the wings. It is quite impossible to place it, except by merest conjecture.

The last instance is to be found in the paper by Brauer, Redtenbacher, and Ganglbaur (Mém. Acad. St. Petersb., (7), xxxvi, 1889) on the Jurassic insects of eastern Siberia; these authors mention among the "Dubiosa," a dipterous pupa "somewhat resembling that of Ptychoptera."

From these unsatisfactory data we may conclude that Tipulinæ and perhaps Limnobinæ have probably been found as far back as the Jura, but that further details regarding specimens will need to be published before the evidence is satisfactory.

VI. FAMILY TIPULIDE.

The two subfamilies of Tipulidæ may be separated by means of the structure of the wings (often the only characteristic part remaining in fair preservation among the fossils) in the following

manner:

Auxiliary vein usually ending in the costa and connected by a cross vein with the first longitudinal vein; the latter ends in the costa without aiding in the formation of a trapezoidal cell; last posterior cell in broad contact with the discal cell...... Limnobina.

Auxiliary vein ending in the first longitudinal vein by abruptly curving down to it, but otherwise free from it; first longitudinal vein, by an apical incurvation and the emission at its base of an oblique costal cross vein, enclosing a trapezoidal cell at the distal extremity of the stigma; last posterior cell touching the discal cell at only one corner Tipulina.

In the enumeration of the specimens at the end of the following specific descriptions the numbers of the obverse and reverse of the same specimen are always connected by "and" without any intervening comma, and this typographical method is employed only in expressing this relation.

VII. THE SUBFAMILY LIMNOBINE.

For the sake merely of simplicity, I use this term to include all the Tipulidæ brevipalpi of authors. There seems to be a greater diversity of structure among them than among the Tipulinæ proper, and they have been divided by Osten Sacken into eight groups,

which may be regarded as tribes, the relative importance of which in recent and ancient times has been pointed out in a preceding table, from which it will appear that every one of them has been recognized among the fossils.

The Limnobini take precedence as they do among modern types, while the Rhamphidini (in Europe in amber only), the Eriopterini, and Limnophilini follow in numbers, the remaining groups being of least importance and three of them altogether lacking in American deposits; while the Cylindrotomini, the only remaining tribe appearing in America, is lacking in the European tertiaries. The Anisomerini are represented in Europe by three species of Eriocera in amber, the Amalopini by four species of Ula in amber, and the Ptychopterini by a single species of Idioplasta in the same and by a Ptychoptera at Krottensee. It thus appears that with the exception of the Ptychopterini all the tribes represented in the European rock deposits occur also in America, while America is also well represented in the tribes Rhamphidini and Cylindrotomini. American genus Pronophlebia cannot yet be placed.

The

Especially difficult of determination among the Limnobini has been the position or the absence of the subcostal and marginal cross veins, which play so important a part in the arrangement and distinction of the genera. It is by no means impossible that in some instances I may have erred in my interpretation of marks upon the stones, but I have endeavored to give all points a rigid scrutiny. It is certainly here that errors are most likely to have been made.

Nearly one hundred additional specimens from Florissant, more or less imperfect, but certainly belonging to this subfamily, await study; and I may add that there is a specimen in the collection of the U. S. Geological Survey (No. 1470) which represents an interesting new genus falling near Atarba with distinct tibial spurs, but which I refrain from characterizing here, as I can give now no illustration of it.

As the number of genera in no one of the tribes exceeds three, I have thought it best to include all the genera of the subfamily in one table, as follows:

Table of the Genera of American Fossil Limnobinæ.

A1. Only a single submarginal cell.

b'. The normal first posterior cell, lying (in forms having but one submarginal cell) between the second and third longitudinal veins, closed, forming at base a supernumerary discal cell......

b2. First posterior cell open throughout.

cl. Five posterior cells....

c2. Four posterior cells.

Cyttaromyia.

Oryctogma.

d'. The first longitudinal vein ends in the second longitudinal vein. e1. Discal cell closed; submarginal much longer than first

e. Discal cell open; submarginal scarcely longer than first
posterior cell.. . . . . .

d2. The first longitudinal vein ends in the costa.
el. A marginal cross vein.

Spiladomyia.

f1. Marginal cell excessively long, much exceeding in length the breadth of the wing.. ......Limnocema. f2. Marginal cell normal, not so long as breadth of cell (applicable to fossil species only).... . . . . . . . . Antocha. Rhamphidia.

e2. No marginal cross vein...

A. Two submarginal cells.

b'. Third longitudinal vein arising at normal distance from the second.
c'. First submarginal cell not, or hardly more than, half as long as the
...Gonomyia.

second

c2. First submarginal cell much more than half as long as the second. d'. Tibia without spurs at the tip.

e1. Auxiliary vein ending at a distance beyond the origin of

the second longitudinal vein considerably more than equal to the breadth of the wing; third posterior cell petiolate.... Cladoneura.

e2. Auxiliary vein (in the fossil species) ending at a distance beyond the origin of the second longitudinal vein less than, sometimes only about half, the breadth of the wing; second posterior cell petiolate...... Cladura. Limnophila.

d. Tibia with spurs at the tip..

b2. Third longitudinal vein arising from the second at a very short distance beyond the base of the latter...... Pronophlebia.

Tribe CYLINDROTOMINI.

This tribe is placed first in the series instead of near the end really because the arrangement of the table just given requires the early appearance of the two genera recognized in the American

rocks. Its place in the arrangement given by Osten Sacken is rather at the other end of the series in nearer proximity to the Tipulina; but it may be noticed that in some of the features in the neuration of the wing, as in the arrangement in the vicinity of the stigma, in which it approaches the Tipulinæ, it also shows most resemblance to the Limnobini.

Although Loew referred some amber species to Cylindrotoma, Osten Sacken, who has since examined them, says they are all Limnophile, so that the species described below, six species of two genera, both extinct, are the only ones known in a fossil state.

CYTTAROMYIA Scudder.

Cyttaromyia Scudd., Bull. U. S. Geol. Geogr. Surv. Terr., iii, 751 (1877); Tert. Ins. N. A., 574-575 (1891).

This genus was founded, in 1877, upon a specimen showing the apical half of a single wing, somewhat distorted by folding, and rather obscurely preserved, found by Denton among the first known tertiary insects of North America, on the lower White River of Colorado and Utah. A number of specimens and several species of the same genus have since been obtained by me from the same spot, while exploring for the U. S. Geological Survey, but no further specimens of the same species. The beds at Florissant have also yielded several species of the genus and permit a more accurate and complete account of the generic characteristics. These specimens show that my original description was faulty in its interpretation of the structural elements of the wing: the cells lying beyond the "secondary discal cell" were wrongly regarded as submarginal cells, for they belong to the "posterior" series, and all the errors of statement followed from this wrong interpretation; but the neuration is none the less remarkable, and, so far as I have been able to discover, unique.

The wings are very long and slender, four or more times as long as broad. The auxiliary vein ends in the costa, without any sudden curve, at the beginning of the stigma. The first longitudinal vein, by very gradually curving downwards, ends in the second, which curves upward to meet it, forming a long and slender marginal cell ; there is neither subcostal nor marginal cross vein. The second longitudinal vein arises near the middle of the wing, varying in the species, is generally considerably arcuate at the base, the præfurca

considerably shorter than the rest of the vein, which terminates above the apex of the wing. The third longitudinal vein arises. from the second a little before the tip of the auxiliary vein and is met by the short cross vein at a distance from its origin equal to the length of the short cross vein. The space between the third longitudinal vein and the upper branch of the fourth longitudinal vein, normally-and so far as I am aware invariably-open in all Limnobinæ, is here closed about half way to the tip of the wing (varying in different species) by a cross vein, from which springs an intercalary vein, thus doubling the upper posterior cell at the apex and forming of its basal portion a supernumerary discal cell, essentially a counterpart of the normal discal cell and overlying it; it would seem to be formed by a mesial forking of the third longitudinal vein, and the base of the fork then connected by a cross vein to the uppermost branch of the fourth longitudinal vein. Both discal cells are usually very elongate (least so in the species upon which the genus was founded), the upper, or supernumerary, usually the longer and narrower, though they are subequal in length. There are five posterior cells and the great cross vein strikes the fourth longitudinal vein at the discal cell close to the base of the latter. The fifth longitudinal vein is very gently arcuate beyond this cross vein, while the sixth and seventh are straight throughout, the latter, however, arcuate at the extreme tip and almost reaching the middle of the wing. The legs are long and slender and the tibia without spurs at the tip.

This genus was well developed in the American oligocene, especially in the White River basin, where it seems to include the larger number of species of Limnobinæ. I leave their description, however, to another occasion and characterize at this time only the species found at Florissant, none of which appear to be identical with those from the White River basin. The genus does not appear to be found among the European fossil Limnobinæ heretofore published.

This genus, it seems to me with little doubt, must fall into the Cylindrotomini, although the tibiæ lack spurs. I am forced to this conclusion by the close resemblance of the neuration to that of Cylindrotoma and Liogma, notwithstanding the striking differences. Especially the formation of the marginal cell is essentially the same, while the absence of the anterior cross veins and the general behavior of the auxiliary vein sustain this view. It seems to me very