Hind wings shaped somewhat as in Hypanartia, the costal border beyond the great rounded prominence of the extreme base being very gently convex, the outer margin full on the upper half, the upper outer angle broadly rounded; the upper median nervule is developed in the middle of the wing into a long, slender, tapering tail, and the lower half of the wing is strongly crenulate, and especially roundly excised in the lower median interspace and lobed on the lowest median nervule; the lower outer angle is well rounded; the inner margin plainly forms a gutter for the reception of the abdomen. The costal and precostal veins are very doubtful, being exceedingly obscure on the specimen; but the former apparently arises from the common stem of the costal and subcostal veins at right angles to it shortly beyond the base, and then curves strongly outward subparallel to the costal margin, striking the latter in the middle of its apical half; while the precostal is a simple recurved vein, directed inward and forward at the sharpest point of the costal curve. The subcostal vein is peculiar in that its first branch, originating only a little beyond the costal, approximates so closely to the costal margin as to strike scarcely outside of the upper outer angle of the wing, a place usually reserved for the apex of the costal vein; the subcostal forks again, scarcely more than one-quarter way from the base of its first branch to the margin, the middle branch continuing the curve of the main stem, and the lower branch diverging very gradually from it, and widely distant from the median vein. The main stem of the latter, with its upper branch, forms a gentle sinuous curve scarcely approaching the subcostal vein (the cell being open), and emits its first branch in the middle of the cell, or scarcely more than half-way from the base to its final divarication. This latter is unusually slight, the middle branch keeping throughout very close to the upper and distant from the lower branch. The submedian strikes the angle of the wing as far from the lower branch as it is from the middle branch of the median. The internal nervule cannot be determined.

The abdomen is full, with the third and fourth joints longest, the whole nearly twice as long, and in the middle fully as broad as the thorax.

This is the first butterfly that has been found fossil in America, and as only nine species are known from the well-worked Tertiary strata of Europe, it may properly be esteemed an especial rarity. Besides this it has a double value: first, in that it is far more perfect than any of the European specimens (nearly all of which I have seen); and, second, in presenting, as none of the others do to any conspicuous degree, a marked divergence from living types, combined with some characters of an inferior organization. When first received, the tails of the hind wings and the tips of the antennæ were hidden by flakes of stone, and it was taken, both by myself and by every entomologist to whom I showed it, to be a Hesperian, the lowest family of butterflies. The neuration, however, which, although mostly very obscure, can be deterBull. iv. No. 2- -14

mined with certainty, shows it to be a Nymphalid, the highest family, with which the structure of the antennæ and palpi and the outline of the hind wings, now entirely uncovered, perfectly agree. The first inference was drawn principally from the robustness of the body and the form, proportions, and markings of the front wings. The latter are unusually long for a Nymphalid of this type, have a remarkably straight costa, an outer border bent at the middle instead of far above it, and are possessed of a nearly transverse, median, light-colored belt on a dark ground, a subapical row of small spots depending from the costa, a spot in continuity with them in the upper median interspace, and beyond them, parallel to the outer border, in the costo-subcostal interspace, a pair of minute spots,-all characters perfectly consonant with Hesperian affinities; never combined, and each very rare in the Nymphales. It is not a little strange, however, that while the form and markings of the fore wings are hesperidiform, those of the hind wings are decidedly nymphalidiform. That the exact opposite should be a far more probable occurrence, follows as an assumption from the fact that, as a general rule, the front wings only of the lower Lepidoptera are ornamented, and that therefore the ornamentation of the hind wings is a more recent development. The somewhat variegated markings of the hind wings are indeed similar to what we find in certain Urbicolæ, such as Pythonides, but they are far more common in Nymphales, while the wing-contour is of a high nymphalideous type, quite above anything. we ever find in Urbicola.

I am at a loss to suggest any really plausible explanation of the mode of development through which the hind wing should have attained an ornamentation consistent with its organization, while the ornamentation of the fore wings, whose structural framework has kept pace with that of the hind wings, has not advanced a single step beyond a type common to the lowest family of butterflies. It may, however, be suggested as a mere speculation that the position in which the wings of many Urbicola are held in repose (the front wings oblique or suberect, while the hind wings are horizontal, and therefore more fully exposed to view) might be productive of such a result. In this case, we should anticipate further indications of such a feature, at least in fossil forms. We are acquainted with the upper surface-markings of both pairs of wings in extinct butterflies only in Neorinopis sepulta (Boisd.) Butl. and Thaites ruminiana Heer. It had escaped notice in my original study of these,* that when they are compared with living types, indications appear of precisely the same nature, although by no means so conspicuous. The rude patches of color that mark the discoidal area of the front wings of N. sepulta, and the repetition of almost similar, unbroken, transverse bars on the same portion of the front wings of T. ruminiana, when compared with these parts in their nearest living allies, are clearly

* Mem. Amer. Assoc. Adv. Sc. i, 1875.

indications of an inferior as well as an earlier type, while no such contrast is presented in the delicate shading and more complicated pattern of the hind wings. But, again, a partial comparison may be made with the markings of the front wing alone, and in the seven other species of described fossil butterflies there is not one, with the possible exception of Eugonia atava* (Heer) Scudd., in which the markings may be looked upon as less highly developed than in the living types.

Instances could, of course, be easily given from among living types in which the ornamentation of the upper surface is less variegated in the fore wings than in the hind pair, but it might readily be doubted whether this should be looked upon as having any direct bearing upon this subject; yet, even if none could be cited, it may fairly be urged that the lapse of time since the Florissant beds were deposited is amply sufficient for the loss of any such indication of hesperidiform affinities in a group of insects so pliable in ornamentation as butterflies are shown to be by the mere facts of mimicry.

Prodryas shows further peculiarities when compared with its nearest living allies. In the Tropical American genus Hypanartia, which seems to be its nearest neighbor, as in all those closely allied to it at the present day, the costal margin beyond the base is uniformly arched throughout; and the outer margin, angulated in the upper half of the wing, is roundly excised below it, giving these butterflies the common name of 66 'angle wings". They are insects of strong and rapid flight, capable of the most abrupt and unanticipated movements, making them very difficult of capture on the wing. The straight, strong costa and more elongated wing of Prodryas, on the other hand, with its nearly uniform straight, outer border, combined with the robustness of the body, indicate great strength of wing and a rapid direct flight, as in the Hesperides, but not the power of sudden turning.

In Hypanartia and its immediate allies, the cell of the front wing is closed, although by a feeble vein, and the superior subcostal nervules take their rise at more or less irregular distances apart, and run long distances crowded side by side; while in Prodryas the cell is open, and the subcostal nervules are much shorter and very uniform in their distribution; the inferior subcostal nervules also originate in Prodryas in a much simpler fashion, indicating that its ancestors never had the cell closed, although a foreshadowing of the closure may be seen in a row of special scales (or a line of color) at the supposititious termination of the cell. That this can hardly indicate a true vein appears from the fact that there is not the slightest tendency of the opposing veins to approach each other at its extremities-a tendency which it would seem should naturally precede the formation of a vein; the second inferior subcostal nervule takes its rise from the first in just about the same manner as

* The remnant of this insect's front wing is certainly simpler in markings than the upper surface of allied living Eugonias, but it may represent an inferior surface, in which case there is no special difference.

the second superior nervule originates from the main stem, neither its basal portion nor that of the first inferior nervule showing any noticeable tendency to bend abruptly and to help form the termination of the cell, as now appears in all Præfecti to a greater or less extent, and which, in some open-celled genera, seems to indicate the loss of a transverse discoidal veinlet after a previous possession. The presence of a transversely disposed pair of spots in the costo-subcostal interspace also indicates the probability that this interspace had hitherto never been narrower nor bridged by a vein.

In the hind wings, there are two features of importance, besides the unusual openness of the cell, which is scarcely narrowed apically. The first is the course of the first subcostal veinlet, which originates far toward the base of the wing, and terminates where the costal nervure is sure to end in nearly all Præfecti,* at the upper outer angle of the wing. This necessitates a shortening of the costal nervure. I do not know of a single instance of such a feature among the members of this group of Nymphales, but it is an almost persistent character in the Pierids, and very common in the Satyrids. The other point is the extreme narrowness of the upper as compared with the lower median interspace, the former being scarcely more than half as broad as the latter, owing to the slight divergence and continual proximity of the outer branches of the median vein. The only other feature in which it differs unusually from its allies is in the brevity of the antennæ.

Prodryas persephone.-A single specimen (No. 394) was found in the Tertiary strata of Florissant, Colo., by Mrs. Charlotte Hill. It is in a wonderful state of preservation, the wings expanded as if in readiness for the cabinet and absolutely perfect, with the exception of the tail of the right hind wing. The thorax and abdomen are perfectly preserved, but indications only of the legs are seen beneath the wings. The head is twisted so as to throw both antennæ upon one side, and to exhibit the palpi better than would otherwise be the case. The tongue is doubtless preserved, but the danger of injuring the palpi prevents me from chipping the stone to find it. The antennæ are nearly perfect, but the stalk is covered with a thin film of stone, which will not scale, and thus conceals the joints. The markings of the wings are perfectly preserved, but on the costal area of the hind wings are partially concealed by the overlapping of the front wings. In many parts of the wings, the form of the scales even can be determined under the microscope. This I was unable to do in any of the European fossil butterflies, although in some the points where they were inserted could be seen.

The wings are rather dark brown, deepening in tint on the front wings toward the extreme base and along the immediate costal edge, ornamented with pale markings, which were, perhaps, bright-colored in life. Front wings with a mesial, transverse, slightly arcuate band, extending *In Polygonia and some of its immediate allies, the upper outer angle of the hind wing is curiously excised, throwing the costal nervure back some distance.

across the wing at right angles to the costal border, just failing to reach either margin, divided by every nervule, its inner margin continuous and nearly straight, its outer strongly crenulate, being gently convex n the discoidal cell (more below than above), strongly convex in the ower median and submedio-internal interspaces, and strongly sinuous in the medio-submedian interspace; its upper extremity is before the middle of the wing, and incloses in its middle the base of the first superior subcostal nervule; its outer border is bent inward below the cell, exactly to the last divarication of the median nervure, and it reaches the anal area of the wing two-thirds the distance from the base. A row of five unequal pale spots crosses the wing in a straight line, extending from the lower outer angle to the costal margin at two-thirds the distance from the base; four of these are approximated in the subcostal interspaces; the fifth and largest is in the middle of the upper median interspace, but nearer the middle than the upper median nervule; it is broadly ovate and obliquely placed, subparallel to the mesial band, its broader extremity above; the lower of the subcostal spots, before the middle of the lowest subcostal interspace, is obovate, still more oblique, pointing toward the upper of the subcosto-median spots to be mentioned, and only a little smaller than the median spot. The three spots above this are equal, about half as large as the previous, twice as long as broad, rounded, subquadrate, each occupying nearly the entire breadth of the subcostal interspaces next succeeding; the upper two appear as a single spot, being scarcely divided by the intervening third superíor subcostal nervule. Still nearer the outer margin of the wing, and parallel to the row of spots just mentioned, are two subequal, rounded, obovate spots, slightly broader than long, the upper a little the larger, together occupying the entire breadth of the subcosto-median interspace, removed by less than twice their width from the row of spots previously mentioned; the fringe of the wing appears to be slightly darker than the ground-color.

Hind wings with a very large pale spot occupying the entire upper outer angle of the wing, reaching from the outer margin nearly halfway to the base, and from the costal margin to the upper median nervule; its basal margin is convex in the subcosto-median interspace, following what would perhaps naturally be the outer limits of the cell, while on either side of the lowest subcostal nervule the spot is separated from the outer margin of the wing by a narrow dark edging. On the irreg ular border which faces the median nervure, this pale spot emits three long, more or less sinuous tongues of pale color: one a very narrow, nearly straight, stripe or line along the margin itself, which only extends to the elongated upper median nervule, the breadth of the spot being less toward the margin than in the middle of the wing; a second, subparallel to the outer border, and therefore arcuate, as well as slightly sinuous, subequal, more or less broken into transverse spots, extending to the inner margin, and distant, beyond the middle median nervule,