iii, has not moved at all from the original position within the cell. It is thus more generalized than either of its associates. From Callidryas-like ancestors may rather have sprung the curious form Colias rhamni, belonging to the genus Rhodocera, or again Gonepteryx of authors, but, according to Scudder, wrongly so referred.

In this genus in which the wings have probably been transformed by mimicry to copy the shape of a leaf, vein iii, keeps its original place of exit before the cross-vein; consequently it cannot have been derived from forms among which this vein was shifting. It must have been thrown off before Meganostoma-like forms appeared and probably Callidryas represents very nearly its direct line of descent. It is more specialized than Callidryas, not only in the remarkable shape of its wings, but because it has lost by absorption vein i of hind wings, the "præcostal spur" of some writers, which is still retained by Callidryas. The specialization runs in this respect parallel with the branch Eurymus-Meganostoma. In the latter genus a remainder of the vanishing vein i is to be seen which has become lost in Eurymus. The specialization on this phylogenetic line of the typical "Yellows" has not apparently developed a three-branched descendant, at least in the holarctic fauna, and so far as my studies now go. Nor have I yet found the five-branched generalized form, which might represent its more remote ancestry.

Turning to the next line of non-typical "Yellows," the Euremini, we find the three-branched descendant reached in Nathalis. This form has evidently emerged from four-branched ancestors, represented in America by Eurema and Terias, forms which so very nearly agree that I am even at a loss to distinguish them. I make out vein viii of primaries to be quite distinct and relatively strong in Terias, and conclude this may be the subspecialized form of the two. I cannot now connect this line with the typical "Yellows," and its ancestry must be apparently sought for in more southern regions.

3+4

3+4+5

We will now take up the "typical Whites." The three-branched condition is attained by Mancipium brassica. Here the little remaining branchlet iii,+ of Pieris has at last vanished. But the vein iii, in which it has lost itself is a little bent at this place. I should not wonder if examples of the "large Cabbage White" might be found retaining some trace of this vanished veinlet. In Pieris I have examined rape and napi, while Prof. Comstock's beautiful figure of protodice appears to agree (Evolution and Taxonomy, Pl. ii, Fig. 3).

PROC. AMER. PHILOS. SOC. XXXVII. 157. c. PRINTED MAY 18, 1898.

In all these the little vein iii,, remains distinct and has not been lost. Evidently Pieris represents the ancestral form of Mancipium and has perhaps been thrown off before the specialization of Pieris has progressed so far. Notwithstanding the similarity of the ornamentation I am not sure that P. rape is on the direct line of descent. As between rape and napi I incline to consider the latter at present the more specialized. Aporia cratagi is evidently a more generalized form, standing a little apart. Vein iii, is quite a long furcation, and measures its distance from Pieris. The skeleton of the wing is more powerfully built and vein viii of primaries stronger than in Pieris, in which it seems little better than a scar. The gradation by which this vein, which appears usually like a loop, strap or support to vii at the base, passes into obliteration is so entire that the exact statement of its condition is often difficult either to correctly grasp or record. The tubular character disappears by minute gradations; the "scar" aspect and the "tubular" shape are easy to detect, but where the one commences and the other ends it is often hard for me to say. In the holarctic fauna

I do not find any form to represent the probably actual fivebranched condition of Pieris, but here several types are wanting to me which I should like to have examined. In the genealogical tree of the holarctic butterflies the more generalized Anthocharini must take the place of the common five-branched ancestor of the whole Pierinæ. But this seems to me to stand upon a separate immediate phylogenetic line of its own, notwithstanding some common features of color and marking. With this Anthocharid line we must now in concluding concern ourselves.

Among the Anthocharini, or what we may call the "non-typical Whites," we have, in Pontia daplidice, the attainment of the threebranched condition. This butterfly appears to me to have no immediate connection with the "typical Whites," but to be a descendant of Anthocharid ancestry. It is true that Mr. Meyrick refers it without comment to the genus Pieris (Handbook, 353), but it is also true that Mr. Meyrick, in the same publication, precedes Pieris by Leptidia (Leucophasia) and this again by Euchloe, and, to make the mixture complete, Gonepteryx (Colias). This sort of work appears to me to prove that Mr. Meyrick's studies are not yet sufficiently correlated" with the actual facts of structure. If, indeed, the picture which Mr. Meyrick has received of the neuration at all resembles the figures with which his publications are adorned,

no proper judgment could, in my opinion, be formed upon it, and this would perhaps account in part for the seemingly extraordinarily unnatural sequences adopted by him.

The coincidence between the neuration of Pontia daplidice and that of Mancipium brassicæ is so great, that I am at a loss to give good characters of distinction. But showing, as I do, that the three-branched character of the Pierid primary wing is attained upon obviously distinct lines (e. g., Euremini), this coincidence will not of itself determine the phylogeny. The shape of the wings and the pattern of ornamentation of Pontia are both Anthocharid. It is not conceivable how either could have been derived from Pieris and the "typical Whites." We should have to suppose that the four-branched Pieris threw off the three-branched Mancipium and also the three-branched Pontia; an inference which, considering the want of any near resemblance in the shape and pattern of the wings between the two descendants, or between one of these (Pontia) and the supposed parent stem, must be set down as untenable. More than this, we have in Pontia a similar secondary sexual character in the shape and extent of the wings to that we find in the Anthocharini, no trace of which is evident in Pieris or Mancipium. This character has evidently been retained by Pontia, through an ancestry of which I find one existing representative form, extending back to the five-branched representative of a remote phase which is brought before us now in Anthocharis and Euchloe. I believe that these facts show, that the phylogenetic position heretofore assigned to Pontia, is a discordant one and should be corrected. We may now leave Pontia and look over the more generalized and the typical Anthocharini with their five-branched radius.

Mr. Scudder (Historical Sketch, 113) says, regarding the use of the generic term Anthocharis : "As Euchloe must be used for the European species, genutia should be considered the type of this

This would seem to imply that all the European species. were generically distinct from all the American and that the latter should alone be referred to Anthocharis. I do not agree with this statement at all, and I can show grounds for referring American species, with orange blotch in the male, to Euchloe, and for considering that the white species of both continents are slightly more specialized and might be kept under the separate title of Anthocharis.

I regret not to have genutia to examine and I use Antho

66

charis for the type belemia, which is, perhaps, identical. The subjective question of whether there are two genera" to be considered is not of any importance to me at all. I recognize two five-branched types: the one specialized, which I seem warranted in calling Anthocharis under Boisduval's original use of that term; the other, relatively generalized, which I call Euchloe, with the type given by Mr. Scudder of cardamines.

The white Anthocharids differ from the type of Euchloe cardamines in that vein iii, has moved from the original position and is given off opposite, or even beyond the cross-vein. In A. ausonides, which is slightly the more specialized of the three examined, it has even passed the extremity of the cell for a considerable distance. Therefore the specialization runs here upon the same line as in the case of Meganostoma and Eurymus. The generic title Anthocharis should have, I believe, the type belemia, in case genutia does not share these essential characters and is not, in the sense here proposed, an Anthocharis. It is clear from the above citation from the Historical Sketch, that Mr. Scudder has misapprehended the state of affairs in this group; for I am quite unable to find any neurational differences between the North American E. stella and the European type of Euchloe. In both insects vein iii, retains its original position above the cell. And the chances seem to be that this will be the case with most of the species, carrying an orange blotch on the male primary, irrespective of locality. In any case, that which interests us here especially is the development of a specializing movement tending generally in the direction of a reduction in the number of the radial branches, but here taking a special and, looking through the day butterflies, perhaps an unusual direction. I find it, besides in these two instances, in the Pieridæ, in Euptoieta, Melitæa, Euphydryas, Araschnia and Heliconius. But when we examine Pontia, we find that, although the fivebranched radius has become a three-branched, still vein iii, has not changed its place. The reduction has been effected by other means than the shifting of iii, in the direction of the apex of the wing. Into the details of the physiological process of absorption I cannot now enter, sufficient for my present purpose is the fact, that Pontia represents a clean descent from Euchloe-like forms and that it has not passed through Anthocharid-like forms upon its way. The absorption of iii, has proceeded to a varying extent in these species of Anthocharis. The little branch remaining has

become very short indeed in A. ausonides. A. belemia would be the most generalized form, since iii, has not, or hardly, passed the cross-vein. In both belia and ausonides this halting place has been passed by. But in Tetracharis (n. g.) cethura Feld., sp., we have a four-branched Euchloe; one which represents an intermediate stage between the five-branched Euchloe and the three-branched Pontia. Tetracharis may be represented also by other species, since I have not been able to examine all the forms of the Anthocharini.

This survey of the Pierina has shown us that the Anthocharini represent the most generalized forms apparently in the holarctic fauna, and that they are probably the survivors, not on the direct line, of a former five-branched condition of the family. There remains one more five-branched form to examine: Leptidia (Leucophasia), but this presents so strange a neurational pattern, that it must have come into its present company by a vastly different route. Of its peculiar type it may be a specialized form, although, in comparison with the Pierinæ, it seems generalized. Its white color has come to it, I think, subsequently; as to its origin― unde et quomodo-I have no idea which is not imaginary.

To touch finally another aspect of our subject-a study of the dynamics of the butterfly wing has been somewhat neglected. From the details of the changes in the position of the veins, it may be concluded that the movements have a mechanical cause. Since this inquiry belongs to a department of direct observation upon which we can obtain absolute knowledge, without employing reconstructive methods, it may be painfully followed up, in field and cabinet, until the subject becomes clear. The butterflies certainly owe a part of their attractiveness to the fact of their seasonal appearance. They recur at a certain niveau in the biological circle, thus relieving the mind through their plain testimony from doubting that the principle of existence is succession.