V. Nabokov.
Notes on the Morphology of the Genus Lycaeides (Lycaenidae, Lepidoptera).
Psyche 51(3-4):104-138, 1944.

This article at Hindawi Publishing: https://doi.org/10.1155/1944/24365
CEC's scan of this article: http://psyche.entclub.org/pdf/51/51-104.pdf, 3388K
This landing page: http://psyche.entclub.org/51/51-104.html

The following unprocessed text is extracted automatically from the PDF file, and is likely to be both incomplete and full of errors. Please consult the PDF file for the complete article.

Psyche [Sept.-Dec.
NOTES ON THE MORPHOLOGY OF THE
GENUS LYCBIDES
(LYCLENIDB, LEPIDOPTERA)
Out of the hundred or so holarctic Lycsenids distributed among at least sixteen genera of the subfamily Plebejince (defi- nitely fixed by Stempffer, 193 7, Bull. soc. ent. France 42 : 2 11, etc.; not covering the superficial concept of "Blues" for which no systematic term or division can exist), only fourteen species or so, two of which are obvious invaders from the Tropics, occur in the nearctic region (north of the 30th parallel). These be- long to seven genera, four of which (the first four in the list given below) are holarctic and contain together six species of which one half is common to both regions. All three exclusively American genera have the free portion of the sedeagus elon- gated; all the exclusively palearctic genera, except Aricia R. L. (and the, mainly tropical, Chilades Moore and Freyeria Cour- voisier) have stubby or proximally "bulbous" free portions. Of the four genera common to both regions one half belongs to the first type and one half to the second.
The only Plebejince, so far known to exist in the nearctic re- gion, are: 1. Agriades Hubner : glandon Prunner (holarctic) ; 2. Vacciniina Tutt: optilete Knoch (holarctic); 3. Lycceides Hubner : argyrognomon Bergstrasser (holarctic) , scudderi Ed- wards, melissa Edwards; 4. Plebejus Kluk: scepiolus Boisdu- val; 5. Plebulina, n.g. (remarkably amalgamating the Plebejus or Lycceides sedeagus with the valval processus superior and uncus + falces of Albulina Tutt): emigdionis Grinnell (geno- type) ; 6. Icaricia, n.g. (allied to Aricia R.L. in sedeagus; some- what to Polyommatus Latreille in general type of uncus as seen ventrally; close to both in processus superior of valve; dis- tinguishable by the underdeveloped, i.e. devoid distally of any Published with the aid of a grant from the Museum of Comparative Zoiilogy at Harvard College.
Unexpectedly represented by speciosa Staudinger in the Andes.



================================================================================

19441 Genus Lycceides 105
semblance of hook, triangular, laminate, proximally very broad falx, its very gradually tapering apex hardly exceeding in height the level of its strongly humped humerulus) : icarioides Boisdu- val (genotype) with its various subspecies (clamoring for a reviser) and four other species, viz.: demon Doubleday-Hewit- son, spt indeterm. (? chlorine, Skinner), neurona Skinner and shastu Edwards; these four structurally smaller than the geno- type (with an uncus lobe distally somewhat grooved in lateral view but not actually revealing Sternpffer's process as it occurs in Arich anteros Freyer), and 7. Hem.iargus Hiibner: a curi- ously aberrant genus (somewhat allied to Chilades Moore) which is represented by hanno Stoll and in which I very provi- sionally retain isoh Reakirt. An unexpected fdtura superior is present in the former and is monstrously developed in the latter.
For some time I have been especially concerned with the genus Lycosides. In a preliminary paper (Nabokov, 1943 [March, 19441, Psyche 50 :87 etc.) an attempt was made to clear up several taxonomic points mainly in regard to the ne- arctic section; the palearctic one is still badly confused taxo- nomically, especially because the type specimens of a number of races have never been examined structurally (German au- thors, for instance, blindly relying upon the haphazard commer- cial identifications of the Staudinger firm). These matters I shall discuss elsewhere, but it is necessary to make a" few com- ments regarding the genotype.
This is the "argus Linn." of Hiibner ( [ 182 3 1, Verz. bekannt. Schmett. 5 :69), nec Linn., which was selected as the type by Scudder (1872, 4th Ann. Rep. Peabody Acad. Sci. 1871:54; 1875, Roc. Amer. Acad. Arts Sci., Boston 10:208), and since Hiibner's argus is the "Argus" of Reverdin (1917, in Oberthur, Et. lbp. comp. 14: 2 2, fig. 3, uncus) it follows that it is also the "argyrognomon Bergstrasser" of Tutt [and Chapman] (1909, Brit. Butt. 3:205-208, pi. 50, fig. 2, uncus) and thus not the "Ligurka?' of Reverdin (191 7, op. cit.: 22, fig. 4, uncus) which is the ''ismenias Meigen" of Heydemann (1931, Int. ent. Zft. 25: 129) and the^wgyrognomon Bergstrasser" of Forster (1938, 'With an incidental suggestion (LC. : 88, not4 that cleobis Bremer falls to mbsohnus Eversmann. I now find that Hemming (1938, Froc. R. Ent. Soc. London, 7 (I), B : 5-7, fig., mate, type) had already come to the same conclusion.



================================================================================

106 Psyche [Sept-Dec.
Mitt. Munchner ent. ges. 28: 1 1 ) , wrongly, and belatedly, se- lected by the latter author as "type" with thfe suggestion that readers look up for themselves Hiibner's plate. They do, and find (Hiibner, Samml. europ. Schmett. pi. 64 [1800]) that fig. 316, to which Scudder referred when selecting the type, can be easily matched by German maim of the "Argus" of Reverdin and of the "argyrognomm Bergstr." of Tutt and, consequently, of Hemming (1934, Gen. names hoi. butt. 1: 1081, who defi- nitely fixed it (thus excluding the other species of Lycazides which he knew well) as the type of the genus, and this clinches the matter, whatever the two species be called. The publication of Beuret's important paper ( 1935, Lambillonea 35: 162, etc.) has led to attempts to transfer the name argyrognomon Berg- strasser (1 779, Nomenclatur, 2: 76-77, pi. 46, fig. 1,2) from the short-falx species (the genotype) to which it was applied by Tutt (1909) and which we shall term for the moment species X, to the long-falx species, ismenks Meigen, 1830 (Heyde- mann? 193 1) which we shall term species Y. These attempts have been prompted by the fact that female specimens appar- ently belonging to Y (Beuret, LC., does not give the reasons for his determination), casually collected in the type locality of argyrognomon Bergstr,, proved to be closer to Bergstrasser's equivocal figures than sympatric females of X. One cannot deny that the figures apply better to the general run of Y fe- males than to the general run of X females; but pending further investigation, or some formal decision on the part'of a special commission, I am compelled to use in this paper the name weyrognomun Bergstr, for X because of the following consid- erations: 1. As noted and illustrated by Beuret himself (1934, Lambillonea 34: 119) at a time when he still called X by the name argyrognomw, absolute similarity to Bergstrkser's fig- ures is exhibited by what he (inconsequently) named argyro- gnomon rauraca Beuret (I.c. pi. 5, 5a, fig. 9, 10. See also Beuret, 1928, Soc. Ent. 43, fig. 5, 10, uncus, argyrognomun, 'Augst") . This, now extinct, colony was discovered on a plot of ground, a thousand feet long and 1/6 of this broad, near Augst in the Aargau, N. Switzerland, Le., some 200 miles south from the type locality (BruchkGbel Forest, in the Hesse-Nassau dis- trict, Central Germany) of argyrogwmon Bergstr.; but mor- 'Whose clumsy fixation I reluctantly adopt.



================================================================================

19441 Genus Lyceeides 107
phologicaUy, i.e., apart from current geographic obsessions and notwithstanding the inconvenience of the thing not flying where it ought to fly, rawaca Beuret was when discovered, and in my opinion remains so BOW, an absolute synonym of orgyro- gnomon argyrogmon Bergstr., since in genitalia it corresponds to Tutt's argyrognomon Bergstr, and in the appearance of the female to Bergstrbser's figures; 2. There is no guarantee that the next German, or British, collector in the Hesse-Nassau dis- trict will not come across chance specimens or a little colony of X, different from the race of X (lyctdasOtdes Beuret, 1934), assigned to the general region, and similar to Beuret's Aargau series-in which case the whole question would have to be brought up again (Tutt remaining the first reviser 5, ; and 3. It is not at all clear what name should be used for X if "argyro- gnomon" is switched to Y. The name acreon Fabricius (1787 Mantissa 2: 76), on the basis of a worn specimen of argus auct (which combined at least X and Y) in the Banksian collection was assigned to the latter omnibus species by Butler (1869, Cat. Diurn. Lep., descr. by Fabricius, in coll. B.M.: 17 1 ) which leaves us none the wiser, even if Butler did see "the type female in Copenhagen" as stated by Heydemann (1931, Int. ent. Zft. 25: 150) who anyway had not seen it himself and thus was per- f &tly unjustified in using the name (I.c. pi. 1, fig. 4, 12) for a race of X. The name' caltiofis Boisduval ( [ 18321 1c. hist. It$. Europe 1 : 58, fig. 4,s) suggested by Hemming (1938, Proc. R. Ent. Soc. London 7,B:4) also cannot be used for X, until the female type (from Grenoble, France) and the Uriage male as- signed to calliopsis by Oberthur (1896, Et. ent. 20, pi. 5, fig. 64) are critically investigated in the B.M. collection. In view of the fantastic misadventures which names have undergone in this genus, pedantic care must be taken, so as to avoid some new nomenclatorial trouble in the future.
The genus Lyccsides, of which argymgnomon Bergstr.-Tutt is the type, is characterized by an uncus (including the fakes) exceedingly different from the corresponding structure found in other subdivisions of the Ptebem, and as I think it advisable to base specific unities upon the intrageneric variation of that character which intergenerically is responsible for the greatest uIn the sense that by figuring the male genitalia he first applied the name argyrognomon Bergstr. (which previously to 1909 had covered at least two Lyeceides species and a form of Plebe* argus Linn.) to a definite species.



================================================================================

108 Psyche [Sept.-Dec.
hiatus, it is the uncus that I have selected (partly in develop- ment of Reverdin's, Chapman's, and Sterppffer's views) for differentiating species in the Lycosides. The male armature consists of a dorsal (in regard to the body) portion (the uncus) and of a ventral one (the valves - which have a constant fishlike shape in the Plebejiwe,'), The two are hinged to each other somewhat in the way of the lids of a shell and appear "closed" when viewed in situ. When teased out of the tissues and viewed ventrally, i-e., when the whole organ is forced open oysterwise so that its symmetrically ex- tended valves continue to point down, whereas the uncus lobes point distad from the observer, the most conspicuous thing about the upper portion is the presence of a pair of formidable semi-translucent hooks (the subunci or fakes -of a peculiar shape not found in allied genera), produced from the opposite side of the distally twinned uncus and facing each other in the manner of the stolidly raised fists of two pugilists (of the old school) with the uncus hoods lending a Ku-Klux Klan touch to the picture. The flame-shaped distal part of the candle-shaped aedeagus reaches a point between their elbows, while its proxi- mal part is propped by the fultura in/erior (furca) at the root of the valves.
In the paper already referred to, I introduced the following terms: F. for the length of the upright portion, or forearm, of the falx measured from its distal point to the apex of its elbow; H. for the length of the herdus of the falx, from the apex of its elbow to the apex of its shoulder; and U. for the length of the uncus lobe from its distal point to the apex of the shoulder of the falx. In the majority of some 500 preparations, regard- less of whether the elbow of the falx happened to be raised (in the follow-through of an "uppercut," to pursue the pugilistic image) as it is for instance in fig, ARGA. of pi. 1, or whether it remained in its normal position (i.e. with the forearm parallel to the axis of the uncus lobe), a rather curious fact was noticed, namely that the distance between the tip of the falx and the apex of the shoulder exactly equalled V. This suggested the tracing of a triangle, FHU, its lines joining three points: apex of forearm, apex of elbow, and apex of shoulder. A glance at fig. 1 will show that, according to the dimensions of forearm, humerulus and uncus lobe, this triangle assumes a different size (showing the gradual generic development) and a different shape (showing the specific relative dimensions of parts).



================================================================================

19441 Genus Lyc~ides
t
ARG. C
FIG. 1. EVOLUTION AND SPECIATION OF UNCUS IN LYCBIDES (All the figures are X 33)
F- length of forearm of falx.
H- length of humerulus of falx.
u - length of uncus lobe, equal to distance between apex of falx and apex of shoulder.
FHU- triangle for measuring relative dimensions of parts. X -hypothetical ancestor; FHU = 0.25 + 0.22 + 0.22 = 0.69 mm. AGN - agnata agnata Staudinger, prep. 193, "Maralbaschi [Maralbashi, W. Sinkiang,
Central Asia]" ex coll. Weeks, M.C.Z.;
FHU = 0.33 +
0.26 + 0.30 = 0.89 mm.
ARG.A - argyrognomon Bergstrasser ssp. (ssp. anna Edw. prox.) , prep. 348, "Brewster, Washington [N. Americal, 18-VII-1940" coll. Stallings-
Turner; FHU = 0.36 + 0.33 + 0.27 = 0.96 mm. ARG.B - argyrognomon bellie~i Oberthuz, prep. 189, "Corsica [S. Europe]'' ex coll. Weeks, M.C.Z.; FHU = 0.33 + 0.30 + 0.25 = 0.88 mm. ARG.C-a~gy~ognomon Bengstr. ssp.
(ssp. opulenta Verity prox.), prep. 211, "Alto Adigo [N. Italyl 3-VII-1930,'' ex coll. Weeks, M.C.Z.; FHU = 0.39 + 0.40 + 0.27 = 1.06 mm.
SCU- scudderi scudderi Edwards, prep. 168, neotype, "Saskatchewan [N. Americal
[leg.] Kennicott," M.C.Z.; FHU = 0.45 + 0.34 + 0.34 = 1.13 mm.
SUB - subsolanus Bremer ssp., prep. 242, "Korea [E. Asia], 27-VII-1933, leg. Suk," M.C.Z.; FHU = 0.44 + 0.39 + 0.39 = 1.22 mm. MEL- wzelissa samuelis Nabokov, prep. 338, holotype, "[Albany, New Yorkl Orig. Pl. 6, fig. 6, Butt. N. Engl. Cab. S.H. Scudder," M.C.Z.; FHU = 0.57 + 0.3.5 + 0.44 = 1.36 mm.
ISM - ismenias calabricola Verity, prep. 152, "San Fili (Cosenza) , Calabria [Italyl 17-VI-1920 [leg. fam.] Querci," ex coll. Weeks, M.C.Z.; FHU = 0.74 + 0.56 + 0.49 = 1-79 mm.




================================================================================

110 Psyche Bept .-Dec.
1 view evdution in Lymides as a twofold process of growth: 1. as a generic growth - involving the who& of the male geni- talic structure, so that the absolute size of the uncus (i~depmd- ently from the size of the wings) in its general graduation from the most primitive stmctures (F -I- H 4- U = about 0.9 mm,) to the most specii-dized ones (F + H + U = about 1.8 mm.) is doubled at the maximum limit of development; and 2. as a spe- cific growth - a process acting upon the relation of parts F, and U, attacking one part more strongly than the other, where- upon the latter tends to catch up with the former, producing at a certain stage stabilization and equilibrium, which eventually are again broken by unequal growth Details camot be dis- cussed here, but it may be noted that the generic growth pro- duces more robust structures in the palearctic section than it does in the nearctic one; that there is also a difference in the rhythm of the specific growth (H being the part conspicuaus1y affected in the palearctic branch, while it is the reIation U/H which grows in the nearctic branch where H is more cramped and sluggish); and that throughout the general process stunted by-products ocmr (holarctiml~y), reduction in absdute size of structure synchronizing here with reduction in size of wings. 1 have separated the extremely munerous subspecies of which some 120, most of them badly chosen and poorly described, have names (with up to four synonyms in some cases) into six specific groups. In each there is a considerabk range of racfd fluctuation in the general size of the structure, arid in F/lJ and a more limited indivkIua1 fluctuation in H/U, but there is a con- venient constance in the structural proportions (and in other structural details not mentioned here 1 of foms clustering around the main peaks of speciation. These peaks are: ugnata Staudinger : small structure, with H smaller than F and slightly smaller than U;
wgymgnomm Bergstriker: small to average? with H sub- equal to F and greater than U;
subsolums Eversmann: average, with H smaller than F and equal to u;
scudderi Edwards: small to average, with H still smaller than F and equal to U;
is not improbable that qmta produm in Turkestan a form paralleling scddm' (see Nabokov, 1.c. : g5, nota) .




================================================================================

19441 Gems Lyc~ide~ 1 11
mdissa Edwards: average to large, with H much smaller than F and smaller than U ;
1
isme~ias Meigen: fairly large to very large, with H much smaller than F and greater than U,
From the arrangement on fig. 1 where selected examples of proportions are given, it will be seen that wgymgnomon, coming from an ancestral structure from which agnata was also derived (and which on the basis of certain data provided by other genera 1 am tempted, being human, to furnish with certain characters, namely with H and W both equal to 0.2 and slightly smaller than the small F), produces two branches, which run pardel to each other in the general growth of parts. A complete sequence of intergrades (more complete than I ariginaIIy thought) exists between argyrognomon and smdderi in the palearctic branch and between wgyrognumm and ~~bsoZu~w in the narctic me; and 1 would not hesitate a moment to assign to suho.Janus and sqdderi a subspecific position with the polytypic argyo- gnomon had they not been centers radiating as it were their om forms and, on the other hand, had they been separated from melissu and ismenias respectively by a definite hiatus, which is not so, since racial intergrades (with a corresponding combina- tion of pattern and structure) exist here too, It may be. added that the genus is distributed from the polar regions to just below latitude. 40' in Europe and eastern North America, and to at least 30' in western North America and Asia. Its cradle is a lost country of plenty beyond the Arctic circle of today; its nurseries are the mountains of central Asia, the Alps, and the Rockies. Seldom more than two and never more than three species are known to occur in a given geographical region, and so far as records go, not more than two species have ever been seen frequenting the same puddle or the same flowery bank. When about to draw up detailed comparative descriptions of the numerous forms, some of them new, involved in my exami- nation of this genus, 1 was confronted by the fact that the pat- tern of the Lycxnidz had never been adequately analyzed by systematists, On the other hand, none of the works especially devoted to schemes of stripes or lines deal with that family nor can I adapt anything they contain to my needs, since pattern development and correspondence in design vahes are discussed



================================================================================

112 PSYC~? [Sepbhc.
by authors (Eimer, Kmnezov, Schwanwitsch, and others) from a point of view with which I entirely disagrke? Thus 1 have been forced to devise a scheme of my om. Before passing on to this scheme, certain methodoIogka1 points must be explained An extremely exact and simple method of mapping the wing characters has been suggested by the fact that the wing is crossed by a set of concentric scale lines of equal breadth (very constant1y about 0.06 mm.; sinking to 0.05 only in dwarfs and rising to 0.07 only in giants). Although a few of these lines may fork a here and there, their curved course is, on the whole, remarkably regular) and easily followed from costa to dorsm. By stating the meridian of the scale line and the parallel of the vein, the position of any pint on the wing can be given) anti by counting the scale lines occupied by a marking, the extension of the latter can be adequately measured both in its absolute size and in relation to the whole expanse of the wing. At the mot of the wing the scale 1hes are badly blurred, since the scales here are coarse and irregular, I have thus taken for 0 the scde line crossing the wing through the base of Cu? which is especially convenient as then Lhe axis of the forewing discoida1 made (ie. the two discaim or cross veins) coincides more or less with the course of the hundredth scale line {from about the 95th in average shed specimens). Out of a great number of specimens examined and rneas~ired, an average looking Lyc~ides was selected the discuida1 made of which lay exactly upon the hundredth scale ]hie (see pl. V? the model of which was a Colorado male of melissa m&sa Edwards, to which macdes R2 and R8 have been added from other hdividuals) .
When prolonged beyond the wing, the scale lines are seen to form concentric circles (the curvature of the central and distal lines, forewing, and that of the distaI ones, hindwing, showing almost gmrnetricd regularity)' These, however, are not con- centric with the termen (especially in the forewing) ' and thus 'Whiie deeply enjoying the profusion of fascinating figures provided by those authors; and of course Kumezov's mwterpiece (1915, hszctes 16pkIopt4res (Nasekomye &&uekryIye) x (I), in Fame de la Russie) is nnsurpassed by any other general survey of the morphology of Lepidopfera. 'This seems to be a more frequent occurrena in large race than in small ones, and takm place more often distally than basalTy but 1 have not yet come to any conclusion regarding the morphological value of this character. *They are concentric to the termen h wresentatives of other subfamiIies, e.g. in Thech Fabriuus (sJ.].




================================================================================

the outline of the latter seem as it were carved out (as if soms body had taken a sheet of paper that happened to be neatly ruled and had cut out a butterfly, ignoring the lines), after which the transversal dispositbn of the markings was more or less adapted to the new shape (especially in the case of the mom distal markings) in consequence of which they ceased to follow the curvature of the scale lines, Its center in regard to the fore- wing lies outside the root of the latter at a point corresponding to the root of the forewing on the opposite side of the thorax, ie., at a distance from the base of the wing equal to the breadth of the body at that point; the hindwing center, however, is situated at the very root of the wing (base of costa), so that in order to make the two curvatures coincide, the right hindwing must be placed upon the right forewing in such a way as to have its hub coincide with the root of the left forewing (see plate V) . My ignorance of mathematical and mecharkal mat- ters is prodigious, ad thus 1 am quite incapable of fo1Iowing up certain lines of thought which these curious facts suggest. Four veins haw been lost in the course of the development of the Lycmidze or of their ancestors. The first to go was an additional radial nerde between ScR and Rs. The next to go was 3A of hindwing. Its more recent disappearance is suggested by the rather constant rhmiform shape of macule 2A and by a slight halving of the cretule (q.v.1 due to the occurrence of a line of weak scales (or a very slight scar) following the old 3A course upon a slightly darker ground. The last two *ins to go were 1A and M, probably more or less ~muhaneous~y, their remnants being very similar. These remnants are: the still quite definite separation of first made (q.~.) in 1A from that in Cu2 (the oldest set), the somewhat less definite (in hindwing especially so) separation of the second made (q-v.) in 1 A from that in Cu2 (a more recently evolved .set) and the distinct scar of vein IA. 1 haw treated it as an existing vein in my classifica- tion of macules, A similar scar is visible in cell RM, the intra- cellular mcuk of the hindwing being placed udw that scar (in other genera there is also an upper made), and conse- quently 1 ca11 it M. The discoidal double macde (RM) placed upon two very weak and often partly obsolescent discales, is very like macules Cu2 + 1A (the + denoting their frequent fusion). It seems likely that the third macdes in Sc and Cu2 of the hindwing travelled to their present positions dktad after



================================================================================

Psyche [Sept,-Dec.
VOL. 51, PLATE V




================================================================================

the disappearance of the veins that had once halved their cells. In the forewing the last radial is numbered R+ since I have not come to any conclusion as to which of the initial five veins disappeared. The stalking of Rg and R4 seems to have occurred after the (rare and weak) first and second macules in E& reached their present position from a point adjacent to cell RM; their weak condition seems due to the subsequent segrega- tion in the prison of the shortened and narrowed interspace. An examination of all known genera of Lycsenidse, clues provided by aberrational individuals and certain ontogenetic data suggest that the maculation of a given interspace develops phylogenetically in result of a series of recurrent waves or rays of pigment, each shorter than its predecessor. An initial wedge- like or gusset-like infuscation, in the proximal corner (against * cell RM) of a neutrally colored interspace, grows distad, ex- tending along the interneural fold. This ray broadens distally; the limit (and transverse breadth) it attains varies, and this variability is responsible for the variable position and inter- neural breadth (filled completely in "striped" forms) of the subsequent macule. The latter- is formed by a gradual deepen- ing and concentration of the fuscous pigment at its maximal distal limit, which in the case of the first made to be evolved, is subterminal. The rest of the fuscous extension is weakened, owing to this local concentration, and finally degenerates and disappears, leaving only the residue of its distal limit and the initial wedge-shaped store of fuscous in the proximal comer, whereupon the whole process is repeated (in the majority of the Lycddas). It is repeated with a little less vigour but with more variety in the limit of the fuscous extension and hence in the position and size of the second maczde which is formed discally in the same way as the first was formed subterminally. In some interspaces the number of which varies in the Lycse- nidae, a proximal wedge still remains, even after the terrnina- tion of the second process. At this point it may not have suffi- cient strength to extend again but a certain concentration of fuscous does occur, with the formation of a half hala dfshlly, (see halo), this gusset-like macule appearing to the eye as a sessile third macde ready to emerge completely and creep in the wake of the second one. However, in certain interspaces a third wave of fuscous may extend as freely as it had done in the second process and a third macule is formed more or less dis-



================================================================================

EXPLANATION OF PLATE V
Disposition of pigmented wing-markings in average Lycceides FOREWING
Terminal
line
Praeterminal
mark
(outer first)
From base
of interspace
Macule
(second)
Stretch
(11-1)
Semimacule
(inner first)
Interval I
with aurora
Terminal
space
Termination
of vein
cua
144-146
Other macules: First discoidal RM (= 10 scale lines, on 100) ; Lateral macule Ra (= 5 scale lines, on 125) ; Second macule Rs (= 4 scale lines, on 145) ; First macule R3 (evanescent).



================================================================================

HINDWING
cui I
cui
at 151
cui
105-112
cui
104-109
cui
130-143
cui
147-150
cu2
140-143
Cuo
at 145
Other macules: Third in Sc (19-27) ; Third in Cu2 (37-43) ; One in 4A (43-46) ; First RM (= 5 scale lines, on 57) ; Second (R) M (= 6 scale lines, on 32).




================================================================================

118 Psyche [Sept .-Dec.
cally. The occurrence of yet a fourth process has been noted only in a limited number of forms (e.g. in the Lycaenidae like patterns of certain Riodinidae) .
Having retained a certain vitality even after it has been formed (or owing to an extension of the wing membrane in the termen) the first macule splits, i.e., the distal part stretches and snaps off and then a fissure is formed, within which very often the neutral ground undergoes an auroral andor structural differ- entiation. In certain species where the general process started very early (e.g. in Tomares) a splitting occurs too in the second macule of the interspace (and the resulting fissure is also differ- entiated aurorally from the ground, or, e.g. in Cosmolyce boeticus Linn. (Catochrysopim) is filled with white structural scales).
Thus the difference we see in the position of the same made when comparing two specimens is really a matter of different limits attained by the sequence of initial rays. In comparing specimens, however, the eye sees those differences as the result of the actual "movement" of this or that made distad and this is a true impression, inasmuch as a macule is formed at different limits of the distally progressing infuscations. On the other hand, the white cretule capping a semimacule proximally (and produced not only by a gradual draining of the ground on the part of the first macule but also by the force of the stretch attending the splitting of the latter), is not at all "growing basad" as one is tempted to see it in some forms: in direction of growth and in shape it adheres to the general standard, for it should be noted that the essential shape of a macule and its halo, of a semimacule and its cretule, of an interval and its aurora, of a prseterminal mark and its scintilla, is obovate, sagittate, cor- date, arcuate, with the wider part directed distad; this outline repeats that of a sessile macule which in its turn conforms to the shape of the apex of the cell; or in other words, the shape of any of these markings renders macrocosmically the shape of each distally broadening scale and microcosmically the general fan- wise expansion of the wing and its cells, and is influenced in de- tails of outline and direction by the apical andor cubito-anal development of the termen (alone the ciliary markings, lying as they do beyond the membrane of the wing, point distad). I see no trace or possibility of the basally directed development of



================================================================================

1944 I Genus Lyc~ides 119
markings postulated by authors to explain certain phenomena of pattern. I
Pseudo-linear arrangements of markings, insofar as they oc- cur in the Lyczenid~~ must be also briefly noted. The terminal line is the only sequence of interspatial-markings for which I employ the word "line" at all, as it is the simplest term. Al- though it may be the remaining maximal limit of an infuscation preceding the formation of the first macule, its connection with ciliary elements places it in a separate class (submarkings) from the macules. It would not have mattered much had I called it "limbal" with Herrich Schaffer or "extreme" with Schroder, or 'cmarginal" with the British authors. But if I called it "Line I" with Eimer (who has eleven of them numbered basad) or "XII'7 with Verity (who has twelve of them, numbered distad)? or "22d" with Kusnezov (who has twenty-two) or "external I" with Schwanwitsch (who has three such external ones) or "Randbinde I" with Suffert (who has two such "Randbinden"), then I would be instantly involved in a wild confusion of man- made patterns. I fail to perceive in the Lyczenid wing any sug- gestion whatsoever of initial transverse lines or stripes forming, or having formed, an integral part of the pattern and lending themselves to classification and "homologisation." In Lepidop- tera generally, the limit of a lost ancestral infuscation in any place within a given cell, may produce? in combination with a similar limit occurring at more or less the same point in an ad- jacent cell, what may be loosely termed a line. When this occurs in several interspaces without- a special macular differentiation in any, and is followed by various adjustments and adaptations to the distal outline of the wing in the course of more or less synchronized stages of posteri6 and anterior development of the termen? then the line may seem very perfect to the eye, but it is the result of those processes and not a "primitive" line which Mother Nature automatically traced with her brush on one butterfly after another as soon as she had stuck on the wings.
It is never the line as such that 'Lbreaks" into ocelloid mac- ules. Such macules are formed by the initial spread of fuscous, or not at all; and sometimes when the latter had been strong enough interneurally to span that space, the resulting macule may be broad enough to "~onnect'~ with any other macule (not



================================================================================

120 Psyche [Sept.-Dec.
necessarily of its %wn,') ie., synchronous series) farmed in an adjacent cell; or) more seIdom, during the procdss of concentra- tion + draining + isolation the mcule may steal additional pigment from the ground of a neighboring unoccupied inter- space and form therein part of its halo. Even in the most zebroid species of Cutochysopim or The- cZim, the macdes peep through their linear disguise. If on the bask of same synthetic ((prototype" we tried to classify these lines (say Lx? Ly, Lz), we would be cmtinuousIy mistaking proximal and distal parts of split macdes for components of different hear sets, or, in other cases, would come to the non- sensical conclusion that the same macu1es (q,, the second mc- LI~E of the posterior interspaces) fom the lower part of Lx in one species, the lower part of Lz in another, and m intermediate Ly in a thjrd. The illusion of a stripe in the subfamilies men- tioned is due to several varbusly combined factors. The mac- u1e.s in two or more adjacent cells may be bar-like, with halos formed on1 y laterdly. Sessile third macules (half hdoed, ix., ody dish1Iy) wedged prmima11y in their interspaces, e.g., in R4 (just above the outer part of a split discoidal macule) and in Ma (just between the discoidal outer portion and the second macule in Cul), comkiined with a posterior sequence of second macdes in Cul, Cu2, and A1 may comp1ete the ilhsion of a stripe cross- ing the wing radianalIy. Moreover? when these macules are com- paratively weakly pigmented, the eye tends to copfuse them with pmtions of ground color; or a complete tramverse sec- tion of brown ground between ('white lines'' (formed by the inversely in regard to each other directed half halos of two dif- ferent macular series) may be mistaken for a %tripe." Re- markable cryptic phenomena in some genera produce yet other illusory patterns, and a &'white line'' that the eye folhs across two cells my rdly consist of a proxima1 half-ha10 in one and a distal one in the other. Finally, it should be kept in mind that among the second macdes my three may be dwys seen in line provided that two of them (such as Al and Cu2 or MI and M2) are those which, throughout the family, are more or less linked together in their movement distad, Although quite possibly my judgment may be affected by -the fact that the genus which 1 have especially studied and to w11kh we must now turn is most honestly %potted)7 - and also by the fact that I am interested more in what happens within a given interspace than in the



================================================================================

wing pattern as a whole, still I am quite sure that it would be a waste of time to try and twist this or that Uqsion created by a transverse combination of Lycmid macdes into this or that "pmtotypicd line.''
The categories to be discussed are: 1. Size and shape, 11. Ground. 111. Cyanic overlay, IV. Vadosal elements, V, Scintil- hnt elements. VI. ~rscales, VIT. Terminal submarkings, VIII. Macdation. (Number of specimens of Lyc&des forms examined: 959).
1. Size and Shape.
Length of forewing (from base of Cu to end of MI) in small- est individual measured: 7.5 mm, with length of hindwing (from base of Cu to end of M2) : 6.5 m. ; in largest individual measured these lengths are: 18.5 and 1.5 mm. respectively, th~s giving a range of 11 mm. and 8.5 mm. Number of scale lines ranging from 140 in forewing and 115 in hindwing to 260 and 210 respective1y. In average sized forms the number of scale lines varies from 190 to 2 10 in the forewing and from 160 to 170 in the hindwing. The hindwing varies less than the forming in the number of scale lines but more in shape. The most distal point of the termen of t he hindwing lies either rather anteriorly (high aagted shape), namely between MI and M2, or more posteriorly (low an&d shape), between M2 arid M3, or rather exactly at the end of Mz (average shape) ; or the termen is evenly rounded, ie., rum almost concentrical1y to the sale lines in the stretch from MI to Cul this however only occurring in stunted indi- viduals. In especially high-angle individuals the scale line which in the hindwing coincides with the tip of Cq (further on termed s.1.C~~) abuts anteriorly at the tip of Rs and cuts off a terminal segment of about 20 scale lines at the point of its greatest expanse (in interspace M2) ; but an- other individual with the same number of sale lines in M2 will seem less conspicu~usly angled if s.1.C~~ reaches anteriorly a more distal point (say, between Rs and MI) since the segment cut off by the line will occupy a lesser number of scale lines. In low-angled forms s.1.C~~ may abut at MI, thus cutthg off the terminal parts of ody two interspaces instead of four. Fi- nally the segment itself my be either of a fuller or more apid



================================================================================

122 Psyche Kept .-Dec.
shape, and when this difference exists the wing of one individual may look rounded and that of another angular1 though actually both are high-angled (the tips of Cul and Rs being connected by the same scale line in both). In the ciycula~ shapes) found in stunted specimens? sc.l.Cul abuts at M2, practically coincid- ing with the termen and thus cutting off no segment at all. In the forewing the variations are less conspicuous but there is gen- erally some correspondence between the wings since in high- angled forms the forewing is apt to be "pointed?" i.e., with the scale line which connects the tips of M3 and R4 cutting off a larger segment (about ten scale lines in a "rounded)' forewing and about twenty in a '(pointed" one). Short forewings (where the proportion between breadth and length is less than five to four) and long forewings (when more than five to four) may have? together with difference in shape, a certain significance in subspecific values. It may be added that there is a certain con- nection between shape (Le. vigor of growth in termen) and color (vigor of pigmentation). A low-angled shape is generally asso- ciated with weakly pigmented undersides? and these are gen- erally strongly pigmented in races with high-angled hindwings. 11. Ground.
Upperside) both sexes: ranging from neutral fuscous or weak brown to blackish. Costa in hindwing above Sc of a scaly neutral fuscous still weakened by the addition of colorless or very faintly iridescent scales. In a few female forms, with greatly developed upperside aurorz (see VIII 4), the fuscous ground may be intermixed with sparse auroral scales (the beginning of a brightening of the ground which in both sexes of PZebuZina is well on the way to complete predominance? as occurring in Lyc~nina) .
Underside, both sexes: ranging from fawn to brownish; or from white (colorless scales completely covering some, or all? neutral ground areas) to whitish fawn; or producing a greyish or bluish effect due to the even admixture of colorless or faintly iridescent scales with a more or less developed ground pigmen- tation. Occasionally the veins and the vein scars appear marked in a lighter shade. The forewing is generally of a slightly more diluted and smoother tone than the hindwing) and in one and same race the ground of the female is generally slightly richer than that of the male.




================================================================================

111. Cy&c meday.
Upperside, both sexes: structural scales inva&g the ground from the base with more or less vivid violet blue; partly (a) or almost completdy (b);
{a) clothing or dusting only certain
areas (ix., absent discally, or only empwpling the cretules (q. v.) in the female) or reduced to a few scales at the base; (b) overlaying the ground evenly or more or less sparsely (&em, leav- ing out minute bald patches and the vadosal elements, q.n) but always keeping clear of the costa h both wings, of most of the subc~std area in the hindwing (see further, V, 1 and IV, $6) and reaching &stad a maximum limit situated at a distance of about three scale-lines from the termen (see IV, 4) and less sharply defined in the female than in the corresponding male; the intensity and tint of the videt blue depending upon the density of the scaling producing it, as well as upon the funda- mental pigmentation of the wings.l0 Reduced or absent in the female considerab~y more often than in the male, where its corn- pIete absence occurs only in a few races. IV. Vadosd dements.
Racially more or less characteristic portions of fuscous upper- side ground inasmuch as they are isolated, defhed, and strongly pigmented in forms (mainly male) with dense cyank overlay which in its spread distad leaves 'tdry'' or fails to reach always three (fourth, fifth, and most of sixth), but often all of the fold Iowhg ten ground elements: the (1) uadosa proper: a longi- tudinal stretch of ground thickly or finely sheathing a vein throughout its course (or only ternidly : (2 ) twmimd mdosa) , often broadening towards its tip (on veins R4 /Rs/ down to IA) to form there the basally tapering (3) wudosd triangle (in shape and position a more or less exaggerated silhouette of the corresponding inner triangle q.u. of the underside) which may occur independently and which in its turn ft~ses with (and rep- CuIlhg at random definitions of thae shades from m&d descripfions of Lycmides fom, I find: dull violet, shiny bluel glossy violet blue, dky lilac blue, deep puwIe, hyssop violet, lamder bluel pruinose blue, ,pinky Hat, violet with a phk tinge, and at least two authon have found in thew races a greenkh cast. All th~c, more or Iess mbjective, color impressions arc worMess as racial charactm unless the combination of the two factors producing thc color efkct (in fresh spccimem) be mrcfully malyzcd in cornparkan with fresh specimens of ather races (of the same and of different species).



================================================================================

124 Psyche [Sept.-Dec.
resents the neural thickening of) the (4) vadum.ll a linear or more extensive marginal space of pigmented ground, from apex to tornus, between the limbal limit of the overlay and the ter- men, at its narrowest reproducing the terminal line of the under- side, apically turning into the delicate (5) costal vadum of the forewing, merging with the distally fuscous Sc area ((6) sub- costal vadum) of the hindwing, distally connected with the vadosal fringe q.v., and with the outer triangles q.v. and proxi- mally (in the hindwing) often joined more or less thickly by means of an (7) interneural vadosa with the (8) insula proper (as differing from (9) insula Rs II and (10) insula RM) which is a frequently occurring, more or less isolated, roundish blotch or point of conspicuous fuscous repeating in all or some inter- spaces the corresponding prseterminal mark q.v. of the under- side (also, but usually faintly, macule Rs 11, and in some cases, mainly in females with strong overlay and mainly in forewing, macule RM), and sometimes appearing as a blacker spot within the vadum when the latter is extensive enough to surround it, but not sufficiently dark to merge with its pigmentation. V. Schtillant elements.
1. The scintillant pulvis: structural scales more or less ex- tensively dusting with metallic greenish blue (in strongly pig- mented forms) or turquoise (in weakly pigmented or white forms) the ground at the base and in the anal interspaces of the underside; mainly in hindwing; sometimes quite absent or re- duced to a few scales next to the body. Upperside: confined to the dorsum and to the proximal and posterior part of the sub- costal interspace of the hindwing and intergrading there with the main overlay; in a few female forms, occurring also on the upperside of the forewing where it clothes the costa and lines the veins discally (i.e., more or less corresponding to the distri- bution of short white hairscales in the male) ; consisting there of rather coarse scales of a dull turquoise tone suggesting "dead" parts of the cyanic overlay.
2. The scintilla: l2 a variable number of scintillant scales more or less thickly and evenly grouped, overlaying the pigment "- "fuscous border," "bordure noire," "Distal Rand," "terminal border," "kraie- vaya polosa7" "marginal streak," etc., of authors. Possibly remnants of a dense scintillant pulvis which had covered the whole of the hindwing, completely swamping all its markings, at some period in the evolution of the Lycsenidae, as it still does in certain Asiatic species of Albulina,



================================================================================

19441 Genus Lymides 125
of each prseterminal mark of hindwing underside; tending to be gradually reduced from Mg or Ma costad, and often lacking in the anterior interspaces, but seldom missing in the posterior ones; very poorly developed in some forms but only individually quite absent; in most cases placed rather proximally upon the mark, Le., not reaching its distal limit, so that the latter spreads out beardlike from underneath the scintillant incrustation, if viewed from the termen; (the following more individual than racial variations in position are to be noted since any one of them can be stabilized specifically in other genera) sometimes coming in complete contact with the aurora (q.v.1, but often well separated from it by a tendency to occupy a median, or even distal, position within the mark; sometimes absent from a more or less conspicuous point in the center (upon the inter- neural fold) which thus forms a blackish pupil; in some cases agglomerating band-like across the mark; or distributed un- evenly, with patches and dots of black showing at different points; but in a few cases overlaying the mark. completely (with or without a pupil), or, as it were, overlapping or replacing it in cases when the pigment of the mark tends to obsolescence or is quite gone; in shade varying (racially, inasmuch as the pigmentation varies racially) to the naked eye from turquoise (in poorly pigmented forms) through peacock blue (at an aver- age or reduced development in well-pigmented forms) to golden green (when completely overlaying a strongly pigmented mark), but hardly distinguishable from the scintillant pulvis under lens (both sets of scales being turquoise), the aforesaid variations in color depending on the angle of light, the compactness of scales, the pigmental basis and frame -and a subjective approach on the part of the observer.
VI. Hairscales (and androconid scales)
1. Hairscales of forewing, in male: very short, white, bluish, or pale violet blue (according to light) ; of a bristly appearance under lens; projecting distad (apically and tornad when paired on a radial vein, on each of which they may form a sequence of basally pointed arrowheads) and sparsely to rather densely dis- tributed (more or less distally) within cell RM and throughout the circumcellular area distad, lacking at the base of its posterior Glawopsyc.he, Lyci~na, and Tomares, and which subsequently had disappeared, leaving the scintilla as seapools are left by the sea at low tide.



================================================================================

126 Psyche tsept .-D~C.
part, stopping or diminishing in number at or beyond limit I1 (i.e. the limit corresponding to the emplacement of second mac- ules, q.v., of the underside) but sometimes just reaching in hindwing (where however they are somewhat less conspicuous throughout than in the forewing) limit I (i.e., the limit corre- sponding to the emplacement of the semimacules q.v. of the underside). Above RM in forewing mainly along veins Sc, Rl, RB, and RS, agglomerating on their slopes (giving the vein a pinnate appearance) when the overlay is dense enough to elimi- nate the vadosse in the costal area which then seems, especially in freshly emerged specimens, rather densely powdered with white (costal pdvis) .
2. Male androcoÌàw scales: a microscopical character: mi- nute battledore-shaped scales, in outline, size, length of pedicule and number and density of "knots" varying in individuals of some forms, racially more or less constant in others (especially in stunted or overdeveloped forms), and of ten duplicated by specifically different races (and thus lacking the specific im- portance assigned them by Courvoisier, 1917, on the strength of scanty, and more or less misidentified, European material). 3. Costal fringe: short hairscales (allied to the male hair- scales) in both sexes rimming the costa with white and very conspicuous in specimens with a strongly pigmented costal vadum (IV, 5).
4. Basal cilia: long and very long silvery white, bluish or drab hairscales clothing basally the upperside of the hindwing (reduced in forewing), sweeping in a distal and then downward direction across the proximal part of cell RM, extending rather far into interspace Cul where they just reach limit II, still fur- ther in Cua + 1A (almost to limit I), and spreading from base into 2A and 4A, where they stipple the scintillant pulvis of the dorsum.
5. Dorsal cilia: white, or producing on the upperside a light blue effect as if daintily dyed. Springing from a very faintly fuscous dorsal margin and sometimes slightly infuscated them- selves. Equal to about 10 s,l,, somewhat shorter in forewing. 6. Terminal cilia as seen from the underside: long hairscales (equal to about 10 s.1. in forewing and to 12 s.1. in hindwing) ,
attached to the termen, proximally denser than distally; silvery white or with slight bluish or mother-of-pearl reflections in certain lights; sometimes, especially in females, more or less



================================================================================

19441 Genus Lyccekies 127
infuscated, ranging in shade from drab or pale fawn to brown- ish; completely pigmented or in part, ie., only proximally or only distally and then either from tornus to apex throughout, or only along a limited section of that range. 7. Vadosd fringe: consisting of rather short pigmented hairscales rooted in the vadum, upperside, and thus doubling basally the terminal cilia of the underside; usually equal to 4 s.1.; tending towards the fuscous of the vadum from which sometimes it may be almost indistinguishable to the naked eye, Viewed in cross-section the short dark vadosal fringe (rooted in the still darker vadum) is seen to overlay the long white underside cilia (rooted in the distal edge of the terminal line) to about 2/5 of their length. The very slight jutting of the dark hairscales of the terminal line just beyond the rim of the mem- brane forms a kind of prop for the base of the ciliary hairs which thus are encased between it and the vadosal fringe. If the cilia are viewed from the under surface by the naked eye, an illusory more or less dark ciliary line seems to run along the middle of their transverse stretch: this is due, first, to the cilia abruptly losing their quilted appearance at 2/5 of their length where the edge of the upperside vadosal fringe stops, and sec- ond, to minute portions of this edge being discernible in be- tween the white ciliary hairs, as they become less dense distally. If, moreover, the distal part of the cilia on the underside hap- pens to be infuscated and if this infuscation begins at just over 2/5 of the length distad, then on the upperside too there is a similar illusion of a ciliary line (but of a light one this time), due to a narrow stretch of unpigmented cilia showing between the distal infuscation and the edge of the dark vadosal fringe which shuts off most of the white basal part of the underside cilia. My abundant material has not proved the occurrence of a true ciliary line in LyciEides, i.e., of an actual infuscation of each ciliary hair only at its middle, or of shorter hairs (among the longer ones) infuscated only at the very tips. VII, Terminal submarkings of underside.
1. The terminal line: edging the termen proximally with more or less dense fuscous from about the middle of iSc. in secondaries, and from the tip of R< in primaries, to the tornus; consisting of very short distally directed hairscales (which very slightly jut beyond the termen), and in its interspatial aspect



================================================================================

128 Psyche [Septa-Dec
resembling a garland more or less raised and thickened at both ends. Very thin and faint in weakly pigmented form. 2. The inner triangle: a fuscous triangular basally tapering ' mark formed upon the termination of each vein (mainly from Mi to Cu2) by the meeting of the thickened prseneural ends of two adjacent sections of the terminal line. Not necessarily absent in weakly pigmented forms. See also IV, 3. 3. The outer triangle: a fuscous subtriangular distally taper- ing mark formed upon the proximal part of the terminal cilia (and also occurring sometimes upon the vadosal fringe of the upper side) independently from the general pigmentation, if any, of the latter and placed directly opposite (base to base) the inner (or the vadosal) triangle, which it repeats in reverse, except that its base is usually narrower and its point more or less truncated. Mainly in hindwing. Seldom leading to any conspicuous scuttelation in the forewing. VIII. Maculation of underside.
Counting from termen basad a first macule (split into inner and outer part) and a second macule are both represented in Sc., Rs, and 2A of hindwing, in R3 (where, however, they occur seldom and are always much reduced) and R4 of forewing, and in Mi, M2, M3, Cul, Cua, and 1A (small in the two last) of both wings. A third made supplements the set in Sc and Cua of hindwing. Moreover, there is a small lateral macde in 4A of hindwing caught in the blind alley of the dorsum and sometimes a small lateral macule is somewhat similarly trapped in R2 of forewing (where the eye sees it as "belonging" to the transverse series of second macules). In both wings the discoidal cell (a double interspace R andM) has its own first (double) discoidal macule,13 the rheniform RM, traversed by the discales (the outer segment of its R part and the outer segment of its M part form in relation to the second macules Mi and M2 a pair of sessile third macules - an important point in the case of certain other genera). In the hindwing there is a second (single) discoidal macule within the cell under the scar of vein M. All the macules are of a more or less deep fuscous and are rimmed with struc- tural scales, i.e., halos (produced by the macule having drained during its period of formation and concentration the initial pig- "Among the names employed by authors for this double macule are "discal streak," "bar," and "disco-cellular lunule."



================================================================================

19441 Genus Lycoeides 129
ment in its immediate neighborhood). The halves of the halo which has split together with each first rnacule; are termed cretdes but only the proximal one is represented in full. The inner part of each first made is the semimule (capped or rimmed proximally by the cretule) and the outer part is the prfetermind mark (adorned in hindwing with the scintilla). The fissure between the two parts is the interval (extending in average size races from about 5 to 20 scale lines in Cui of fore- wing, always correspondingly more in hindwing) which may be, and generally is, more or less completely filled by the auroral element - an agglomeration of brightly colored scales differ- entiated from the ground, and associated with the splitting of macules. In the female the semimacules and praeterminal marks may appear in darker pigment within the fuscous ground of the upperside, and the series of aurora; is often repeated, completely or in part; but in the males with average overlay only the praetermioal marks appear (as insul~) although in very rare aberrations the posterior aurorse of hindwing may be repeated (as happens more often in forms of Plebejas argus L., becomes fairly normal in its allies and is characteristic of the smaller Icaricia where the auroral development resembles that found in certain Glaucopsychince). All parts of the first macule are less developed in the forewing than in the hindwing, where again those in Mg and especially Cui (there extremely devel- oped in "tailed" genera) are stronger than in the rest of the in terspaces.
Villa. Elements of First Macule.
1. Semimacde: l4 generally crescentic, sagittate, or deltoid (pointing basad upon the interneural fold) in hindwing (from iSc to As incl.) ; when well developed, spanning almost the whole breadth of the interspace, except in Sc, Rs, and Mi, where it is shorter and often reduced to an uneven bar-like shape; often tending to the latter shape in all interspaces of forewing (from i.R4 to i.lA id.) where each is shorter than the corre- sponding one of the hindwing and may seem blurred to the naked eye owing to a weaker pigmentation. Variable in longi- tudinal extension; quite absent only in extreme individuals of very weakly pigmented races.
"The "rather narrow bent Sunule" of Scudder and the "crescent," "flat cres- cent," "arrowhead," and "chevron" of Chapman; those two authors have left by far the best descriptions of the Lycaeides pattern.



================================================================================

130 Psyche [Sept.-Dec.
2. Crefde: capping proximally each semimacule in both wings; more or less conspicuously white or whitish (almost in- visible on the powdery white ground of some forms) or, very rarely, retaining some diffuse pigment; crescentic or sagittate or squarish, i.e., more or less in keeping with the shape and size of the corresponding semimacule, but usually somewhat more pointed and larger; sometimes so greatly developed as to seem to fuse with the halo of the corresponding second macule (actu- ally it is the halo which intrudes), and then appearing raylike if the whole system, of macule I is transversally reduced; at other times, however, especially in hindwing where the semimacules are better developed, and more often in females, occupying the whole breadth of the interspace for a certain distance basad from semirnacule before "terminating" crescentically or taper- ing to a point (phylogenetically, however, expanding distad from that point), so that the sequence of cretules (especially if they fuse with the halos of series 11) has been described by observers as a "white band"; in some well pigmented forms very small or quite absent, especially in forewing. Appearing on the upperside in some females, in whitish or bluish (or violet as portions of the overlay).
3. The prcetmhaal mark: tending to be heart-shaped (ex- panding distad) in hindwing where it is generally strongly fus- cow and contains the scintilla; roughly rhomboidal or (when reduced) bar-like in forewing where its pigmentation is weaker ; situated in the same interspaces as the semimaciile distally to the latter, and varying in size accordingly; tending to complete obsolescence in some weakly pigmented forms, although the scintilla may be retained (see VJ).
4. (The remnants of an) outer cretule: colorless (white) scales diffused in the ground of the crescentic terminal space with which, when the latter lacks pigment altogether, it is prac- tically synonymous; usually more conspicuously white in hind- wing but sometimes very much so in Cut of forewing in other- wise well pigmented forms. Appearing on the upperside in some females with the same variations as 2.
5. The aurora: racially varying in extension (together with that of the interval) and in transversal development, (to- gether with that of the semimacule) ; on the underside in both sexes( but somewhat better developed in the female); rang- ing there from light yellowish to deep reddish orange; of



================================================================================

19441 Genus Lycceides 131
a more velvety appearance on the upperside of the female where it may be under or overdeveloped in comparisonsto the under- side in the same specimen and where its color ranges from a bleached neutral shade to a rich fulvous (the slight discrep- ancy in tone between the two surfaces being due to a difference in the degree of the ground pigmentation as well as to the sparser spread of colored scales forming the average upperside aurora) ; at its full development on the underside snugly fitting into the interval between the semimacule which caps it and the corresponding prseterminal mark which it caps in its turn; often represented in all intervals; tending, however, to be ill- formed, underdeveloped,. or absent in the primaries, especially in RI, Mh Ma, Cua, and A^ (termed the weak interspaces') of the male underside and of the female upperside; in a few female forms, however, hypertrophied on the upperside and espe- cially conspicuously so in the forewing, the sequence reach- ing there from costa to dorsum and swamping a stretch corre- sponding to that occupied on the underside by the inner cretule + semimacule + aurora, thus forming a broad "band" with a more or less diffuse proximal edge (see also 11) ; when under- developed the aurora edges the interval always proximally, Le., does not reach the prseterminal mark in its growth distad from "beneath" the semimacule (the remaining gap being either concolorous with the ground or colorless). It is the first to develop, or the last to go, in Cui (with its neighbor in Mg following closely). Completely absent only in extreme indi- viduals of weakly pigmented forms.
5a. Cusps: when fully developed and especially in Cui and M, of the hindwing underside, the crescent of the aurora is prolonged distad by two (inner and outer) pairs of cusps and occupies the whole breadth of the interspace; the inner cusp clasps the prseterminal mark laterally, the outer one runs next to the vein and fuses upon the vein with the outer cusp of the adjacent aurora to finally penetrate and bisect the inner triangle of the terminal line; in the forewing and in the anterior inter- spaces of the hindwing the outer cusp tends to be absent, so that the auroras (and their semimacules) do not touch the veins and are separate from each other.
5b. Lacrimce: in some richly pigmented and strongly devel- oped forms there are on the underside two or four streamlets of blurred auroral pigment coming as it were from beneath the



================================================================================

132 Psyche [Sept.-Dec.
prseterminal mark, and "trickling" distad across the terminal space (one, or a pair, on each side of the intesneural fold). VIIIb. Second, Third, Discoidal, and Lateral Macules. Second macules: if a Lycceides forewing is placed with its base towards the observer and its discal constellation is viewed from an imaginary horizontal line joining the opposite ends of the discales, the second macules, (b), R^, Mi, Mg, Ma, and Cui l5 all of which have radiated from positions adjacent to cell RM, are seen to form a fairly regular rather weak arc, sloping somewhat sideways in relation to the rheniform macule RM, as if tipped by a slight apical pull (process 3). The twin macules Cua and 1A lie outside the lower end of the arc, i.e., lead an independent existence, having reached their present position (phylogenetically, from an enormously remote starting point in comparison to the starting points of the other macules) in result of a (process 2) cubito-anal stretch of the membrane (so conspicuously retained in some genera) that had occurred at some period in the evolution of the narrow and ovoid ancestral forewing (with a similarly shaped hindwing) prior to the comparatively recent generic and tribal apical de- velopment (process 3) which in a way has tended to repeat the initial growth and elongation of the ancestral wing (process 1). These stages of unequal growth and of subsequent compensa- tory readjustment may be compared to the already discussed evolutionary phenomena in the case of the uncus. * It would be necessary to analyze a great number of ge- neric patterns (in the Plebejince alone striking variations on a P-shaped basis occur in Agriades, and a remarkable apicoid angle is formed by the macular constellation in AlbuUna) in order to bring out certain features of the position of second macules in Lycceides, but this would transcend the scope of this paper. In selecting the three positions (1 proximal, 2 cen- tral, and 3 distal) given for this genus, stress has been laid on the progress of macule Cui, but actually this may be combined, at these and intermediate stages, with shiftings on the part of the anterior series which may be removed from RM further than it is shown here. Fig. 2a shows the generic starting point of R4 whose initial rather distal position (in regard to even such "When no Roman number is appended to the symbol of the macule, the reference is to the second macule (e.g. CUI = Cui 11).



================================================================================

19441 Genus Lycasides 133
genera in Plebejiws where the constellation is of the same type) coupled with the also rather distal position of R& or Rg, when occurring (not shown in the figures), is instrumental in weaken- ing the curvature of the arc and producing its "sideways)' posi- tion already discussed. The same figure also shows the most proximal position of macule Cui which is at this stage in an oblique line (the radwnol slant, reoccurring throughout the family) with RM and Cua + 1A. Under RM this Imaginary line diverges distad from the latter's scale line to finally cross the scale-line of rnacule Cue. Fig. 2b shows a middle position which is most frequently found in this genus. Fig. 2c shows the most distal position of Cui (except that the whole series can move still further if the semimacules are further removed than they are in average forms) when the series is roughly adapted to the sequence of the first macules which in its turn is sub- parallel to the outline of the termen. The tendency to assume one of the two extreme positions (a, b) is sometimes a racial character.
In forewings of average extension (about 200 s.L in Mi and 185 in Cui) and with semirnacule Cui having reached s.1. 150 or thereabouts, the range of movement of the center of macule Cui (and it is this center which is referred to throughout), is from s.1. 105, at which initial point in Lycceides it is about 50 s.1. removed from the apex of its cell (which thus is less than made R4 has travelled from the apex of cell R4 but more than the distance covered by the other anterior macules iii regard to their respective cells - although curiously enough all de- scribers, being obsessed by the notion that macules must form "lines," speak of Cui in this position as "advanced basally") to s.1. 135, at which point it has SO scale lines to go if it wishes to reach the termen, which of course it cannot, since the split first macule occupies the remaining space. Thus its range of activity is 30 s.1. which is somewhat less than 1/4 of the length of its interspace and about 2/5 of the distance from the proximal position of Cui to the termen (this range varies racially). The width of the interval between semirnacule (inner I) and prse- terminal mark (outer I) (see fig. 2d, e, f) is mainly dependent on the position which the former had reached when the made I split (the outer part wandering distad) . The breadth of the fis- sure (interval I) ranges from 4 s.1. to at least 20 (average sized males). The space available for the progress of macule Cui



================================================================================

134 Psyche [Sept.-Dec.
depends on the position reached by the center of the correspond-^ ing semimacule (this is about s.1. 145, proximal limit, about 155, average, and about 165, distal). Thus when the latter r
reaches its distal limit (resulting in a narrow interval I, since the prseterminal mark cannot wander away beyond a certain limit), the increase in the 11-1 stretch allows macule Cul a FIG. 2. a, b, c, Forewing discal constellation in Lycceides, showing proximal, central and distal position of second macule Cul. d, e, f, Cell Cul of forewing, showing relation in position between semimacule (inner part of first macule) and second macule.
greater range (at least 40 s.1.). A terminal extension of the wing even to only 195 in Cui may produce a veritable wilderness for Cul to traverse. These phenomena have great racial im- portance.
In the hindwing the position of macule Cui varies less con-



================================================================================

19441 Genus Lycasides 135
spicuously. The same discal arc as in forewing is readily per- ceived, but the eye sees Rs in a more proximal position in relation to Mi than R4 appears in relation to Mi in the forewing (actually, both Rs and R< - especially the latter - have pro- gressed further distad from their apices than the corresponding macules Mi have progressed in their cells). Macules ScIII and Sell at one end and Ck II at the other prolong the discal arc (Rs, Mi, M2, Mg, Cui) in such a way as to form a horseshoe arrangement around cell RM: this circumcellvlar arc becomes practically a circle in some genera, where the second macule in cell RM (or a third one) is placed basally enough to act as a link. Posteriorly to this, roacules Cua + 1A, 2A and 4A form a short weak additional arc or parenthesis (also a special fea- ture in certain other genera) with its concavity toward the proximal stretch of vein Cu. The radianal slant connects mac- ules ScIII, RM, Cui (when lagging) the colon (its 1A part, however, generally "diverges basad") and the semimacule in 2A by a regular but perfectly imaginary line, traversing the wing and very conspicuous and perfect in butterflies where the anal part of the termen has been stretched tailwise. When examining Lyc<smdce patterns for systematic pur- poses, loose impressionistic descriptions will inevitably result (and I have erred myself in this respect) if the describer does not take into account the actual distances of the macules from the apices of their cells and from the termen, the actual and comparative positions of the split first macules, the extension of the split in comparison to the whole wing, the development of the terminal space, and the relation between the size of the macules and the entire number of scale-lines. I shall limit my- self here to a few words regarding the dimensions of macules in this genus.
Divided by three, the sum of scale lines occupied by the three median macules II in a specimen gives pretty exactly the mean size of the whole discal maculation in that specimen. When the relation of this number to the alar expanse in scale-lines (see . category I) is around 1/20 for each wing, the maculation in the specimen or in the race may be said to be of "average" develop- ment in both wings. Below this, it is "reduced"; above, it is "en- larged." In the forewing mule Mi is often equal to Ma but their elongation and direction may be different. R4 is smaller than Mi and both tend to be ovoids slanted towards the wing apex, these



================================================================================

136 Psyche Kept.-Dec.
two being especially sensitive to the apical pull. Ma when ovoid, tends to be slanted towards the stem of veins Rg and R4 just above the discoidal. Ma when ovoid and tending towards a proximally pointed cordate shape has its axis directed anteriorly towards the discoidal and posteriorly tornad. The same is true of Cui which is usually the largest in the series and is often very conspicuously elongated (in all these cases, of course, the actual extension and expansion is essentially in a distal direc- tion). Macules Cua and 1A which are often well separate and smaller than R4 (except when the latter is very much reduced) form together a colon, the axis of which is directed either to- wards the discoidal, and appears more or less in line with the latter's axis, or towards the apex of interspace Rg and then fol- lows its scale-line (which is most frequently the one traversing the point of forking of veins Rg and R4), as the discoidal does its own, in which case, since both lie on different sections of their respective scale-lines, discoidal and colon do not appear parallel to each other, the former slanting tonad and the colon remaining "straight," i.e., at right angles to the dorsum. The rare Rz is smaller than Cua (or 1A) while the slightly more fre- quent Rg is scarcely perceptible to the naked eye. In a general way and disregarding the difference in elongation, the dimen- sional sequence of macules runs as follows: Ra, Ra, Ai, Cua, It4, Mi, M2, Ma, Cui, with the rheniform RM (R + M) slightly broader than colon (Cut + 1A).
In the hindwing the macules forming the ~ir~mcdular arc are generally subequal, with CusIH and Mg often tending to be smaller than the rest, while Rs tends to be slightly enlarged and Sell is still more so (sometimes vaguely suggesting a very an- cient fusion of two spots in adjacent interspaces where the par- tition has been lost). Thus there is a gradual reduction in size from Sell to Mo with Cui subequal to Mi and Mg. Ch and 1A are the smallest in series II (and even slightly smaller than CuaIII) and are apt to be fused forming an hour glass-shaped or rheniform (distally convex) spot not unlike the discoidal (R + M) and of approximately the same size but having a different curvature of axis since they lie upon different sections of their respective scale-lines. The extension of 2A is almost that of Ma but (transversally to the veins) it is longer and forms a roughly rheniform blotch suggesting a more complete fusion of adjacent macules in 2A and 3A (an extinct vein) than that



================================================================================

19441 Genus Lycosides 137
of the macules in CuJI and 1AII and often slightly diverted anteriorly and outwardly from the scale-line of itp axis. Mac- ule M in the discoidal cell is intermediate in size between ScIII and Cu&I while the latter is intermediate between M and 4A which is the smallest of all. In a general way, and taking ac- count of their tendencies, the sequence in size of the macules is as follows: 4A, CuJII, M, Ma, ScIII, Mi, Me, Cu,, Rs, Sell. The sequence for the rheniforrn macules (length) is: 2A, Cua + 1A, RM. These sequences are important as they give the order in which macules in both wings tend to disappear in some races. Their reduction (racial) in one wing, however, is not necessarily accompanied by reduction in the macules of the other, nor do the rest of the markings and the general pigmenta- tion always follow suit.
In conclusion a few words may be said concerning the specific repetition, rhythm, scope, and expression of the generic characters supplied by the eight categories discussed. "Repeti- tion" when affecting a conspicuous character or a great number of characters, produces striking resemblances between certain forms (which may be widely allopatric and associated with totally different surroundings) belonging to two or more dif- ferent species of Lycasides, and this kind of resemblance I term homopsis since I cannot use "isomorphism," (the mimetic im- plications of which would be quite irrelevant in the case of this genus), or "parallelism" (which I restrict to resemblances in structural characters), or "analogy" (which is a minor form of homopsis affecting allopatric races of the same species); inter- specific homopsis to be precise - for remarkable hornoptic forms may be also supplied by generically and tribally different Lyc~enids. "Rhythm" depends on the following: if 3, L, P, T represent in one species of Lycosides certain combinations of characters as revealed by definite subspecies, and if in another species the combination L fails to be represented at all, while on the other hand P is not represented by a single definite sub- species, but is spread over several, these omissions, gaps, fusions, and syncoptic jerks will produce in one species a variational rhythm different from that of another. "Scope" refers to range of variation in a species in comparison to that of another species and in its approach towards the generic range. A species may set a unique record in one character or category, while lagging



================================================================================

Psyche [SepLDec.
behind in the others, or it may attain a good average in most characters. Finally, "expression" means the slight differences by which even the most strikingly hornoptic forms (i.e., belong- ing to different species) may be distinguished without an exam- ination of the genitalic structure,
A priori, 1 had assumed that in the course of the combination and segregation of generic characters in various racial forms (and this is incidentally the meaning I attach to the term "form") each of the six structurally different groups ($.en, spe- cies) of Lycceides would be seen to repeat certain stages of the same general (i-e., generic) variation, but would reveal differ- ences in rhythm, scope, and expression, the total of which would produce the synthetic character of one species as differing from the synthetic character of another. This has proved correct insofar as the species are known at present, although certain aspects of rhythm are exaggerated or, inversely, blurred by erratic taxonomy and by the tendency to create a new form not because of its marking some important combinational stage in the morphologic development of the species, but because of its coming from some new locality. New localities, however, are most welcome in themselves, for it should not be forgotten that immense areas, practically all of European and Asiatic Russia, as well as China, and numerous more limited areas in the pale- arctic and nearctic regions are more or less terra incognita in regard to these butterflies (although no doubt much precious material from there lies unsorted or misidentified in museums), so that one can still hope to obtain an agnata with white under- side, a subsoianus as blue as melissa, and a meha with a heavy vadum.
In delineating in this manner the principles I intend to follow in rnv subseauent discussion of racial variation in Lvca'Mes spec&, I ambided among other things by the belief that the systematist may fare better when keeping to the all important morphological moment, than when giving comprehensive geo- graphic names (the whole of China, the whole of the Moon) to hypothetical "populations" (a dreadfully misused term - and a hideous word, anyway) on the basis of half a dozen specimens taken by somebody between climb and cloud on some mountain thousands of miles away from the describer's desk.



================================================================================