descriptions. Occasionally it has been done with an eye toward interpreting structures of the mature larvae and correctly homologizing them with those of other groups. This has been singularly successful, since the first instar represents a link with the past in that primitive characters may occur in this stage which are completely modified after the first molt. This paper will consider first instar larvae in their own right with only passing reference to other stages of development. Such concentration may seem arbitrary until an attempt is made to classify first in-star larvae using a key constructed for mature larvae. The number and degree of structural changes taking place during the first molt require that a separate study be made of the first instar just as exclusive comparative studies have been made of adults, pupae, and later instar larvae.

The original intention was to include only the Nearctic genera of the Saturniidae, but a few species from other areas were available and since their structures suggested some interesting problems, they are included. The subfamilies represented are the Saturniinae, Hemileucinae, Citheroniinae, Rhescyn-tidinae, and Agliinae.

There are about as many systems of setal nomenclature as there have been workers in the field. Only a few have been applied to the first instar larvae of the Saturniidae. Packard (1905,1914), following Dyar, used numbers to identify a few setae, numbering them from the dorso- to ventromeson. He depicted first instar larvae of many species of Saturniidae in color and by line drawings. Unfortunately, setae are omitted from many of the color figures, and some plates do not agree with descriptions in the text (e.g., Coloradia pandora). Abdominal segments eight, nine, and ten are also sometimes confused. On the other hand, the line drawings, which are Dr. Packard's own, agree more closely with actual specimens. As a means to quick visual identi-

89

90

Pease: First instar larvae

Vol.14: no.2

fication of larvae, the color figures are, of course, unsurpassed, but they should not be considered structurally complete.

Fracker (1929) constructed a key to the mature larvae of North American Saturniidae using his system of Greek letters to identify setae. Characters used in his key will sometimes hold for the first instar. Particularly important are the relationships of the most dorsal setae on segments eight and nine of the abdomen.

In the following discussion the setal nomenclature of Gerasimov (1935) as modified by Hinton (1947) is followed as closely as possible. The almost mathematically precise definitions of groups of setae could be applied to the Saturniidae even though this family was not included in Hinton's paper. The notations of Fracker, Forbes, or Heinrich are alternate systems. For comparisons and synonymy see Hinton (1947) and Peterson (1948).

Those setas which occur throughout the Lepidoptera in the first instar are called primary setae. Secondary setae appear in succeeding instars but are usually absent in the first. It is an indication of advancement in the Saturniidae that some of them have numerous secondary setas in the first instar. Hinton recognized two types of setae: long or tactile setae and microscopic or proprioceptor setae. The former are generally distributed over the body, while the latter are restricted to areas where overlapping of cuticle is likely to occur, such as at the juncture of segments. Most of the microscopic setae on specimens preserved in alcohol were not clearly visible with the resolution and magnification (150X) used. Except for a single series of aberrant Hemi-leuca larvae, the microscopic setae will be omitted from this discussion.

Characters such as crochets, head structures, and thoracic legs are probably as diagnostic as body setae, but they have not been so widely used and are less easily studied. Some reference will be made to the crochets on the prolegs. All structures except the most prominent may be ruined if specimens preserved in alcohol were ever allowed to dry out. The more conspicuous setae are usually in satisfactory condition even in some of the most severely shrunken specimens. It is hoped that the key presented here will eventually be extended to include all the prominent structures of first instar larvae.

A few terms need to be redefined for use with the first instar. If a seta is referred to as PRIMITIVE, it consists of a single stiff filament rising directly from the body wall. Presumably, this is the ancestral form. Chalazae are extensions of cuticle bearing one to several setae. Fracker (1929) defines scolus as a "spinose projection" of the body wall, with reference to the Saturniidae. The scoli of later instars, however, are homologous to structures in the first instar which are clearly divided into a chalaza and a seta. Here the term CHALAZA will be used for any projection on the body wall bearing primary setae which are still distinguishable from secondary setae. SCOLUS will be reserved for similar structures in the Saturniinae having secondary setae on the chalazae which cannot be separated from the primary setae. As a general rule, the setae of first instar larvae of Saturniidae never become fused. It is the chalazae which are fused to a varying degree. For ex-

1960

Journal of the Lepidopterists' Society

91

ample, on the dorsomeson of the ninth abdominal segment, all Hemileucinae have a single chalaza. This bears two primary setae, one from each side of the body, whose chalazae are united. Very rarely an individual larva will have a seta split or forked, but in two individuals seen (one each of Arsenura and Hemileuca), the fork occurred on only one side of the body.

Primary and Secondary Setye

The (primary setae are divided into several groups defined by their position and recurring on each segment. They are: dorsals (Dl and D2), subdorsals (SD1 and SD2), laterals (LI and L2), subventrals (SV1 and SV2), and ventral (VI). In addition there are two extra setae (XD1 and XD2) on the cervical shield of the prothoracic segment. The homologies of the setae on the tenth abdominal segment are not clear at present. The abbreviations T-l, T-2, T-3, and A-l to A-10 will be used to indicate thoracic and abdominal segments numbered from anterior to posterior. When fusion of chalazae takes place, it occurs first within groups of setae and then between different groups. Only the letters of the setal groups will be used to describe such united structures. For example, D would refer to a structure homologous to the dorsal setae (Dl and D2) and XD-SD would be a structure made up of the two SD setae combined with the two XD setae of the pro-thoracic segment.

On the cervical shield of the prothorax (T-l) there are six primary setae. In most Saturniidae four of these are represented by two large bifurcate chalazae or scoli on the anterior part of the shield and two primitive setae on the posterior margin (Fig.l). In the Saturniinae the bifurcation has disappeared, and the two scoli bear numerous secondary setae. Ant her tea and Actias show increasing degrees of fusion of the bases of the two scoli, the top being plainly double in the former and single in the latter (Figs.3 and 4). On the posterior edge of the plate, two single setae rise directly from the shield without chalazae. One is usually mesocaudad of the first chalaza, and the second is about midway between the two chalazae. Dryocampa rubicunda (Fig.2) offers a chance to observe the relationships of the setae when their chalazae are reduced and almost wholly absent. The setae of the more dorsal anterior pair are of subequal length, while one seta of the pair below them is longer than the other. In the higher Lepidoptera, XD1 and XD2 are of equal length and are on the anterior (part of the cervical shield. The subdorsal group (SD1 and SD2) is below these with SD1 longer than SD2. The dorsal group (Dl and D2) forms the remaining setae, with Dl shorter and more dorsal than D2. Therefore, in the Saturniidae the more dorsal of the two anterior pairs of setae on the cervical shield will be called XD, the more lateral SD, the more dorsal of the single setae Dl, and the more lateral single seta D2. Dl and D2 vary in their length relationship. Those on T-l are of nearly equal length in most Saturniidae. In Eacles imperialis Dl is bullet-shaped and much shorter than the normally developed D2.

92

Pease: First instar larvae

Vol.14: no.2

On the meso- and metathorax (T-2 and T-3) the dorsal group is borne on a single chalaza or scolus in all Saturniidae. On the abdominal segments Dl is antero-mesad of D2 and is usually on a chalaza. Dl is always present on A-1 to A-9 but may be fused on A-8 across the dorsomeson, in which case it is flanked on either side laterocaudad by D2 (Fig. 15). D2 is present on A-1 to A-9 in the Hemileucinae, Citheroniinae, and Rhescyntidinae but is fused on the dorsomeson of A-9 (Fig. 18). The Saturniinae may have D2 conspicuous on all abdominal segments (Saturnia — Fig.22), reduced to a weak bristle {Actias — Fig.23), or missing on A-9 and conspicuous on the other abdominal segments (Hyalophora, Samia — Fig.24).

The SD group consists of two setae below the D group. SD1 and SD2 are both tactile on the thoracic segments and are borne on fused chalazae. On abdominal segments, SD2 is prespiracular and microscopic. It usually looks like a small puncture when visible at all. SD1 is borne on a chalaza and is similar to abdominal Dl.

The lateral group (L) is subspiracular on the abdomen and prespiracular on the prothorax. It consists of two setae (LI and L2) on T-l and A-1 to A-8. Only LI is present on T-2 and T-3 although it is usually specialized in the same manner as LI on the abdomen. In the Saturniinae, LI is a well developed scolus on all segments except A-9, while L2 is of primitive form occurring caudad to LI on A-1 to A-8 (Fig. 15). On the prothorax L2 is mixed with the secondary setae of the lateral scolus (LI), but it can often be identified by its thinner more fragile appearance. In subfamilies other than the Saturniinae, LI and L2 may be borne on the same chalaza at equal heights (Eacles imperialis on A-1 to A-7 — Fig.7), on different chalazae (Rhescyntis meander — Fig.8), or on the same chalaza with L2 projecting caudad at halfmast (Auto?neris to — Fig.6). L2 is sometimes inserted slightly ventrad of L2 on A-1 to A-6 and slightly dorsad of L2 on A-7 and A-8. E. imperialis and Neocarnegia basirei have L2 projecting caudad from the base of LI on T-l and A-8. Also there are one to three secondary setae on L of T-l in these species.

The subventral group (SV) normally consists of two setae on T-l and A-3 to A-6, one seta on T-2 and T-3, one or two setae on A-1, A-2, A-7, A-8, and one seta on A-9. The group is situated just above the true legs on the thoracic segments. If more than the allotted number of setae occur, there is usually a larger number on T-l than on T-2 and T-3. The Saturniinae have at least two subventral setae on all thoracic segments arranged in a more or less horizontal line (Fig.5). SV is a scolus on T-l in Rothschildia. The Citheroniinae, Rhescyntidinae, and Hemileucinae have two setae on T-l and only one on T-2 and T-3. On the proleg-bearing abdominal segments, the SV are primitive setae borne on the distal part of the proleg, one lateral and one anterior. In some Saturniinae one or more setae are arrayed horizontally between the sclerotized plate bearing the secondary setae on the prolegs of A-3 to A-6 and the body (Fig.36). None of the species which lacked secondary setae had SV1 and SV2 in this position. Therefore; SV1 and SV2 are probably mixed with the secondary setae.

SATURNIID^

PLATE 1

First thoracic segment (T-l) : fig. 1 — Automeris io; fig. 2 -— Dryocampa rub't-cunda; fig. 3 — Anthercea polyphemus ; fig. 4 — Actias luna. Second thoracic segment (T-2) : fig. 5 — Hyalophora cecropia. Fourth abdominal segment (A-4) : fig. 6 — Automeris io; fig. 7 ■— Eacles imperialis; fig. 8 — Rhescyntis meander) fig. 9 — Samia cynthia. Eighth abdominal segment (A-8) : fig. 10 — Hemileuca maia; fig. 11 — D. rubicunda; fig. 12 — Syssphinx heiligbrodti.

94

Pease: First instar larvae

Vol.14: no.2

On abdominal segments A-l, A-2, A-7, and A-8, the number of sub-ventral setae is variable. Uusually there are two setae arranged vertically on the first two abdominal segments. A third seta sometimes appears between the two. Specimens from the same batch of eggs have differed in the number of setae which they possess in this group even from one side of the abdomen to the other. Except for SV1 on A-l, A-2, A-7, A-8, and A-9, the subventral setae are primitive. SV1 on some or all of these segments may be on a chalaza and follow the modification of other setae of the species under consideration. In some Hemileucinae SV1 of A-7 and A-9 are on chalazae while the chalazae of A-8 are much reduced or absent. In general, the instability of the subventral setae on segments without prolegs impairs their usefulness as taxonomic characters. However, the larger number of setae in this group on the thoracic segments of the Saturniinae in comparison with other subfamilies is a constant character.

There is some question as to whether a lateral or a subventral seta occurs on the ninth abdominal segment (A-9) in the Saturniidae. The setae on A-9 are arranged in a more or less vertical array unlike other abdominal segments. The three most dorsal setae (two in some Saturniinae) are easily identified as the two (or one) members of the dorsal group (Dl and D2) and the tactile seta of the subdorsal group (SD1). The seta close to the ventromeson is obviously a member of the ventral group (VI). The Hemileucinae (Fig. 16) possess one other seta situated below the level of the lateral group on other segments. Either the lateral or the subventral group is missing. The seta and chalaza are similar to the subventral group of A-7 in some species, so the seta is probably SV1. The Saturniinae (Fig.22) have two setae in this area, both of primitive form. A specimen of Hyalophora cecropia was found in which both setae were present on the left side of A-9, but only the outer one occurred on the right side. This suggests that both setae are members of the subventral group in the Saturniinae with the lower seta secondary. The Citheroniinae and Rhescyntidinae also have two setae, but their form varies greatly. Syssphinx heiligbrodti (Fig.18) and Dryocampa rubicunda (Fig.17) have both setae unmodified. However, the outer seta is larger, has a more prominent chalaza, and is level with L of A-8 or above it. In contrast the lower seta of E. vmperialu (Fig.20) and Citheronia is more prominent and resembles LI of other abdominal segments. The other seta is always small in comparison. Although the size relationship varies, the position of the two setae is very nearly the same in Syssphinx, Dryocampa, Eacles, and Citheronia. The higher seta of Rhescyntis meander (Fig.21) has the saw-shaped structure peculiar to most setae on the upper two-thirds of the body in this species, but the lower seta is of primitive form.

A probable interpretation is that the lateral group is typically absent on A-9 of the Saturniidae with the exception of the Citheroniinae and perhaps the Rhescyntidinae. Since A-9 is smaller than the other abdominal segments, the loss of LI may have been due to crowding of the setae and to the increased flexibility of the segment which would result from a more spacious arrange-

saturniim;

PLATE 2

Eighth abdominal segment (cont.) : fig. 13 — Eacles imperialis ; fig. 14 — An-thercea polyphemus; fig. 15 — Actias luna. Ninth abdominal segment (A-9) : fig. 16 Hemileuca maia; fig. 17 — Dryocampa rubicunda; fig. 18 — Syssphinx heiligbrodti; fig. 19 — Citheronia regalis; fig. 20 — E. imperialis; fig. 21 — Rhescyntis meander; fig. 22 — Saturnia pavonia; fig. 23 —Actias luna; fig. 24 — Hyalophora cecropia.

96

Pease: First instar larvae

Vol.14: no.2

ment. Development of a subventral chalaza as in Eacles, Citheronia, and Automeris with a corresponding loss of the lateral seta would serve to protect the larva's flank and to eliminate crowding. Numerous secondary setae on nearby chalazae and on the anal prolegs in the Saturniinae may be sufficient protection without modification of the subventral setse.

The ventral seta VI is posterior to the thoracic legs on T-l to T-3. It is usually difficult to detect. On A-l and A-2 VI is mesad of the subventral setae. It is on the inner side of the prolegs of A-3 to A-6 and near the meson on A-7 to A-9. On A-8 it is often nearer the meson than on A-7 and A-9.

The tenth abdominal segment in its probable primitive form (Hemi-leucinae) bears five setae on the suranal plate, five on the lateral sclerite of the proleg, one seta on the anterior of the proleg, and three stubby setae on the mesal side of the anal proleg (Fig.25). There is a tendency towards fusion of the three most anterior setae on the suranal plate. All setae on this plate are separate in Rhescyntis meander (Rhescyntidinae) and in Dryocampa and Anisota (Citheroniinae). Progressive fusion is clearly indicated by comparison of Syssphinx heiligbrodti (Fig.28), Citheronia regalis (Fig.29), and Eacles imperialis (Fig.30). In the first of these, the three setae are fused only at the base, in C. regalis they are borne well up on the chalaza, and in E. imperialis a secondary seta or two are present. Saturniinae have a scolus on the suranal plate with two or three setae inserted behind it (Fig.33).

Secondary setae, when present, are usually limited to the lateral plates of the prolegs and to the chalazae of the primary setae. All Saturniinae have secondary setae on the more prominent chalazae. New World genera may have thorny cuticular eruptions of the cuticle of the chalazae but usually have no secondary setae on them. Eacles has extra setae on the chalaza L of T-l and the fused chalazae of the suranal plate. Comparison of Hemileuca, Automeris, and Hyalophora cecropia might suggest that the secondary setae of the Saturniinae, which are mixed with and inseparable from the primary setae, have developed directly from the cuticular processes of the Hemileucinae. These range from the tiny bristles of Hemileuca to the conspicuous spines in Automeris. That this is not the case is indicated by the fact that some genera of the Saturniinae (Saturnia and Calosaturnid) lack fusion of Dl on A8, and all have D2 of A-9 unfused. These setae are almost always fused on the dorsomeson in New World species (exceptions are Dryocampa and Anisota which have only D2 of A-9 fused). In Aglia tau Dl is fused on A-8, but because of the size of the chalazae on A-9 it is not certain whether they were once fused or whether they merely cover more surface of the dorsum and hence appear to approach fusion.

A series of specimens of Hemileuca maia from Long Island may shed a little light on the method of development of secondary setae. In these, extra tactile setae occur on the dorsum of the mesothoracic to eighth abdominal segments and tend to displace the microscopic seta MXD1 which is lateroanterior to Dl (Fig.48). Three times an extra seta was also found near D2. In one larva, both MXD1 and the tactile seta are present anterior to Dl (Fig.51).

SATURNIID^

PLATE 3

Last abdominal segment: fig. 25 — Automeris io — dorsal plate, lateral aspect of proleg and ventral view (half only) ; fig. 26 — Eacles imperialis — lateral plate of proleg; fig. 27 — Neocarnegia basirei — lateral plate of proleg. Dorsal plate: fig. 28 — Syssphinx heiligbrodti; fig. 29 — Citheronia regalis; fig. 30 — E. imperialis; fig. 31 — N. basirei; fig. 32 — Rhescyntis meander; fig. 33 — Samia cynthia; fig. 34 — Aglia tau.

PLATE 4

SATURNIID^E

36

39

38

41

Abdominal prolegs (A-3) : fig. 35 — Actios selene; fig. 36 — Hyalophora ce-cropia; fig. 37 — Attacus atlas; fig. 38 — H. cecropia —■ anal proleg; fig. 39 — Citheronia regalis with crochet remnants; fig. 40 — Rhescyntis meander crochets; fig. 41 — Rothschildia orizaha crochets.

1960

Journal of the Lepidopterists' Society

99

In all other individuals, when a tactile seta occurs anterior to Dl, its position is close to where MXDl would normally be, and MXDl is not visible. In view of the presence and proximity of both MXDl and the tactile seta on one specimen, it is not certain that this tactile seta represents a mutation of the microscopic seta. The occurrence of the extra tactile seta was not constant from segment to segment nor even symmetrical from left to right. On the same specimens, there is a tendency for the chalazae D2 of A-9, which are fused on the dorsomeson, to separate and fuse with Dl. The tactile seta in some instances even migrated part way up the chalaza of Dl so that the structure began to resemble the fused D group of the thoracic segments (Fig. 50). Jt is interesting to note that Pseudohazis has three setae rising from chalazae XD, SD on T-l, and SD on T-2 and T-3. This seems to be constant. Packard (1914) figures first instar larvae of Pseudohazis with three-forked chalazae. Related species have only two setae on these chalazae. The source of the third seta is not clear.

The secondary tactile setae of the Saturniinae may have appeared by a process of random mutation, with selection collecting the setae on the chalazae. It may be significant that the setae sometimes migrated up the stem of the chalaza in some of the Hemileuca maia. [See the chart below for a complete list of these setae.] Another explanation would be that the secondary setae of the Saturniinae arose by a lengthening of cuticular processes similar to the development of the processes indicated by comparison of Hemileuca and Automeris. As mentioned above, the Saturniinae could not arise from the Hemileucinae directly.

Hemileuca maia — distribution of aberrant MXDl setae on 5 specimens from West Hampton, Long Island, New York, ova collected by S. A. Hessel.

SPECIMEN

I               TYPE OF SETA ON SUCCESSIVE SEGMENTS 1

T-2 T-3 A-l A-2 A^-3 A-4 A-5 A-6 A-7 A-8 A-9

1  Symbols: M = microscopic; TfD — tactile seta fused to Dl; T c= tactile.

2 D2 double.

s #1 has no fusion of D2 on dorsomeson of A-9.

PLATE 5

SATURNIID^E

HEAD-}

Dorsal setae (D) of second thoracic segment (T-2) : fig. 42 — Hemileuca maia; fig. 43 — Automeris io; fig. 44 — Citheronia regalis; fig. 45 — Syssphinx heilig-brodti; fig. 46 — A. io showing position of eversible gland on first abdominal segment (A-1). Aberrant setae of Hemileuca maia: fig. 48 — normal left side A-2 of specimen #1; fig. 49 — MXDl abnormal right side A-2 specimen #1; fig. 50 — extra setae associated with Dl and MXDl apparently missing on right side A-1 specimen #5; fig. 51 — extra seta associated with Dl and MXDl present on right side A-2 specimen #4.

1960

Journal of the Lepidopterists' Society

101

Aglia is unique among the species examined in that it possesses a lengthy XD group on T-l and D group on T-3, while T-2 has a much smaller D-chalaza bearing two long and two shorter setae. Several setae are behind this chalaza on the segment. It is probable that the two long setae represent Dl and D2, which are fused on T-l and T-3. There are a few secondary setae scattered over the body, and eruptions of cuticle on the long chalazae of T-l, T-2, and A-8 bear single secondary setae. Aglia also has a remarkable chalaza projecting caudally on the dorsomeson of the tenth abdominal segment (Fig.34).

Crochets

The most general arrangement of crochets seems to be a single series of crochets of similar size on the inner side of the distal part of each proleg. The embedded ends are distinctly bent or hooked as well as the free ends of the crochets except in the Saturniinae, which have the embedded ends blunt. When the crochets are deeply set in the fleshy pad (spatula) on the proleg, the connections between the ends are obscured and there is an appearance of two rows of crochets where only one occurs, as in Rhescyntis meander (Fig.40).

Crochet remnants, from what may once have been an outer row, are found in some specimens of Citheronia regalis and Eacles imperialis (Fig.39). Not all specimens have them. In the Saturniinae and in atheroma, Eacles, and Syssphinx of the Citheroniinae, the crochets in the middle of the series are shortened forming a more or less C-shaped pattern (Fig.41).

The number of crochets seems to be distinctive, as the following counts for Saturniinae show:

Ant her tea, Actias, Calosaturnia                        18 or more

Saturnia                                                              15-18

Hyalophora, Samia                                            13-16

Roths childia                                                       10-12

Counts apply to prolegs on the third to sixth abdominal segments. A larger number of crochets always appears to the anal prolegs.

The number of crochets may be related to the size of the larva, at least in some of the New World forms, atheroma, Eacles, and Rhescyntis meander are large and have at least 20 on each segment, while the small Dryo-campa ruhicunda has 7 to 9 crochets, and the Hemileucinae have 6 to 9. Attacus atlas, however, has only 13 to 16 crochets, as do Hyalophora and Samia.

Evolution and Taxonomic Value of Larvze

Characters in the immature stages of Lepidoptera have long been used for aid in defining taxonomic groups. First instar larvae, in particular, show less variation within taxa than adults. Species within genera may be almost identical.

102

Pease: First instar larvae

Vol.14: no.2

Because of the almost cataclysmic changes which take place during metamorphosis, speculation is justified as to whether larvae, including the first instar, could diverge faster than their imagines in the course of evolution. This would seem possible, since in Lepidoptera most external structures of the imago do not develop from external larval structures but instead arise at the end of the larval period from small, previously quiescent groups of cells, the "imaginal discs." Larval-specific selective factors could not affect them nearly as much as the larval structures. Predators and parasites are examples of such selective factors. Many are particular as to the stage (egg, larval instar, pupa, or adult) which they attack but not as particular about the species (Howard and Williston in Scudder, 1889). Since immature and mature stages need not occur at the same season, the effects of climate may differ. The very great mortality which occurs between oviposition and pupal eclosion and the relatively long duration of this period in most species of Lepidoptera, indicates that the brunt of natural selection (aside from courtship recognition factors) is borne by the immature stages. For example, in a quantitative study of natural populations of Phlegethontius sextus (Johan.) and P. qiiinquernaculatus (Haw.) (Sphingidae) Lawson (1959) found 98 per cent mortality in larval stages alone. Yet, species differences are usually assumed to be greater in adults than larvae.

A model which might result in larval divergence follows. Suppose that some selective factor was focused on larvae of several species but affected the imagines not at all. Random variation could remain the same for individuals in all stages, but since selection was concentrated on larvae, net evolution would be greater in larval characters than in those of the adult. It is hardly to be expected that all species would adapt in the same way. Some would become more unlike as a result of selection. Meanwhile, no adaptation to this factor has been required of the adult. Thus, it seems possible that larval differences could accumulate faster than those in adult structures.

Species could not evolve solely by larval adaptation since the sexually mature stages must be reproductively isolated. Thus, for populations of the same species which were geographically isolated and had evolved different larval characters, the removal of spatial barriers would at first permit interbreeding with resulting larval polymorphism.

A behavioral difference in the imago, however, might serve to isolate reunited forms although they retained similar imaginal morphology. To a tax-onomist looking at museum specimens, the adults would appear similar but the larvae different. Eventually, the imagines would probably become more unlike as selective factors acted upon them more strongly than on the larvae, their evolutionary divergence overtaking that of the immatures.

The following examples may make this more convincing. A favorite behavioral mechanism for isolating species in the Saturniidae is the time of day or night during which a female will release the scent which attracts males. The time is characteristically different for each species. Fortunately for tax-onomists, adults of these moths usually have conspicuous differences. But, in the Notodontidae, the adults of the genus Datana are difficult to separate al-

1960                                   Journal of the Lepidopterists' Society                                     103

though mature larvae can be identified at a glance. Sibling species such as Erynnis lucilius Scud. & Burg., E. persius Scud., and E. baptisice Forbes (Hes-periidae) have different food-plants, but adults are difficult or impossible to determine in unlabelled dried specimens even from genitalia. A conspicuous color difference makes separation of the first instar larvae of two of the species a simple matter. First instar larvae of E. baptisia are orange while those of E. lucilius are pale greenish white.

First instar larvae should be particularly helpful in taxonomic studies either (1) when morphology of the imagines differs greatly but the larvae retain similar characters, or (2) when larvae have changed and the mature stages have not diverged greatly. Each requires evolutionary conservatism in one stage and divergence in another. The majority of cases fall into the first category. However, a classic example of dissimilar larvae but superficially similar adults is that of Automeris (Hemileucinae) and the Saturniinae. Until Mosher (1914) classed Automeris with the Hemileucinae on the basis of pupal characters and Fracker (1929) reached the same conclusion on larval characters, the genus was often placed in the Saturniinae. Characters have since been found in the adult for placing the genus correctly.

In passing, it should be noted that the setae are quite homogeneous in the Hemileucinae but that there is a diversity of types in the Citheroniinae and Rhescyntidinae. In the first group the larvae have stinging setae, and larval evolution may have proceeded along physiological lines. In the latter groups the product of evolution is more readily seen as the result of setal modification.

Key to First Instar Larvae of Saturniid^:

The key was constructed from the following species, from North America unless a locality is noted. All specimens are in the larval collection at Yale University. HEMILEUCINAE: Automeris coresus Bdv. (Argentina), A. io Fabr., A. pamina Neum., Coloradia pandora Blake, Dirphia baroma Schaus (Brasil), D. curitiba Draudt (Brasil), D. epiolina Felder (Brasil), Hemi-leuca maia Drury, H. neumoegeni Hy. Edw., H. nevadensis Stretch, Hylesia nigricans Berg (Argent™), Pseudohazis spp ; CITHERONIINAE : Syssphinx (Bouvierina) heiligbrodti Harvey, Adeloneivaia apicalis Bouvier (Brasil), Anisota oslari Roths., Dryocampa rubicunda Fabr., Cither oma brisottii Bdv. (Argentina), C. regalis Fabr., Eacles imperialis Drury, Neocarnegia basirei Schaus; RHESCYNTIDIN^: Rhescyntis (Arsenura) meander Walker (Brasil) ; SATURNIIN^: Actias luna L., A. selene Hubner (India), Anthertea polyphemus Cramer, Attacus atlas L. (Borneo), Calosaturnia waltero-rum Hogue & Johnson, Dictyoploca japonica Moore (Japan), Hyalophora (H.) cecropia L., H. (H.) euryalus Bdv., H. (Callosamia) angulifera Walker, EI. (C.) Carolina Jones, //. (C) promethea Drury, Rothschildia orizaba West., Rhodinia fugax Butler (Japan), Saturnia(S.)pavonia L. (Germany), Samia cynthia L.; AGLIINAE: Aglia tau L. (Germany). The Brasilian larvae were preserved by F. Plaumann, those from Argentina by J. Forster, those from Germany by J. Reich el, and those from India, Japan, Borneo by C. L. Remington. The nomenclature largely follows Michener (1952).

104

Pease: First instar larvae

Vol.14: no.2

Notes:

1.   Descriptions are for one half of the body, i.e. reference to two setae would mean the larva actually .possessed four, with two arranged symmetrically on each side.

2.   Whenever the word "fused" is used, it applies to the chalazae of the setae, i.e. "Dl fused on dorsomeson" means that the chalazae of the first dorsal seta from each side of the body are united so that the two setae now arise from a common base.

3.   Segments are numbered posteriorly T-1, T-2, T-3 for the three thoracic segments and A-l to A-10 for the abdominal segments.

The Key

1.  Pro- and metathorax (T-1 and T-3) with dorsal chalazae much longer than those

of mesothorax (T-2) ; dorsal setas of T-2 short, bearing 4 branches with 2 of these longer than the others; chalaza of dorsal seta Dl of abdominal segment eight (A-8) fused on dorsomeson; no fusion on A-9; last abdominal segment (A-10) with a conspicuous posteriorly projecting chalaza on dorsomeson (Fig. 34) ; elongated chalazae have numerous thorny projections of cuticle usually bearing a secondary seta; some secondary setae on body, lateral plates of prolegs, and anal proleg........................................................................ Subfamily AGLIIN^ (Aglia tau)

Chalazae of T-1 and T-3 never prolonged without similar extension of chalazae of T-2; A-10 never with posteriorly projecting chalaza on dorsomeson; if cuticle of chalazae is erupted with thorny projections, secondary setae are usually absent from their tips; secondary setae, if present, usually confined to the lateral plates of the prolegs (short hairlike setae on the elongated chalazae in Syssphinx)...................... 2

2.   Scoli (usually XD, SD, L, D) on thoracic and abdominal segments bear 5-12 setae

of subequal length usually attached near the top (Fig. 5) ; dorsal scoli on A-9 never fused on dorsomeson; second dorsal seta (D2) of A-9 often absent; scoli subequal in height or at least none several times as long as others; anal prolegs with as many as 18 setae on lateral sclerotized plate; second lateral seta (L2) on prothorax (T-1) often not distinguishable from secondary setae of scolus LI; on A-l to A-8 second lateral seta (L2) is of primitive form and separated from secondary setae of scolus LI (Fig.9). (Subfamily SATURNIIN^E)..................... 6

Chalazae usually bearing only one or two primary setae at the top; some times 4-6 shorter processes around the setae; second dorsal seta (D2) always present on A-9, fused on dorsomeson and borne on a chalaza (Fig. 16) ; some chalazae may be much longer than homologous chalazae on different segments, particularly those on the thorax and A-8 and A-9; anal prolegs with three setae mesad, one anterior, and five on lateral sclerotized plate (Fig. 25) ; primitive lateral seta (L2) on T-1 present and projecting caudad from the base or near the base of LI; L2 may rise from chalaza of Ll on A-l to A-8..............................................................-......—............ 3

3.  T-2 with chalaza of dorsal seta (D) several times as long as other thoracic chala-

zae; no fusion on dorsomeson of A-8 (Fig.ll). (Some CITHERONIIN/E)............

................................................................................................................ Anisota, Dryocampa.

D on T-2 not the only prolonged thoracic chalaza; Dl fused on dorsomeson of A-8 ..

1960

Journal of the Lepidopterists' Society

105

4.   Chalazae of dorsal (D) and subdorsal (SD) setae of T-2 and T-3 and first dorsal

seta (Dl) of A-8 at least twice as long as chalazae on A-l to A-7, A-9 and A-10; chalazae may have thorny cuticular projections; some setae of suranal plate on A-10 usually borne on prominent chalazae (Figs. 28 to 31) ; primitive setae on posterior edge of cervical shield (Dl, D2) on T-l shorter than chalazae of D on T-2; no eversible glands on A-l and A-7.................................................................................... 5

Chalazae of thoracic setae subequal to those of abdomen; all chalazae without cuticular eruptions except at the top surrounding the primary setae; chalazae of setae on suranal plate inconspicuous; primitive setae on posterior edge of cervical shield (Dl, D2) usually as long as chalazae of D on T-2; eversible glands often visible behind spiracle on A-l and A-7 (Fig. 46) ........._________.................~~....................... 14

5.   Chalazae granulated without prominent cuticular thorns; many setae barbed and

flattened giving appearance of a double-^dged saw; subventral (SVl and SV2) and ventral (VI) setae of A-3 to A-6 without this modification; only one seta on the lateral plate on A-10 has this shape; double-hooked crochets on prolegs deeply embedded in spatula giving the illusion of two rows; lateral setae (Ll and L2)

rising from separate chalazae on A-l to A-8............................................___________......

................................................ Subfamily RHESCYNTIDIN^ {Rhescyntis meander)

Chalazae with cuticular thorns; setae relatively smooth; crochets shallowly embedded in spatula and obviously composed of a single row with occasional remnants (3-4) of an outer row; usually three or more setae of the suranal plate borne on one chalaza (Figs. 28 and 29) ; Ll and L2 with chalazae at least partly fused on A-l to A-8. (Subfamily CITHERONIIN^E).................................................................... 20

6.   Only primary setae (SVl, SV2, VI) present on prolegs of A-3 to A-6; secondary

setae occur on the anal prolegs; D-2 usually present on A-9 (Figs. 22, 35) ............ 7

All prolegs have secondary setae on the lateral sclerotized plates (making a total of 9-12 setae on prolegs of A-3 to A-6) ; D2 never present on A-9 (Figs. 24, 36) .... 10

7.   Chalazae of Dl on A-8 entirely separate; D2 a conspicuous seta on all abdominal

segments.................................................................................................................................... 8

Chalazae of Dl on A-8 partially or wholly fused on dorsomeson; D2 may be missing on A-9 ______....................................................................................................................... 9

8.   Height of SD and D scoli less than twice width; XD-SD of T-l present on an-

terior edge of cervical shield as a continuous band of setae .... Saturnia, Dictyoploca

Height of SD and D scoli greater than twice width; XD-SD of T-l are two separate groups of setae with prominent chalazae............................. Calosaturnia, Caligula

9.   Secondary setae of D and T-2, T-3 and abdominal segments recurved with tips

pointing caudad; width of L scolus on T-l less than half the width of the segment; XD and SD on T-l separate.........................................................Rhodinia fugax

Secondary setae of D and SD straight, pointing in various directions; width of L scolus on T-l more than half the width of the segment; XD and SD on T-l partly or entirely fused; Dl on A-8 fused on dorsomeson (Fig. 4) ..................Actias

10.  Secondary setae on prolegs of A-3 to A-6 arranged on both proximal and distal

parts of sclerotized plate (Fig. 37).......................................................___ Attacus atlas

Secondary setae on prolegs of A-3 to A-6 arranged on distal part of sclerotized plate only (Fig. 36)........................................._____.............................................„........... 11

106

Pease: First instar larvae

Vol.14: no.2

11.  Scoli XD, SD of T-l and Dl of A-8 fused at base only (Fig.3)............ Anthera>a

Scoli XD, SD of T-l separate; Dl on A-8 completely fused on dorsomeson........ 12

12.  SV group on T-l is a scolus with more than 3 setae; on head capsule a light-

colored area appears on each side of the epicranium; 10-12 crochets on A-3 to A-6 ----......................................-------................................................................. Rothschildia

SV group on T-l with at most 3 setae; no light colored area on epicranium of head capsule (may be light colored elsewhere) ; 13-19 crochets on A-3 to A-6............ 13

13. Body appears checkered due to dark areas between pairs of adjacent scoli (Fig.

9).................................................................................................................................. Samia

Body with transverse bands of color or unicolorous but not checkered .... Hyalophora

a.     Head and body unicolorous ..............................__............. Subgenus Hyalophora

b.     Head capsule bicolored; body with transverse dorsal stripes

___..................................................................................._______ Subgenus Callosamia

14.  Chalazas with tiny processes (4-6) at tip which are less than one-tenth the length

of primary setae on the chalazae (Fig. 42)..................................................................... 15

Chalazae with longer processes (4-6) at tip which are one-half to one fifth as long as the primary setae (Fig. 43) __...................................____...........—..........-............... 19

15.  Scoli on anterior edge of cervical shield (SD and XD) of T-l and subdorsal

scoli (SD) of T-2 and T-3 three-pronged at top with a seta on each prong............

............___...................-........................................................................................ Pseudohazis

SD, XD of T-l and SD of T-2 and T-3 two-pronged bearing two setae................ 16

16.  SV chalazas on A-7, A-8, and A-9 subequal in height ........___.................................... 17

SV chalazae on A-7, A-8, and A-9 not all subequal in height.................................... 18

17. Eversible glands behind spiracle on A-l and A-7 conspicuous ........ Dirphia curitiba

Eversible glands behind spiracle inconspicuous .........______ D. baroma, D. epiolina

18.  Four longitudinal series of light colored dashes running from T-2 to A-9 and

situated between rows of homologous groups of setae; the dorsal and subdorsal series more conspicuous than the others; SV seta on A-7 with large chalaza while SV chalazae on A-8 and A-9 are reduced or wanting.................................... Coloradia

Body without longitudinal light colored stripes; SV chalazae on A-7 and A-9 of subequal height; A-8 with SV chalaza reduced or wanting (Fig.10) .. Hemileuca

19.  Chalazae about one half the dorsoventral diameter of the body; fused chalazas with

only a slight fork below the insertion of the primary setae; eversible glands inconspicuous ___......____........__.....__......._____......________.............................. Hylesia

Chalazae equal to or greater in length than the dorsoventral diameter of the body; fused chalazae with conspicuous fork below the insertion of primary setae; eversible glands prominent behind spiracle on A-l and A-7............________ Automeris

1960

Journal of the Lepidopterists' Society

107

20.  Elongate chalazae not swollen at top; Dl on T-l and D2 on A-l to A-8 short,

thickened and bullet shaped; primitive lateral seta (L2) on T-l and A-8 usually projects caudad from base of L chalaza (Fig. 13) ; lateral scolus (L) on T-l with 1 to 3 secondary setae; SV chalaza on A-9 prominent (Fig.20) _________............. 21

Elongate chalazae swollen at apex; Dl on T-l and D2 on A-l to A-8 longer and bristle-shaped; L2 on T-l and A-8 usually at top of L chalaza (Fig.12) ; L on T-l with no secondary setae; SV on A-9 relatively inconspicuous (Figs.18 and 19) .................................................................................................................................................. 22

21. One or two secondary setae on prolegs of A-3 to A-6; some fusion on lateral plate

of proleg on A-10 (Fig. 31)....................................._______.......... Neocarnegia basirei

No secondary setae on prolegs of A-3 to A-6; chalazae separate on lateral plate of anal proleg (Fig. 30).............................................................................. Eacles tmperialis

22. XD at least as large as the most prominent structures on T-2 and T-3 and moved

lateral to a level between D and SD on T-2; SD on T-l less than one-tenth as large as XD and severely crowded................................._____ Adeloneivaia apicalis

XD not as large as the structures on T-2 and at nearly the same level as D on T-2; XD on T-l may be subequal or several times as long as SD ........................ 23

23.  Hairlike setae along the entire length of elongated chalazae (D, SD on T-2 and

T-3, Dl on A-8) ; primary setae on fused chalazae parallel to each other; XD and SD on cervical plate subequal and reduced; thorny processes on Dl, SD, and L of A-l to A-7 only at base of chalazae (Figs. 12, 18, 45)............................ Syssphinx

Hairlike setae lacking on all chalazae; primary setae on fused chalazae widely divergent forming acute or obtuse angle; XD several times as long as SD on T-l; thorns on Dl, SD, L of A-l to A-7 along the whole length of chalazae (Figs. 19, 44)............................................................................................................................. Citheronia

Recapitulation of Similarities and Differences

There are two clearly separated groups in the family Saturniidae. One includes the primarily Old World subfamily Saturniidae. The first instar larvae of this group have numerous secondary setae on wartlike processes (scoli). Each scolus with its secondary setae is homologous to one, two, or four primary setae. Secondary setse also occur on some or all of the prolegs. There are tendencies toward fusion of scoli in the middle of the dorsum of the eighth abdominal segment but never on the ninth abdominal segment where the second dorsal seta (not a scolus) tends to be lost.

The other group includes the New World subfamilies Citheroniinae, He-mileucinae, and Rhescyntidinae. Scoli and secondary setae rarely occur. Setae and the chalazae supporting them on the thoracic segments and abdominal segments eight and nine are often quite different in size and form from those on other segments. The first set of dorsal setae on the eighth abdominal segment and the second set of dorsal setas on the ninth abdominal segment are usually fused in the middle of the dorsum on both segments.

The Agliinae have numerous secondary setae, but many of the primary setas can be distinguished from the secondary setae. The first dorsal setae are

108

Pease: First instar larvas

Vol.14: no.2

fused on the eighth abdominal segment and may tend toward fusion on the ninth. A unique large chalaza or scolus with many secondary setae projects caudally from the middle of the tenth or last abdominal segment.

Brahmtea japonica Butler was examined for its relationship to the Sa-turniidae. It has numerous scoli and secondary setae resembling the Saturniinae. Some scoli on the meso- and metathorax and on abdominal segments eight and ten are greatly extended, whereas they tend to be subequal in the Saturniinae. There seems to be no compelling reason for including it in the Saturniidae although one might wonder if larval differences between the Old World and New World subfamilies are greater than the differences between the Saturniinae and Brahmaeidae.

In the Saturniinae a few trends are indicated. The subfamily is divided into: a) those genera which have secondary setae on all prolegs but lack second dorsal seta (D2) on the ninth abdominal segment, and b) those genera which have secondary setae only on the anal prolegs and may have a separate D2 seta on the ninth abdominal segment. In the first group fall Hyalophora, Samia, Roths childia, Anthertea, and Attacus. In the second are Satumia, Calosaturnia, Actias, Caligula, Rhodinia, and Dictyoploca. Caligula, Calosaturnia, Dictyoploca, and Satumia have the dorsal scoli on the eighth abdominal segment wholly separate while Anther-aa, Actias, and Rhodinia exhibit progressive fusion of these structures on the dorsomeson. Antherceci and Adtias also tend to have the two scoli on the prothorax (XD and SD) fused. Satumia and Dictyoploca have the secondary setae of XD-SD on the prothorax sprouting directly from the cervical plate. In Satumia these setae form a more or less ribbonlike band while in Dictyoploca they are separated into two clumps. In Calosaturnia and Caligula these prothoracic setae are borne on separate well developed chalazae.

In the New World group of subfamilies, the Hemileucinae are rather homogeneous. There is a tendency toward lengthening of processes on chalazae in relation to the primary setae — Hemileuca and Pseudohazis with almost none, Dirphia, Hylesia, and finally Automeris with well developed processes. If the genera are arranged by increasing length of processes on chalazae, then the order agrees with the order of branching in Michener's (1952) phylo-genetic tree. Some chalazae in Pseudohazis which should bear only two primary setae had an extra seta, making a three-pronged structure. This could be dismissed as an aberration except that the structures were symmetrical, and Packard (text figs. 10, 11; Figs.5 and 6 on plate 24; 1914) figures three pronged setae also. Although the eversible glands behind the spiracle on the first and seventh abdominal segments are considered ''family characters", there is variation in their prominence. In some species of Pseudohazis none were visible. They were conspicuous in Dirphia curitiha but inconspicuous in D. baroma and D. epiolina.

The single available representative of the Rhescyntidinae has specialized saw-like and finely granulated chalazae without the cuticular eruptions which characterize many Citheroniinae. The setae on the suranal plate were com-

1960

Journal of the Lepidopterists' Society

109

pletely separate as in the Hemileucinae; but there is great difference in length of some thoracic and abdominal chalazae, as in the Citheroniinae. Packard mentions another species, Rhescyntis richardsonii Druce, having enormous armature on the pro- and metathorax.

Within the Citheroniinae, Dryocampa and Anisota are quite unique. The armature in general is greatly reduced and the dorsal setae are separated on the eighth abdominal segment. There seem to be more and greater differences separating these genera from the rest of the Citheroniinae than there are separating the Citheroniinae from the Rhescyntidinae and the Hemi-leucinae. The chalazae on the dorsal half of the three thoracic segments are useful in characterizing genera. On each of these segments there are two pairs of fused and often enlarged chalazae one more dorsal than the other. Each of the six structures can be indicated as very large (VL), large (L), small (S), or very small (VS) in relation to the others of the same species. In summary crude formulae may be written which emphasize differences rather than similarities :

The order of branching in Michener/s phylogentic tree is, from bottom to top, Eacles and Citheronia, Syssphinx, Adeloneivaia, Dryocampa and Anisota, and Neocarnegia. Based on larval setae, Dryocampa and Anisota are separate from the rest, but whether they represent the loss of many specialized characters or a really primitive type is not certain. At the very least, the larvae suggest that these two genera belong either at the top or bottom of the Citheroniinae if not in their own subfamily. Eacles and Neocarnegia have important similarities in the manner of branching and the shape of setae. The two genera seem at least as close as Citheronia and Eacles. Syssphinx is distinct in having hairy chalazae. Adeloneivaia has unique widely spread prothoracic chalazae. The clue to the position of Adelonewaia and Syssphinx probably lies in other species of these or allied genera.

110

Pease: First instar larvae

Vol.14: no.2

There is a general trend toward fusion of setae on the last abdominal segment — Dryocampa and Anisota with none, to Adeloneivaia, Syssphinx, Citheronia, Eacles, and Neocarnegia.

A complete study of the first instar larvae of the Saturniidae may settle phylogenetic relationships precisely. It is usually no more difficult to collect first instar larvae than it is to collect adult female moths, since oviposition readily takes place in almost any container. Preservation in alcohol or fixative is somewhat simpler than preparing adults, and a complete collection of first instar larvae of all the Saturniidae would occupy the space of only three or four museum drawers. In fact, living material of species whose larvae have never been adequately described can be purchased from commercial breeders. The inescapable conclusion is that a taxonomic collection should and easily can include samples of first instar larvae.

Summary

1.     First instar larvae of the Lepidoptera have certain primary setae which are found throughout the order. Other setae which appear in later instars or only in more specialized forms are called secondary setae. The variation of the setae and their supporting structures is useful taxonomically and phylogenetically.

2.     There is a general tendency toward fusion of the supporting structures (chalazae) of certain groups of setae followed by fusion of whole groups particuarly on the dorsum of the thorax and eighth, ninth, and tenth abdominal segments. As a rule, only the supporting structures fuse; the primary setae remain separate although it is useful to say, for example, "the first dorsal setae are fused on the dorsomeson."

3.     The evolutionary significance of first instar larvae and the relative effects of selection on adults and larvae are discussed. A model is given whereby larvae might diverge faster than adults. Several examples are mentioned where larval recognition characters are better than those of adults.

4.     A series of aberrant Hemileuca is discussed in relation to the evolution of secondary setae.

5.     A key to the first instar larvae is presented.

6.     The Old World subfamily Saturniinae and the New World subfamilies Citheroniinae, Hemileucinae, Rhescyntidinae are clearly separated by larval differences. The Agliinae may be a specialized offshoot of the New World group.

7.     Within the Saturniinae, the genera Actias, Caligula, Calosaturnia, Dictyoploca, Rhodinia, and Saturnia are separated from the more specialized genera Attacus, Anther<za, Hyalophora, Rothschildia, and Samia.

8.     The Hemileucinae seem rather homogeneous.

1960                               Journal of the Lepidopterists' Society                                 111

9. Within the Citheroniinae, Citheronia, Eacles, and Neocarnegia are related while Adeloneivaia and Syssphinx are distinct. Dryocampa and Ani-sota are very similar. However, the differences which separate these two genera from the rest of the Citheroniinas seem greater than those which separate the Citheroniinae from the other New World subfamilies.

Acknowledgements

The author is particularly grateful to Dr. C. L. Remington for advice and encouragement while preparing the paper and to Dr. P. F. Bellinger for reading and criticizing the manuscript. Many of the specimens were supplied by F. Plaumann in Brasil, J. Reichel in Germany, and F. P. Sala in California.

References

Fracker, Stanley B., 1929. The classification of lepidopterous larvae, revised. Illinois

Biol. Monographs, vol.2, no.l: 161 pp., 10 pis. Gerasimov, A. M., 1935. Zur Frage der Homodynamie der Borsten von Schmetter-

lingsraupen. Zool. Anz. 112: 177-194, 8 figs. Hinton, H. E., 1946. On the homology and nomenclature of the setae of lepidopterous

larvae, with some notes on the phylogeny of Lepidoptera. Trans, Royal ent. Soc.

London 97: 1-37. Lawson, F. R., 1959. The natural enemies of the Hornworms on tobacco (Lepidoptera:

Sphingidae). Ann. ent. Soc. Amer. 52: 741-755. Michener, Charles D., 1952. The Saturniidae (Lepidoptera) of the Western Hemisphere. Bull. Amer. Mus. nat. Hist. 98: 341-400, figs.1-420. Mosher, Edna, 1914. A classification of the pupae of the Ceratocampidae and Hemi-

leucidae. Ann. ent. Soc. Amer. 7: 277-300. Packard, A. S., 1905. Monograph of the bombycine moths of North America, Part II.

Mem. nat. Acad. Sci. (Washington) 9: 1-151, 61 pis. __......., 1914. Monograph of the bombycine moths of North America, Part III. Mem.

nat. Acad. Sci. 12: pp.ix -f 1-276, 503-516, pls.1-113. Peterson, Alvah, 1948. Larva of insects. Part I. Lepidoptera and Hymenoptera. 315

pp., 70 pis. Scudder, S. H., 1889. Butterflies of the Eastern U. S. and Canada, vol. 3: pp. 1869-

1924, pis. 88, 89. Schiissler, H., 1933-1936. Lepidopterorum Catalogus, vol.10 (partes 55, 56, 58, 65, 70,

75, 76): Saturniidae 769 pp.; Syssphingidae 230 pp.

Dept. of Zoology, Yale University, New Haven, Conn., U. S. A.