of which this series is the last, has, however, been attained in demonstrating that modern classificators (I need not recapitulate their names) have been in error in intercalating the Swallowtails between the Blues and the Skippers. In thus doing violence to the characters of the insects, they have separated two allied groups by the interpolation of a third, not at all nearly related to either. I have further shown that we cannot bring in the Papilionides after the Hesperiadæ, since this course would break the sequence of character which allies the Hesperiades with the higher groups of the moths, the Sphingides, Saturniades, Bombycides (Agrotides). I do not assert a belief that the Hesperiades have sprung from common ancestors with these, but I find nothing in the neuration to render the idea improbable. And I am compelled to add, upon such evidence as is accessible to me, that I cannot say the same of the Papilionides, all connection of which with any of the above-mentioned groups appears to me to present a high degree of improbability. What discoveries await us in the future no one can say, and a naturalist can only come to a conclusion upon the material before him.

The notion that the Papilionides are generalized forms appears to me to be alike overstated. I lay especial stress upon the fact that the hind wings show but one internal vein, whereas all the Hesperiades show two at least. How the outer vein has been gotten rid of is plainly to be traced in the Papilionides, viz., in the same manner as the shortening and weakening of the inner vein has been gradually effected, from where it is stout and long, as in Ornithoptera, to where it becomes weaker, more curved and shorter, in Parnassius. This shortening of the vein is accompanied by a hollowing away of the tegument along the inner margin. A perfect parallel in this movement is sustained between the Papilionides and Saturniades. More than this: I believe I have found in the rounded and full inner margin of the secondaries in Ornithoptera an earlier stage of the hollowed margin of Papilio. For this reason, among others, I look upon Ornithoptera as being a relatively generalized form in the group. The ancestors of Papilio might have had two internal veins on secondaries, and in this they would have resembled the Hesperiades. From this point of view the inference is irresistible that we should commence our lists with the Papilionides. The specialization of the radius on fore wings keeps pace with the shortening of vein viii on hind wings in the Parnassiidæ.

In taking a fresh view of the evolution of the neuration, one is only too liable to overlook minor characters, or again to lay undue stress upon them. But, if I am correct in my major conclusions, if the theory in the main adequately explains the facts, such faults may be eliminated and forgiven. A well-founded division of the butterflies into two larger groups, of equal morphological value, must be considered as a gain to taxonomy, while it eases the study of the waste of butterflies to a hitherto unknown extent. I try to show, not only that the Parnassi-Papilionidæ belong together, but that all the rest of the butterflies are correctly associated in a second group. I have broken the Papilionides away, not only from their supposed affinities with the Whites, but from their incongruous placement with the Skippers also. In the former case I differ from Mr. Reuter, in the latter from Mr. Scudder. It seems to me that light is thus thrown upon a subject on which much has been written which must now appear purposeless, while the new course would enable us to carry with greater certainty the apparatus of classification along the road of phylogeny.

MATERIAL.

The authority for the generic names and types which I have here consulted, is Mr. Scudder's Historical Sketch of the Generic Names Proposed for Butterflies, Salem, 1875. I regret not to have been able to procure information or material of several generic types published since this date in the Papilionidæ. For specific determinations I am largely indebted to Dr. O. Staudinger and A. BangHaas, Blasewitz, Dresden. As far as possible I have studied the neuration of the female sex, since here the question of secondary sexual character is for the most part excluded. Where the male alone has been examined the sex is indicated. Following Mr. Scudder's work, I have omitted to cite the author to the generic type.

Classification of the Papilionides.

The characters are taken from the position of the veins and their condition. These characters are offered by the neurational movement peculiar to the Papilionides, of which they constitute existing stages, and are hence secondary and characters of specialization only. They are here amplified from my original communication

contained in Mittheilungen aus dem Roemer-Museum, Hildesheim, No. 8 (February, 1897).

On primaries vein iv, is radial or radially inclined,
vein iv, is cubital.

No traces of cubital cross-vein. . . PARNASSIIDE.
Radius four-branched, specialized Parnassiina.
Radius five-branched, generalized Zerynthiana.
Traces of cubital cross-vein . . . . . TEINOPALPIDÆ.

On primaries vein iv, from, at or near middle of
cross-vein; vein iv, from the lower outer cor-
ner of cell equidistant, or nearly so, between
iv, and iv.

A cubital cross-vein reaching vi or vii. PAPILIOnidÆ.

The radius is four-branched, specialized. There appear to be only two genera in this subfamily, since Doritis, with the type mnemosyne, does not seem distinct from Parnassius, with the type apollo. These two genera may be separated as follows:

Vein iv, from radius beyond the cell.
Vein iv, from cross-vein .

Parnassius apollo.

I. Parnassius.

2. Hypermnestra.

In all the species of Parnassius I have yet seen, vein iv, issues from radius, but since in the succeeding group this vein is fluid in Zerynthia, the character may not always hold. In both apollo and mnemosyne vein iii, arises a little before the point at which the cross-vein joins the radius, but, in Hypermnestra, at this point. I have been hitherto in error as to the absorption of i and ii at base in Parnassius. The lower vein is partially degenerate, but distinctly present in both apollo and mnemosyne. My figures must be corrected in this respect. This Papilionid feature is then retained throughout the group. It is the lower branch, the base of vein ii, which here seems to fade away. The upper branch, vein i, is united by a cross-vein, according to Comstock, with vein ii in Papilio. This short cross-vein appears then to become fused with and a continuous part of vein ii, its real base, while all that remains of it is the "præcostal spur." This may be the true termi

nation of vein i, the alternative being that the "præcostal spur is homologous with the shoulder veins of the Lachneida. But, in Papilio machaon, it is the base of vein i which clearly shrinks, vein ii being strong and continuous. In Parnassius the præcostal spur appears to emerge from vein ii, and the cross-vein has become absorbed.

Hypermnestra helios ().

Characterized by the extreme diminution of the humeral cell of secondaries, which is so reduced as at first to escape notice. The position of vein iii, shows an advance upon Parnassius, while in that of iv, it lags behind its ally. Else the neuration of the two

generally agrees.

Subfam. 2. Zerynthiana.

This subfamily must take its name from Zerynthia Ochs., 1816, because Thais Fabr., 1807, which is the older title for the same type, is preoccupied (Scudder, l. c., 279). Not observing this, I originally used for it the name Thaidina. No other author, so far as I know, had proposed this division of the Parnassiidæ, which in itself seems a natural one. The Zerynthianæ are intermediate, in the specialization of the wings, between the Parnassiidæ and the Papilionidæ.

Radius five-branched, generalized:

Internal vein (viii) of hind wings relatively short, not

reaching anal angle .....

Internal vein (viii) of hind wings reaching anal angle:
Hind wings without special prolongation of

Archon.

A character of specialization is found in the shortened and bent internal vein of secondaries which follows the inward PROC. AMER. PHILOS. SOC. XXXVIII. 159. B. PRINTED JULY 7. 1899.

curve of the margin, as in Parnassius and the Saturniades. This character is probably more recently acquired. The neuration else shares the main features of that of its group. The humeral cell is narrow and reduced, the lower vein (base of ii) being weaker. On primaries vein iv, springs from the cross-vein. Although the pallor of the ground color of the wings suggests the preceding subfamily, the pattern of ornamentation (on the value of which for taxonomical purposes I have been insisting upon for more than thirty years past) is distinctly Zerynthian. The discal blotches are intermediate between the bands of Zerynthia and the spots of the Parnassians and show the origin of the latter. The subterminal red-marked fasciæ are like Zerynthia. Archon may be regarded as a generalized form of Parnassius, showing the origin of the latter from Zerynthia-like ancestry.

Zerynthia polyxena.

The material examined shows that vein iv, is in a fluid state. In a female this vein leaves the extreme upper corner of the cell. In the male it has passed beyond the cell and issues from the radius. In a female of rumina, figured by me (Mittheilung a. d. Roemer-Museum, No. 8, Taf. 1, Fig. 2), it leaves the cross-vein. Not only this vein varies in position, but also vein iii,, which is further removed outwardly in the specimen in which vein iv, leaves the radius. Sufficient material has not been examined to gauge the extent of these variations, but it seems unlikely that they are sexual. As compared with Archon, the internal vein of hind wings is less bent and reaches the anal angle, the humeral cell is a little wider, else the neuration generally agrees. Traces of a splitting of vein ii at base appear on fore wings.

Luehdorfia puziloi.

The neuration of the female examined agrees exactly with that of Zerynthia rumina. Except that the margin of secondaries is more uneven, and that vein iv, is produced into a short "tail," no differences whatever have been noted. Vein iv, leaves the cross-vein. The forms in which vein iv, is still attached to the cross-vein must be regarded as more generalized than those in which it has ascended the radius and is thrown off beyond the cell.