KONINKLIJKE AKADEMIE VAN WETENSCHAPPEN

TE AMSTERDAM.

■\

•»'

The value of generic and specific characters
tested by the wingmarkings of Sphingides

BY

J. F. VAN BEMMELEN

4 ' /
P

i

Reprinted from: Proceedings Vol. XXI. N°. 8.

Si S'- 73

34i~ /

ci

5*

«x

r_

rn

V

*- 1

Zoology. — “The value of generic and specific characters, tested
by the wingmarkings of Sphmgides” . By Prof. J. F. van Bemmelen.

•

(Communicated in the meeting of October 26, 1918).

Supposing the rules for the colour-pattern of the wings, which
I deduced from former investigations by others as well as by
myself — to be valid, they ought to prove lit as guides in the
choice of a point of issue, when entering on the investigation of
a new group, that is to say when searching for a form which
shows the general pattern in its most original, least altered condition.

Judging by those rules, I believe that among Sphingides, as fai
as I am acquainted with - them, Smerinthus populi is a very original
form, in spite of the covering of red hairs, spread over the upper
side of the rootfield of the hind-wings, there hiding the primitive
pattern.

The arguments for this opinion are the far-going similarity of for e-
and hind-wing, both on the upper and the underside, and the pre-
sence of a pattern, which over the entire wing-surface is built after
the same simple motive, viz. regular alternation of darker and lighter
transversal lines and bands, each composed of spots. In both the
dark and the light bands the spots show a strong tendency to the
semilunar shape (the convex side turned outward), but here and
there they clearly approach the biconcave (hourglass) form. As to
the shades occurring as well in the dark as in the light bands of
spots, I pass them over for the present.

On the upper side of the fore- wing two of the darker lines run
on both sides of the light discoi'dal spot, and at a certain distance
from it, thus separating a darker median field from two lighter
transversal bands, which in their turn are again bordered by a
similar ribbonlike series of dark spots. Across this dark central area,
at the outside of the light discoi'dal mark, another dark bar may
be distinguished, and in the distal part of the area, between the
last mentioned bar and the outer borderline, there also occurs a
series of spots, which however are far fainter.

Moreover the anterior edge of the wing, on the inner side of the
discoidal spot, shows a lighter hue than the rest of the central area,
which increases in darkness toward the posterior margin. In this

2

992

lighter part two small dark spots touching the front-rim of the wing
may again be distinguished, each of them composed of two trans-
verse striae. In some specimens of Smerinthus populi the proximal
border-rim of the central area is also clearly double. Next to the
wing-root in the lightgrey hue of this part, a faint indication of a
dark ribbon may be detected besides.

Of the various hitherto mentioned bars of dark spots, the outer-
most, which is by far the strongest and completest, consisting of
nine separate elements when accurately counted, takes a sinuous
course. Counting from the front backwards, the fifth spot is situated
furthest inward, it also is the shortest and straightest. In many
specimens this spot is obviously darker in hue than the rest, and
this difference deserves our attention, as it is met again in allied
species, but here increased in intensity and extension.

On the inner and on the outer side of the just mentioned rows
of spots there occurs a broader bar of less obscure and more faintly
circumscribed markings, which however are evidently darker than
the grey shade of the lightest wing-areas, playing the r61e of ground-
colour. The outer of these two collateral bars is separated from the
median series by a narrow sharply traced light interval. From the
internal bar it differs by lesser regularity, some of its components
being broader than the rest, and at the same time darker. This is
especially the case with the spot near the hindborder of the wing,
this spot broadening obliquely in an outward and posterior direction,
and thereby just touching the hinder angle of the wing. A similar
triangular broadening also occurs at the front' end of the bar, near
the apex of the wing, but here it has a lighter hue. I think it
desirable to indicate these spots by special names, e. g. anterior and
posterior triangular spot, as they are found again with increased
clearness and independence in allied species

In the middle part of the bar under discussion four of the spots
clearly show the hourglass-type. In front of them the bar coalesces
with a dark area, extending along the greater part of the outer
margin of the fore-wing. This area forms a large convex blotch,
occupying five internervural cells from the apex backward.

Though at first sight this blotch is not divided into separate spots,
yet three darker centres may be distinguished in it, touching the
foreside of the nervures which take their course through it. A
comparison with other species of the same genus and of different
allied genera again proves, that these darker centres may be con-
sidered as originally independent separate spots — one in each
internervural cell— which have coalesced with each other into a single

993

3

almost homogeneous dark blotch. In front next to the apex, this blotch
is sharply cut off from the above mentioned anterior triangular spot
by an oblique light-hued line, the lastnamed spot moreover often
showing a very light shade itself.

Now directing our attention especially to the points of corre-
4 spondence between the various components of this wing-pattern, and

on the contrary less heeding the differences, we are easily led to
the conclusion that it is composed of seven transverse rows of dark
* spots, separated by lighter bars.

The external of these transverse rows (1) must then be looked
for in the above-mentioned dark blotch along the external margin.

The second (II) is the complete row of nine spots, with its set of
accompanying fainter bars.

The third (III) forms the external border of the dark central field.

The fourth (IV) is the dark line along the outer side of the
discoidal mark, which although somewhat obliquely, may be said
to run across the middle of the central area from fore- to hind-
margin.

The fifth (V) is the inner front-line of the central field, this line

* being sometimes double.

The sixth (VI) the single, curved series of spots over the middle
of the proximal light wing-area.

The seventh (VII) is formed by the faint traces of spots near the
wing-root.

The light intervals between these seven bars may be indicated as
in former publications by the letters A to G. In those intervals
some traces of still other dark bars, varying in distinctness, are
again met with ; so it is not improbable that originally the stronger
transverse bands everywhere alternated with less dark and sharply
marked rows of spots.

In all these features the pattern of popiili remarkably agrees with
that of Arctiids, as I described it in a former paper, and in the
same way with that of numerous other families of Lepidoptera, as

* I hope to show afterwards *).

x) Here l wish already to mention that Annette F. Braun, in a paper : Evolution
f of the Color Pattern in the Microlepidopterous genus Lithocolletis (Journal of the

Academy of Nat. Sc. of Philadelphia XV 2d Ser. 1914) as the result of an onto-
genetic and phylogenetic investigation about the colour-development on the forewings
of these Tineids, gives as her opinion that all patterns of this polymorphic genus
may be derived from seven dark transverse bars, forming to her mind a primary pattern,
on which a secondary one, composed of still darker lines, will later on be so to
say projected, as is also proved by the development inside the pupal sheath.

4

994

Moreover this pattern occurs almost completely on the hind-wing
as well as on the front one. For also there the first row forms a
convex dark blotch occupying five internervural cells along the
outer margin, with three darker centres next to corresponding
veins. In the same way the transverse bar II is composed of dark
spots, curved outward and is accompanied on both sides by a less
dark bar of more diluted spots. The number of components of bar II
is smaller than on the fore-wing, the hinder three being concealed
under the covering of red hairs. The fourth and fifth spot (counting
from the front border) are straighter than the rest, and placed
somewhat more inward, while they show rather a darker hue.

Row III and IY stand in contact with the corresponding ones of
the fore-wing, but disappear under the red covering even sooner
than II. V and VI are only indicated by dark spots along the front
margin, these spots moreover for the greater part being concealed
under the overlapping fore-wing.

VII is totally invisible.

On the under side of the wings the pattern perfectly corresponds
with that on the upper side, but on the fore-wing it is paler and
partly indistinct, on the hind-wing on the contrary it is sharper and
more complete than on the upper surface, because the red hairy
covering is absent on the former. The front-rim of the hind- wing,
remaining uncovered on this side, sharply contrasts both by colour
and pattern with the rest of the wing-surface and wears one espe-
cially dark spot, forming the initial component of Bar IV. It seems
desirable to indicate this peculiar spot by a special name, as it was
also done with those of the fore- wing : viz. the hind-wing-frontborder-
spot. In a single of the specimens at my disposal I also found the
markings along the front-border of the under side of (he fore- wing
differentiated and specially spotted.

On the under side of the hind-wing Bar V and VI are not repre-
sented, either by their initial (frontal) elements or by other spots of
their row, neither can Bar VII be distinguished; the light discoi'dal
marking however being well visible, as it strongly stands out against
the broad anterior part of Bar IV.

Now let us compare this pattern of the Poplar Hawkmoth with
that of the Eyed Hawkmoth. The close kinship of Sm. ocellata
with populi appears from several points of correspondence, but
surely most convincingly from the possibility of crossing these two
species together, the reciprocal hybridisation leading to different
results and the hybrids themselves being again fit for propagation.
On these accounts the classing of these two species into two different

995

5

genera, as proposed by recent systematists, in my opinion does not
give a true representation of these delations, but is only a conse-
quence of the immoderate tendency to splitting up, which nowadays
is so prevailing in systematic zoology.

To me the comparison between the wing-markings of these two
species, so different at first sight, seems highly interesting, especially
if the numerous byforms, which are described partly as independent
species, partly as subspecies, races, varieties, aberrations etc., are
also taken into account. Attention should also be given to the results
of hybridisation. But in the first place the under side of the wings
should be considered just as accurately as the upper one, and more-
over fore- and hind-wing, on both their surfaces, should be compared
to each other in detail.

If we do so with ocellata, it is easy to see that this species forms
one of the innumerable proofs for the assertion, that the difference
between fore- and hind-wing, upper and under side, is a conse-
quence of secondary modification of a general primitive pattern, this
pattern as a rule remaining better preserved on the under side than
on the opposed surface, though as to the latter, the fore-wing usually
has retained clearer and more complete vestiges of the primitive
pattern than the hind-wing.

Starting with the uppfer side of this latter, the conviction is easily
reached, that the eye-spot, in all its conspicuousness, is yet nothing
else but a peculiar modification of parts of three parallel dark bars,
each forming the termination of a transverse ribbon (parallel to the
wing-border), these ribbons again resulting from the coalescence of
a series of internervural spots1). Most convincing for this supposition
is the comparison of ocellata with the nearly related species Sm.
coecus and kindermanni , but it is already rendered highly probable
by the comparative inspection of upper and under side of the hind-
wing of ocellata itself. Such an inspection shows, that on the under
side the three ribbons in question are continuous without interrup-
tion from behind unto the front border, the outmost one causing
near to the hinder angle, where the wing-edge forms an incurvation,
a marginal obscuration, which can be retraced on the upper side in
the peculiar curved little stem, connecting the eye-spot with the
hind border.

On the under side therefore no indication of an eye-spot is present,
the dark bars running from before backward without interruption

x) This proof has been ably and convincingly delivered by Dr. J. Botke, in his
paper : Les motifs primitifs du dessein sur les ailes des Lepidopteres et leur signi-
fication phylogenetique. Tijdscbr. der Nederl. Dierk. Vereeniging XIV, 1916.

6

996

or modification, separated from each other by two narrow, very
light intervals.

Along the outer side as well as toward the wing-root these three
bars are accompanied by rows of dark spots. The row at the marginal
side represents the big semilunar blotch, which occurs on this same
place in other species of Smerinthus, and forms the homologue of
the corresponding patch at the outer margin of the fore-wing, which
occurs in ocellata as well as in numerous other species. We here
find the convincing proof that this dark marginal area is formed by
the coalescence of a row of spots. The middle row of the three

ribbons of dark spots is evidently double, its members forming a
series of square blocks, whose inner and outer side are formed by
dark strokes, sometimes straight, in other cases slightly curved. The
outer of these border-strokes are the darkest. The comparison with
populi proves that these dark strokes represent bar- II. As in populi ,
this bar is therefore accompanied at both sides by a dark seam.
The one on the outer side is much broader, darker and more inde-
pendent than the seam on the inner side. The latter is separated
from Row III by the inner white' band, this row only forming

a narrow line, connected by a dark interspace with the very dark
and complete Bar IV, which runs along the outer side of the light
discoi'dal marking, in the same way as in populi .

On the under side of the fore- wing the same spots and bars can
be found, with the exception of Bar IV, which remains entirely
concealed under the wine-red hairy covering of the root-field, just
like the posterior part of Bar III. It is only the discoi'dal marking,
which maintains itself as a small whitish patch in the middle of this
reddish covering.

But also on the upper surface of the hind-wing traces of these

same bars may be noticed, viz. along the front border, on that part

of the. wing that remains hidden under the fore-wing during flight,
but is protruded in front of it during rest, in consequence of the
peculiar attitude of fore- and hind-wing in regard to each other.
On this part three dark double-lines run backward up to near the
beginning of the red hue, and there end blindly. To my view there
is no reasonable ground for the supposition, that these vestiges of
pattern should have secondarily crossed over from the fore-wing to
the freely protruding part of the hind-wing. On the contrary it
seems justified to assume, that they belong to the primary pattern
of the hind-wing, as well as their homologues on the under side,
or those on the upper side of the hind-wing of populi , and have
remained untouched by the red discoloration.

997

7

Returning to the upper side of the fore-wing, it is at once clear
that the same pattern occurs on it as on the corresponding wing of
populi. It is only somewhat more differentiated: the dark middle
area is broken up into a fore- and a hind-part by a narrow funnel-
like slit of light colour, along the course of the second cubital vein,

* while the middle-member of the dark Ribbon II has increased in
bulk and shade to a very dark square.

In the same way the anterior and posterior triangular spot,

* especially the latter, are much more conspicuous and independent
than in populi.

By proceeding this way we can gradually arrive at the probable
conclusion that the patterns on upper and under side of both fore- and
hind-wing of ocellata repose on one and the same groundplan,- and
that this primitive pattern has suffered the strongest modification on
the upper side of the hind-wing, in consequence of its partial
overshadowing by a red discoloration and of the differentiation of
the back part of the pattern to an eye-spot.

Should further proof be needed, that the pattern of ocellata takes
its issue from the same ground plan as that of populi , this proof, as
already remarked, would be furnished by intermediate forms as
coecus and kindermanni.

As far as the markings on the upper side of the fore-wing, Sm.
coecus corresponds more to populi than to ocellata, the transverse
bars being more complete and more purely traced than in the latter.
Especially the dark middle area is not split up into a fore- and a
back-part, the Bars III, IV and V therefore all running straight
and unbroken from before backward, V in particular being sharp
and dark.

On the hind-wing the eye-spot is less purely circular, because the
external (hinder) dark line and the black pupilla-line are less rounded
and more advanced toward the hind-margin, thereby giving the
impression of fragments of ribbons.

On the other hand the vestiges of original design along the

* front-margin, hidden under the hind-rim of the fore-wing, are less
conspicuous than in ocellata.

On the fore-wing of kindermanni (Fig. 3) the median area is
broken up as in ocellata , and in general the similarity with the last
named species is greater, the design appearing only somewhat
sharper, especially Bar VII looking thereby more conspicuous. The
convex blotch along the outer margin is divided into a smaller
anterior and a bigger posterior part. The pink hue on the hind-wing
is particularly vivid in tone, but the eye-spot is perfectly flat and

8

998

composed of three almost similar pieces of dark ribbons, so that its
eye-character is almost gone *).

On the under side the similarity with ocellata is very striking.

/

By the consideration of the colour-pattern of ocellata , as well in
itself as in comparison to that of populi , we therefore come to the
following general statement:

In contrast to populi, the pattern on the upper side of ocellata ,
deviates from that on the under side, and is moreover composed of
two heterogeneous parts, a far-going difference existing between that
of fore- and hind-wing. Yet it proves possible to deduce the pattern
of both wings from the design of populi , which in this latter is
especially developed on the upper side, but which can be retraced,
be it in a fainter and more reduced condition, on the inferior surface.

The pattern of populi therefore satisfies the general conditions of
a primitive design, that of ocellata those of a secondarily modified,
viz. : dissimilarity between the upper side of the fore- and the hind-
wing, as well as between the upper and the under side of both
wing- pairs, in consequence of deviations of the upper side (of fore-
as well as of hind-wings) from the original, simple and regular
pattern, but this in a different sense for the two wings, the hind-wing
deviating more widely than the fore one. On the first-named a
tendency to annihilation of large parts of the pattern by the influence
of selfcolour prevails, combined with an extraordinary differentiation
of the remaining fragments, this leading to great contrasts between
the areas (eye-spot on pink ground). The fore-wing on the other hand
shows the complete original design, but transformed over all its
components in a more or less similar manner: some parts thereby
prevailing above the rest, without affecting however the general
harmonious character of the whole.

These facts might easily lead to the conclusion, that the peculiarities
in the design of the upper side, by which ocellata differs from

1) In parenthesis it may be observed, that we therefore are able to prove for
the eye-spot on the wings of the imago a similar origin as A. Weissmann so
ably did for those on the body-rings of the larvae of Sphingides, viz. that they
spring from fragments of a set of longitudinal, alternately light and dark stripes,
these fragments becoming independent and differentiated to concentric circles, wdiile
the rest of the stripes disappears totally or nearly so. In caterpillars these stripes run
parallel to the body-axis, on the wings of the imagines they are seemingly trans-
verse. Considering wings to be folds of the body-skin, it is easy to conceive, that
these so called transverse stripes in reality correspond to longitudinal stripes on
the insects’ body. Probably the latter stripes may as well as those on the wings
be considered as rows of spots which have coalesced together.

J. F. VAN

Proceedings F

J. F. VAN BEMMELEN: "Wing-markings of Sphingides”

Proceedings Royal Acad. Amsterdam. Vol. XXI.

E. THEYSSEN, Phot

POPULI 4. TILIAE

999

9

populi, should be considered as specific features of the first-named
form, whereas the points in which the two species resemble each other,
especially on the under side, would possess the significance of generic
characters. This opinion, that e.g. the eye-spot of ocellata cannot
pretend to a higher significance than that of being a specific peculi-
arity, might not only find support in its restriction to the upper
side, but also in the results of hybridisation, showing that the hybrids
from the cross between a male ocellata and a female populi are
ocellata- like in their habitus, though with a faint eye-spot, deprived
of its pupilla, whereas those from the combination of a male populi
with a female ocellata possess such a far-going similarity to populi ,
that they can hardly be distinguished from it, the eye-spot being
wholly absent.

This view about the meaning of the differential features of ocellata
seems the more attractive because a special importance for the chances
of survival of the animal may be ascribed both to the variegated
and marmorated design of the fore-wings and to the eye-spots, in
their monochrome pink field, of the hind ones. As long as the moth
in ifs attitude of rest is suspended on a willow-twig among the
leaves of that food-plant of its caterpillar, the hind- wings are con-
cealed under the fore-wings with the exception of their narrow
fore-rim, and the animal so delusively imitates by its form, colour,
design and proportions of a pair of dry willow leaves, that notwith-
standing its bulky size it can hardly be detected amongst its natural
surroundings, as long as it remains motionless.

When however the sleeping moth gets disturbed by pushing or
hurting, it moves its fore- wings a little forward, thereby suddenly
displaying the eye-spots in their red surroundings, which by their
situation on both sides of the somewhat upheaved abdomen (this
part of the body at the same time making periodical jerks) cause
the illusion of a savage face with wide-opened eyes, thereby (as
experiments have proved) so effectually frightening birds and reptiles,
that they generally abstain from further attempts to devour the moth.

When 1 mention these long known facts from the chapter of
Protective Mimicry, it is because I think it desirable to state once
more, that they can never be used as an explanation of the presence
of markings, hues and shapes, which by their cooperation call forth
the deceitful resemblance. These features owe their presence to causes
of quite another order of things, viz. to the variability, which itself
is a consequence of the coincidence of hereditary factors. When this
coincidence accidentally leads to an effect which in a certain direction
is favourable for chances of survival of the animal (or plant), it

10

1000

will always be preserved and ameliorated by the influence of natural
selection, so that it will give rise to those highly finished cases of
mimicry and protective resemblance which so often raise our aston-
ishment and admiration.

One of the reasons that have made it seem desirable to repeat
once more these opinions, however often proclaimed before, is the
fact that de Meyere in his recent paper “Zur Evolution der Zeich-
nung bei den Holometabolen Insekten”, when arguing on page 59
against Botke’s views about the wing-pattern of Cossids, declares
that he can only see in their design “eine hochgradige Entwicklung
einer sympatischen borkenahnlichen Farbung”, while on page 48 of
his preceding article “Zur Zeichnung des Insekten-, im besonderen
des Dipteren- und Lepidopterenfliigels”, he derives this design “aus
einer Zeuzera-pirina-ahnliehen Fleckenzeichnung”. To this he adds:
“Dieser Weg scheint mir besser verstandlich als der umgekehrte”.
And somewhat further on he says about the transverse markings on
the under side of several Vanessidae: “diese scheinen mir mit dem
primaren Zeichnungsmuster iiberhaupt nichts zu tun zu haben,
sondern es sind eher spat erworbene Elemente der sympathischen
Farbung”.

The point, in these considerations of de Meyere which I want to
discuss, is not his opposition against Botke’s views about the con-
nection between spots, stripes and nets, to which 1 cannot pay full
adhesion either, but his assertion, that by considering a wing-design
as a “sympathetic pattern” an argument is raised for the explana-
tion of the origin and the discovery of the age. of this pattern.
Patterns of all kinds, the most original as well as the most strongly
modified, may produce a mimicking effect, and thereby prove useful
for protective purposes.

E.g. the wing-design of populi has quite as much protective value
as that of ocellata , though only in the sense of resemblance to a
weathered poplar-leaf, and yet it is much more primitive than the
latter. Moreover the same motives and elements of design, which in
one species of animal are the source of highly imitative mimicry,
may also be found in other species, near akin as well as far removed
in a generic sense, but here, b}7 showing a somewhat different form
or by occurring in another part of the body, only cause a feeble
sympathetic resemblance or no mimicking effect at all. Of this so
called false mimicry Eimrr has cited several instances.

Numerous thin, irregular transverse stripes between the veins, in
the sense of Botke’s “traits effiloches”, are found except in Vanessidae*
in many other Lepidoptera of diverse families : also in Sphingides,

1001

11

on both wing-sides, though most frequently on the under side. In a
general survey of the wing-markings of Lepidoplera, which I under-
took long before Botke’s doctor-dissertation, and wrote in English,
but which I did not hitherto publish, I even thought it desirable
to choose a special name for this curious motive of design and

i called it “Cosside markings”.

Now it might very well be, that these markings could also be
reduced to an old and original motive of design, occurring generally

* among insects, and whose connection with the system of internervu-
ral spots still wants elucidating, although Botke has made a notable
attempt to come to a general theory.

That a “sympathetic” design, on account of its mimetic character,
should necessarily be younger than other patterns, I deny most
emphatically. Each of the elements, which by their cooperation
produce the mimetic effect, may in itself depend on hereditary
tendencies of very high phylogenetic antiquity. Only the specific
and special culmination of that cooperation may be young, and
even this need not necessarily be the case. Among Pierids, Papilio-
nids and Nymphalids the mimicrists probably often wear an older

* and more primitive uniform than the remaining so-called typical
members of these families, as I have tried to demonstrate in my
paper read at the International Entomological Congress at Oxford

r in 1912.

In numerous Geometrids, issuing from their pupae in autumn,
the similarity to a weathered leaf reposes on their light-yellow
colour, besides on the broken rim of their wings and the course
and arrangement of dark transverse lines on them, imitating the
veins of the leaf. Must on this ground the yellow hue be younger
than other tints? According to my view this need no more be the
case than it need be assumed for the form of the wing-border or
the pattern on its surface, even when granting that in general a
broken border-line has to be derived from an unbroken, round-
ed one.

( In the same way the evident connection between spots, stripes

and meshes on the wings of Cossids, which can so to say be read on
the wing-surface by simple observation and by comparison with the
Zeuzerids, is in no way brought nearer to an explanation by the
remark, that the preponderance of the net-markings produces a sym-
pathetic resemblance to the bark of trees. The real question remains :
what causes tendency of the Cossid wing-markings to the net-character
and how old is that tendency ? In putting this question we have to
keep in view, that the same tendency occurs in many other insects

12

1002

belonging to different orders and that it may likewise be remarked
in the nervural system, which possesses such intimate and primor-
dial relations to the distribution of pigments in the skin.

Were it only to consider this question from all possible points of
view, the well-foundedness of ’the hypothesis would have to be tried,
whether net-design may be connected with the formation of meshes
in the system of wing-nervures, as is so frequently and specifically
found in Neuroptera and Orthoptera, vicariating with regular trans-
verse venation ; whether therefore the net-design may not be as old
as or even older than spots or stripes. An argument to this assertion
might be found in I he fact, that nets between the longitudinal veins
are characteristic of the nervation of the wings in Palaeozoic
Palaeodictyoptera.

With this inference I do not in the least intend to proclaim, that
I am convinced of this connection of the net-design and of its phy-
logenetic antiquity, but simply that I think the contrary is not
proved either.

Remembering Weismann’s words: “Ohne Hypothese und Theorie
giebt es keine Naturforschung”, I am of opinion that 'the continual
proposing of explaining suppositions about the connection between
corresponding phenomena is necessary condition for fertile scientific
research, and therefore I cannot adhere to de Meyere’s point of
view, where he says: “Ich mochte mich, den Tatsachen entsprechend,
mit Feststellung des Auftretens begniigen und keine ganz hypothe-
tische Verbindungslinien ziehen.”

Returning to the specific differential characters of Sin. ocellata,
I here find the danger to get entangled in purely hypothetical spe-
culations not by any means serious. For it can be easily proved
that all the special characteristics of the upper side of both fore-
and hind-wings occur as well in other species, not only of the genus
Smerinthus, but also of different allied genera.

In the first place the comparison with tiliae is highly instructive.
On the upper side of the fore-wing of the Limetree-Hawkmoth every
peculiarity by which the pattern of ocellata deviates from that of
populi, is again met with, but in a modified form and in other
hues, which together produce a totally different effect of the wing-
design as a whole.

Especially striking is the similarity of the dark median area with
the same wing-part of ocellata ; as in the latter it is cut up into a
fore- and a hind-quarter by funnel-shaped intrudings of the light-
brown ground colour, which may either meet each other or remain
separate. A single look at every somewhat considerable collection

1003

13

of tiliae shows the extreme individual variability of this feature, as
well as of others. The transition of an unbroken middle-field to one
divided into an anterior and a posterior portion we therefore here
see take place under our eyes.

We may likewise notice, that the peculiarity of the central (5th)
mark of the dark Bandline II, to differ in hue and size from the
other members of that series, is also present in tiliae , but in so far
in an opposed sense, that in some specimens it is distinguished by
, a lighter instead of a darker shade. On its underside, tiliae shows

again the same simplified pattern as populi and ocellata , viz. the
two Bandlines II and III, with traces of I, IV and VI.

The right here to speak of simplification, and to connect this
with the covering of light hairs stretching from the wing-root out-
ward as far as the middle area, is strikingly proved in this case by the
vestiges of the opaque central blotch (so strongly developed on the
upper side) which can also be detected on the under side. At the
root of each hair in the area of this dark middle-field a small black
speck may be perceived, and this produces the effect that the field
is seen in its full extension as a collection of specks, when we look
^ obliquely between the hairs.

Still more striking than the resemblance between tiliae and ocellata
is that between both these species and tartarinovii, this latter offering
r so to say a form of transition between the first-named two. Here

the anterior part of the external margin of the opaque middle area
is not convex as in tiliae , but is concave, while a contrast both in
hue and in markings exists between the anal field of the wing and
the rest of its surface, the division of the central field in a fore-
and a hind-part thereby appearing as part of a process which extends
over the whole length of the wing' surface, in the same way as in
so many other Lepidoptera and even in Insects of other orders.

It is likewise remarkable, that the apex of the fore-wing, which
shows a special differentiation identical for all these species, viz.
that it is separated from the remaining markings by the oblique
light stria already described for populi , is dark greenish grey
instead of silvery grey, in contrast with the convex blotch along
the outer margin, which is stained in light grey, while it is dark
in others.

On the hind-wing tartarinovii displays the same pink as ocellata ,
and even traces of an eye-spot.

On the other hand in tiliae a dark band extends over the entire
surface of the hind-wing parallel to the outer margin and at some
distance from it. This band evidently consists of as many components

14

1004

as there are internervural cells. Each component is prolonged wedge-
like in the direction of the wing-root. The extent of this prolongation
is individually different, though in general it may be stated, that
in a backward direction toward the hinder wing-edge the dark
internervural spots get larger and more intensively black, this back-
ward increase in size and darkness being the only indication of a
similarity to the eye-spot of ocellata.

Therefore, though in this latter instance tiliae shows almost as
little likeness to ocellata as does populi, this does not derogate from
the truth and the value of the fact, that in numerous Sphingides the
hinder external angle of fore- as well as of hind-wing shows a dark
pigmentation, which may be differentiated to an eye-spot. This might
also be expressed otherwise, by saying that the above-mentioned
posterior triangular spot is not restricted to the fore-wing, but returns
on the hind-wing.

And also on the fore-wing the spot in question may assume the
character of an eye-spot, as is shown by several Sphingides belonging
to different genera, e.g. Daphnusa ailantha.

On the other hand, as already said, other species possess near to
this hinder external angle of the fore-wing only a single or double,
solid blotch not differentiated to an eye-spot, e.g. Oxyambulyx
canescens. Also this blotch may be repeated on the hind-wing, e.g.
Smerinthus quercus.

Among the Sphingides at my disposal Pholas labrusiae (Fig. 6)
seems to me to possess a highly remarkable colour-pattern. On its
under side the similarity between fore- and hind-wing is very great,
and both show the usual simple design of Lines II and III on a
nearly homogeneous faint greenish-yellow ground, to which only along
the external margin a differently coloured area, separated from the
rest of the wing-field by a zigzag line, and evidently representing
Bar I, is added. But on the upper surface the contrast between the two
wing-pairs is very profound. The fore- wings are almost unicolourous
dark opaque green, exactly corresponding in hue to the entire body
of the moth. Yet several traces of dark transverse lines are well
defined, especially the middle-field between Bars III and IV, which
is conspicuous by a somewhat darker green shade. But at the external
border of this middle-field two square little areas are so to say
spared out from the general green overshading: one nearly in the
centre, the other at the back margin, thereby giving the impression
of brown curtains before two low windows in a green wall. That
impression is strengthened by the fact, that in these brown areas
the curved components of the transverse dark bars are more numerous

1005

15

and far sharper than in the green held. In the central window 4
of these arched stripes are present, in the back one 2.

Comparison with Smerinthus popidi and ocellata as well as so
many other Sphingides, whose ground-colour is brown or grey, and
whose transverse bands are composed of curved, stripy spots, leads
to the supposition that in these windows we have to see remnants
of the original hue and design of the wing, which for the rest has
become indistinct by green discoloration.

As in so many other cases, e. g. the Hepialids, green therefore
would be the secondary, brown the primary hue, the design having
partly get lost in the process of discoloration or at least having
greatly diminished in distinctness. But why these two brown windows
with their trelliswork of curved stripy spots have remained untou-
ched by this process, I cannot as yet explain.

Contrasting with the almost homogeneous green hue of the fore-wing,
the hind one possesses a very showy and variegated pattern : two
jetblack bars standing out against a light yellow ground, bluish-
grey areas occurring at the front border between the black, a
brickred patch vicariating with two black strokes near the inner
margin, while at the outer one a small green field breaks the yellow.

But the most remarkable point in this pattern are two darkbrown,
irregular, denticulate lines, starting at the hinder angle, and running
parallel to the outer margin along its posterior part, to pass into
the broad black bar at the hinder border of the small green field.
These crooked lines represent the posterior part and the pupilla of
the eye-spot in the ocellata-g roup among Smerinthidae, and form
the least-modified part of the hind-wing-pattern of Pholus labrusiae.

I think it highly probable, that this pattern has a protective signi-
ficance for the animal, just as well as the almost homogeneous green
hue of the fore- wings and of the body. The latter give protection
to the sleeping animal by making it hardly visible to enemies that
prey upon it, possibly the brown windows play their part in this
process of concealment, by breaking the anatomical lines of the
rather extensive wingfield.

It certainly would be worth while to make the experiment,
whether the moth when disturbed in its sleep, suddenly displays
its hind-wings and so frightens its enemies away, or whether the
showy colour-composition, which thereby gets visible, has only the
meaning of a warning-pattern, announcing unpalatableness.

Whatever may be the right interpretation, this pattern in any
case ought to be considered as a high and special differentiation of
the original one of the hind-wings, common to all Sphingides ; neither

16

1006

the crying colours nor their queer arrangement justifying the infe-
rence that they could ever be the direct consequences of the useful
effect they produce in favour of the animal.

The comparison of Sm. ocellata with populi, and of both with
other Sphingides, leads me to the following conclusions:

The colour-pattern of populi is more primitive than that of ocel-
lata, it agrees with the conditions which may be posed for a pri-
mordial pattern, and it corresponds to the fundamental plan, as this
is found in Arctiids, and most probably in numerous other families
of Heterocera, possibly also in Rhopalocera. It therefore is not only
older than the genus Smerinthus, but even than the family of Sphin-
gides, perhaps than the entire order of Lepidoptera. So it cannot
without great restriction be qualified as a generic pattern.

The colour-design on the upper side of ocellata can be derived
from that of populi by the assumption, that the ribbons of interner-
vural spots occurring in the latter have been specially transformed
in the former. But each of these transformations in itself is seen as
well in other species of Smerinthus, and even in many other genera
of Sphingides, it is therefore not allowable to assume, that they
should have been acquired during the formation of ocellata from a
populi- like ancestor. Each for itself they are not characteristic of
ocellata, and cannot be taken as specific features of this species. It
is only the peculiar combination of the modifications of the ancestral v

type with the subtle nuances by which in ocellata they are distin-
guished from the similar modifications in allied forms, that in the
end give the specific character to ocellata. At any rate the origin of
the said modifications of the primitive pattern cannot be ascribed to
the influence of protection against enemies, which ocellata obtains
by the use she (instinctively) makes of her eye-spots. The special
refinement however and the elaborate details, by which the pattern
of ocellata surpasses that of other Sphingides near akin, may well
be the consequence of natural selection, which could enter into
action as soon as by coincidence of hereditary variations of the
fundamental Sphingidial pattern with special circumstances of life,
a deceitful likeness had been established to the face of a big-eyed
owl, which frightened away preying little birds and small mammals.

Groningen, October 1918.