Acraspis quercushirta (agamic)

Cynips heldae (agamic)

Cynips mirabilis (agamic)

Acraspis gemula (sexgen)

Cynips quercusechinus (agamic)

Cynips quercusechinus (sexgen)

Paracraspis guadaloupensis (agamic)

Paracraspis insolens (agamic)

Paracraspis patelloides (agamic)

Xanthoteras teres (agamic)

Xanthoteras pulchellum (sexgen)

Cynips multipunctata (agamic)

Cynips conspicuus (agamic)

Cynips maculosus (agamic)

Antron plumbeum (agamic)

Philonix fulvicollis (agamic)

Philonix gigas (agamic)

Philonix lanaeglobuli (agamic)

Atrusca simulatrix (agamic)

Atrusca dugesi (agamic)

Atrusca brevipennata (agamic)

Atrusca pupoides (agamic)

Atrusca bella (agamic)

Atrusca cava (agamic)

Atrusca quercuscentricola (agamic)

Cynips arida (agamic)

Cynips mellea (rydbergiana) (agamic)

Atrusca unica (agamic)

Xanthoteras clavuloides (agamic)

Antron douglasii (agamic)

Sphaeroteras melleum (agamic)

Sphaeroteras carolina (agamic)

Cynips conica (agamic)

Antron quercusnubila (agamic)

Cynips incomptum (agamic)

Antron acraspiformis (agamic)

Acraspis alaria (agamic)

Acraspis villosa (agamic)

Acraspis echini (agamic)

Acraspis prinoides (agamic)

Acraspis erinacei (sexgen)

Acraspis pezomachoides (agamic)

Acraspis erinacei (agamic)

Atrusca congesta (agamic)

Atrusca clivorum (agamic)

Atrusca strians (agamic)

Cynips hirta variety hirta, agamic form

Cynips q hirta
Cynips hirta
Biorhiza hirta
Biorrhiza hirta
Philonix hirta
Philonyx hirta
Trichoteras quercus-hirtum
Philonix pezomachoides (error)
Acraspis hirta

GALL. — Quite spheroidal, moderately large, up to 6.0 mm. in diameter; the faceted surface without projecting spines and consequently smooth in appearance; on leaves of Quercus montana (= Q. Prinus = Q. monticola of authors), Q. Michauxii, and probably related chestnut oaks. 

RANGE. — Toronto (Brodie in U.S. Nat. Mus.). Massachusetts: Worcester? (gall, M. T. Thompson in Boston Soc.). Connecticut: Waterbury (types; Bassett coll.). New York: West Point (Bassett in Mus. Comp. Zook). Storm King (gall, M. D. Leonard acc. Weld 1926). Highbridge (gall, Amer. Mus. coll.). Huntington (J. C. Bridewell in U.S. Nat. Mus.). Ithaca (gall, acc. Weld 1926). New Jersey: Fort Lee (gall, acc. Beutenmiiller 1910). Richland (gall, Kinsey coll.). Pennsylvania: Washingtonville (gall, E. C. Zeliff in Kinsey coll.). D.C. (acc. Beutenmiiller 1909). Virginia: Bluemont (gall, acc. Weld 1926). Blue Ridge Mountains near Natural Bridge Station (gall, Kinsey coll.). Winchester (Kinsey coll.). North Carolina: Asheville (Ashmead in U.S. Nat. Mus.). Kentucky: Livingston (gall, Q. Michauxii, Kinsey coll.). Indiana: Nashville ( Q . Michauxii, Kinsey coll.). Michigan: Agricultural College (Gillette? in U.S. Nat. Mus.). Kansas: state (acc. Bridewell 1899). St. George (in U.S. Nat. Mus.). 

Probably confined to the chestnut oaks in a more northeastern area of the United States and Canada, at least from Connecticut to Indiana, perhaps westward to Missouri. The Atlantic Coastal Plain, the more southern, and the most western records given above need further examination when the bisexual adults become known. 

This is the chestnut oak variety of the species, originally described from Q. Prinus and now recorded from Q. Michauxii. Most of the Cynipidae do not make distinctions between the several species of Middle-Western chestnut oaks. Beutenmuller’s 1900 record (in Ashmead in Smith (1900:548)), and Viereck’s 1916 records of Q. ilicifolia as the host are certainly errors, while the Q. macrocarpa records should apply to variety macrescens. 

Mature galls of this variety in the American Museum are dated September 11 (1904). Bassett collected galls in October, finding that the insects had begun to eat passages thru the walls of the galls on October 20, altho the first emergence occurred on November 29. Galls I collected in Virginia on October 18 (1919) showed exit holes, but whether of gall makers, inquilines, or parasites, I cannot determine. An adult gall maker was alive in a gall I collected on October 31 (in 1920) at Nashville, Indiana. I have bred adults on De- cember 20 and January 5 (1928). Bassett (1870) notes that the peculiar, acid odor characteristic of several short-winged cynipids is found in hirta. 

This insect is very close to variety macrescens from which it appears to be distinguished only by the longer wings and by the large, spherical gall. From variety undulata, hirta appears to differ in having the antennae darker at base, the lateral lines absent, and the mesopleura more hairy. But altho I have examined the holotypes of all of these insects, and made direct comparisons of paratypes of the three, I am not certain that I could determine additional insect material without locality records or galls. On the other hand, the galls of the three are so distinct that, with the additional host and geographic peculiarities of each, the varieties should be maintained as distinct. The much reduced thorax and the mere stubs of wings which these forms of Acraspis show offer very few characters for taxonomic determinations. The characters given for these agamic insects by Weld (1922) in his key to Acraspis did not seem to me to separate the paratypes of the several varieties when I studied them at Washington a couple of years ago. 

[Cynips hirta variety undulata](https://www.biodiversitylibrary.org/page/53516882#page/420/mode/1up)
agamic form

GALL. — Of moderate size, more or less ellipsoid, the faceted surface fairly smooth in appearance; on leaves of Quercus Gambelii and its varieties (NOT on Q. undulata!) . 

RANGE. — Colorado: Manitou (types, Gillette coll.). New Mexico: Kingston (galls, Kinsey coll.). Probably confined to Q. Gambelii and its varieties, in a Southern Rocky Mountain area of southern Colorado and more northern New Mexico. Figure 70. 

There appear to be no data on this insect beyond Gillette’s record of the type collection. His galls, collected on June 30, contained mature adults that ran about when cut out. It is difficult to understand summer maturity, in the light of the data for related varieties of the species, unless these were survivors of the previous year’s growth that had failed for some reason to emerge in the late fall or the ( winter. Gillette stated that from June 30 to November 19 none of the insects emerged, tho a number were cut out of the galls. His further statement that the galls were kept in a warm room after the last of September may offer one more explanation of the failure to obtain normal emergence. 

I have a single gall from Kingston, New Mexico, which might well represent the present variety. The insect had emerged when I collected this gall on December 27 (1919). We must await insect material before we can be certain that undulata extends as far south as Kingston. West of the Rockies, in Utah, there occurs variety packorum with its distinctly blunt-spiny and spherical gall. True undulata does not occur east of Colorado. 

Variety undulata, so named because of Gillette’s confusion of the names of the Rocky Mountain oaks, really occurs on Quercus Gambelii and the closely related varieties or species of oaks. Fragments of leaves with the type material confirm our interpretation of the host. 

Gillette described undulata by comparing it with macrescens (= Bassett’s macrocarpae) . Beutenmuller, examining paratypes of the two in 1907, failed to find these differences and considered the names synonymous. I agree with Beutenmuller that none of the characters noted by Gillette are reasonably appreciable or constant enough among the individuals of the type series to warrant taxonomic recognition. On the other hand, Beutenmuller, perhaps because he worked with a hand lens which would not show the diagnostic characters on the much reduced thoraces of these insects, failed to ob- serve several distinctions which seem to me to be of significance. Other authors have followed Beutenmuller’s synonomy without having examined material. Weld, whose 1922 revision of Acraspis was critical as far as it went, did have types of undulata but none of macrescens for comparison. 

Undulata differs from macrescens chiefly in having more distinct lateral lines, more naked and shining areas on the mesopleuron, and wings that are (Gillette’s original description to the contrary) a little longer than in macrescens. My measurements are made from camera lucida drawings of type material. The discovery of the bisexual generations of these insects may (or may not) offer further data for making dis- tinctions, but the characters now recorded seem to warrant the recognition of these varieties with distinct geographic ranges and distinct hosts. 

[Cynips hirta variety packorum, new variety](https://www.biodiversitylibrary.org/page/53516882#page/422/mode/1up)
agamic form

GALL. — Nearly spheroidal, moderately large, up to 6.0 mm. in diameter, with the faceted surface set with very short, cone-like projections, the gall consequently rough and short-spiny in appearance; on leaves of Quercus utahensis [synonym of gambelii]. Figures 317-318. 

RANGE. — Utah: Wales and Central in Sevier County (galls, Hagen in Kinsey coll.). Salt Lake City (thru C. T. Dodds, in Kinsey coll.). Payson, Bountiful (types), Centerville, Farmington, Layton, and Willard (B. and H. J. Pack in Kinsey coll,). Probably confined to an area west of the Continental Divide. 

This insect is best distinguished from variety undulata and more eastern varieties of the species by its more generally black color (but more rufous legs), by its slightly longer wings, and particularly by its rougher and even spiny, spherical gall. The gall suggests a small gall of an agamic form of Cynips gemmula rather than the smooth, faceted galls of the other varieties of hirta. Undulata is probably confined to an area in southern Colorado and northern New Mexico east of the Continental Divide. The cynipid fauna of Utah is usually related to but never identical with this fauna from the eastern side of the Rockies. 

How far packorum ranges in Utah is not determinable from our present insect material. Many Cynipidae in that state have more northern and more southern varieties whose ranges meet somewhere between Provo and Brigham. Most of our insect collections of packorum were made in this critical territory, but I cannot recognize two varieties in this material. 

Most of our galls of this variety were gathered early in September, at which time they were still young enough to be touched with a rosy pink, but old enough to breed with fair success. Adult insects emerged (out-of-doors, at Bloomington, Indiana) on November 18 and December 10 and 20. 

The type material and fine series of this variety from several other localities were collected by Dr. H. J. Pack, Entomologist of the Utah Agricultural Experiment Station, and his daughter Bessie Pack, for whom I am naming this insect. These collectors have sent me large series well representing the cynipid fauna of northeastern Utah, and they thus will contribute materially to our further revisions of the cynipid genera and our ultimate understanding of the distribution problems involved in that little-explored area west of the Continental Divide.

[Cynips hirta variety obtrectans, new variety](https://www.biodiversitylibrary.org/page/53516882#page/424/mode/1up)
agamic form

GALL. — More ellipsoid than spheroidal; up to 5.5 mm. in length; the faceted surface distinctly smooth in appearance; on leaves of Quercus macrocarpa. 

RANGE. — Texas: Austin (types; Patterson coll.). Oklahoma: Pawnee. (Kinsey coll.). Kansas: Winfield and 10 miles southeast of Winfield (galls, R. Voris in Kinsey coll.). Probably confined to eastern Texas and Oklahoma, or perhaps extending due east from that area as far as the host, Q. macrocarpa, occurs. Figure 70. 

This is the more southern of the burr oak varieties of hirta. The present insect is distinct enough from the more northern varieties macrescens , scelesta , and opima, to allow certain determination. I have only one immature adult from the mate- rial I collected in 1920 at Pawnee, Oklahoma, but it shows the coloration and the lengthened wings of the types from Austin, Texas, and I take it that the range of the variety is largely east and northeast from Austin. Dr. Ralph Voris has collected young but full-sized galls, of this species but doubtfully of this variety, in southeastern Kansas as early as June 6, 1927. At that time the larvae were still microscopic in size. Dr. Patterson collected the type galls in the fall of 1921, cutting two mature (one not fully pigmented) and live adults from the galls on October 26 of the same year. 

[Cynips hirta variety opima, new variety](https://www.biodiversitylibrary.org/page/53516882#page/425/mode/1up)
agamic form

GALL. — Nearly spheroidal, slightly ellipsoid; up to 3.5 mm. in diameter; the surface fairly smooth; on leaves of Quercus macrocarpa. 

RANGE. — Michigan: Wayland. Illinois: Seneca (types, Kinsey coll.). Charleston and Green Valley in Tazewell County (Kinsey coll.). Indiana: Mongo, Delphi, Romney, Morocco, Rogers in Pike County, and Aurora (Kinsey coll.). Centering in more southern Indiana, Illinois, and possibly south- westward into the Ozarks; found northward into Michigan only within populations of the hybrid variety macrescens. Figure 71. 

This is the more southern variety of hirta in the northern Middle West, occurring in Indiana and Illinois where we might expect an Ozark variety, altho we do not have collections of the species from localities within the Ozarks proper. This appears to be the form which, hybridizing with the northern scelesta, gave rise to the hybrid variety macrescens, and segre- gates of opima consequently occur on occasion as far north as macrescens extends. Quite different are the nearly pure populations of opima in Illinois and southern Indiana. Both insects and galls of typical opima are about as distinct from scelesta as anything in the present species ; but the occurrence of the hybrid macrescens in between provides such a gradation from opima to scelesta that determinations of anything less than large series of these insects cannot be more than approximations to natural interpretations. 

Our emergence records for this insect are November 4, 16, 19, and 20, and December 1, 4, 8, 12, 15, 18, 19, 20, 22, and 28. 

[Cynips hirta variety scelesta, new variety](https://www.biodiversitylibrary.org/page/53516882#page/426/mode/1up)
agamic form

GALL. — Always elongate ellipsoid; up to 4.0 mm. in length; the faceted surface quite smooth in appearance; on leaves of Quercus macrocarpa. 

RANGE. — Michigan: Wayland (types, Kinsey coll.) Big Rapids and Tekonshah (Kinsey coll.). Indiana: Morocco, Romney, Delphi, and Mongo (Kinsey coll.). Illinois: Seneca (Kinsey coll.). Centering on the very northern rim of the range of Q. macrocarpa, probably from New York to Minnesota. Remnants of the species scattered southward (by the Pleistocene glaciation) thruout Indiana and Illinois. Figure 72. 

This is the sub-Canadian variety of the species; to be expected everywhere on the most northern rim of the range of Quercus macrocarpa. The extreme form of the insect is distinguished more readily than any of the other eastern varieties of hirta. On the other hand, the hybrid of scelesta and opiwu is the widespread macrescens, and thruout more southern Michigan every transition occurs between scelesta and ma- crescens, while further south scelesta appears here and there as a segregate in the hybrid population. Since the host, Q. macrocarpa, does not extend into the true Canadian zone, it is possible that a pure population of scelesta has not been maintained north of the area of hybridization. Our emergence dates for this insect are November 20 and December 1, 4, 8, 12, 15, and 22. 

[Cynips hirta variety macrescens, new name](https://www.biodiversitylibrary.org/page/53516882#page/427/mode/1up)
agamic form

GALL. — Usually elongate ellipsoid; up to 4.0 mm. in length; the faceted surface hardly with projecting tips and consequently fairly smooth in appearance; on leaves of Quercus macrocarpa. Figures 316, 329. 

RANGE. — Quebec: Dows Lake (gall, Rowland in Gray Herb.). Ontario: province (acc. Jarvis 1907). Toronto (Brodie in U.S. Nat. Mus.). New York: Medina (acc. Weld 1926). St. Lawrence County (acc. Bassett 1890). Manorville (?) and Farmingdale (?) (acc. Weld 1928) . Ohio: Rockport (types; Bassett). Indiana: Delphi, Steubenville, Mongo, Romney, Morocco, Aurora, Clinton, and Rogers in Pike County (Kinsey coll.). Auburn (gall, Kin- sey coll.). Roachdale (gall, G. Blaydes in Kinsey coll.). Michigan: Lansing (acc. Gillette). Tekonsha, Wayland, Owosso, and Three Rivers (Kinsey coll.). Illinois: Evanston, Winnetka, Libertyville, Glen Ellyn, and Moline (acc. Weld 1926). Fountaindale (acc. Weld 1926; also gall, Bebb in Gray Herb.). Charleston, Green Valley in Tazewell County, and Seneca (Kinsey coll.). Minnesota: Hastings Road (gall, Ruggles acc. Felt 1921). Min- neapolis (galls, J. S. Benner in Kinsey coll.). Iowa: Ames (Gillette in 111. Lab. and U.S. Nat. Mus.). “Keo- sauqua” (?) (gall, MacDonald in Gray Herb.). Corinth (gall, C. Bar- racks acc. Weld 1926). Kansas: Cedar Point and Holton (galls, acc. Weld 1926). Probably confined to a more northern range of the host, Q. macro- carpa, from Quebec to Minnesota and Kansas. Replaced on the extreme northern limits of the range of the host by variety scelesta. The more western records need re-determination; the published Colorado records apply to variety undulata. Figure 70. 

This is the common, nearly spineless, Acraspis gall on the burr oak in the more northern Middle West, but not on the very northern rim of the range of Q. macrocarpa. It is un- fortunate that Bassett’s well-known name, macrocarpae , for this insect is already pre-occupied in Cynips (as indicated in the synonomy above), and that it is necessary to introduce the new name macrescens for this insect. 

Macrescens shows the considerable variation which is evident in northern Middle- Western varieties of Cynips fulvicollis and C. pezomachoides. Toward the north macrescens aver- ages smaller and darker, finally giving way in north-central Michigan to the variety scelesta. In southern Indiana and Illinois, and thru parts of more northern Illinois macrescens averages larger and more robust and gives way to purer popu- lations of variety opima. These three insects are not differ- entiated by many characters, probably because of the great simplicity of structure of all the sub-apterous Cynipidae; but their distinction is important because of the light it throws on the origin of a species by the hybridization of more northern and more southern varieties which were brought together during the glacial invasions of the Pleistocene. Thruout the range of macrescens individuals may be found with the characters of pure scelesta or opima, but the mass of the population of macrescens represents an intermediate blending of the parental characters. Variable and poorly fused as this aggregate may be, I feel we should continue to recognize it as a taxonomic unit deserving a name for convenience of reference. 

The galls of scelesta are ellipsoidal, those of opima more nearly (but not quite) spheroidal. In the galls of macrescens the scelesta influence usually dominates, but there is some variation in this gall character. 

Felt recorded young galls of this variety (misdetermined as hirta) from Minnesota as early as July 23. Weld found pupae in galls from northern Illinois early in September. Galls which I collected at Delphi in northern Indiana on September 21 (1926) were just beginning to drop to the ground. Many of them at that time showed exit holes which indicated the emergence of parasites. The cynipids were still larvae in the galls on that date. Weld cut living adults from his Illinois material as early as the last week of September. Normal emergence dates are recorded as follows: October 10 (acc. Gillette 1889) ; November 4 and 10 (Kinsey in 1926) ; November 9 to 14 (Brodie in U.S. Nat. Mus. ; in 1885 and 1893) ; November 14 (acc. Weld 1926) ; November 15 to 30 (Brodie in U.S. Nat. Mus.; in 1892); November 4, 16, 19, 20, and December 1, 4, 8, 12, 15, 18, 20, 22, and 28 (Kinsey 1926 to 1928). While the major part of the large series of material I have bred emerged in the first half of November, it will be seen that many of the insects emerged after December 1 and even as late as December 28. All of the adults had previously emerged from galls collected at Roachdale, Indiana, on December 26 (in 1922) . Some of the many parasites that are to be found in these galls emerge before the gall makers in the fall, but the majority of the parasites and inquilines do not mature until the following May or June. 

Cynips multipuncatata var heldae
agamic form

The agamic galls of all of the species of the genus are produced on the veins of the leaves of white oak, with the single exception of Cynips heldae, which occurs either on leaf veins, petioles, or young stems of the oak.

GALL:
Generally spherical (as seen in cross-section), but so distorted with peculiarly irregular ridges and spines on the surface as to appear very irregular, even cubical; up to 10.0 mm. in diameter, completely covered with an irregular, often twisted mass of broad, blunt spines which are usually flattened and often ridged; the surface of the gall smooth, naked, with a microscopic scurf on the younger galls, this scurf more or less lost on the older galls; young (but fully grown) galls light reddish-brown in color, often lighter gray because of the scurf, the older galls dark, dirty brown. Internally solid with a compact, crystalline substance which is not as hard as in Cynips echinus and which contains more compacted, branched fibers; the larval cell large, up to 4.0 mm. in diameter, central or asymmetrically placed. Occurring singly or in compacted, distorted masses of as many as 8 galls, on the twigs (terminally or laterally), on the leaf petioles, or on the veins on the upper or under surfaces of the leaves of Quercus lobata. Figures 195, 205-206. 

Occurring singly or in compact, distorted masses of as many as 8 galls, on the twigs (terminally or laterally), on the leaf petioles, or on the veins on the upper or under surfaces of the leaves of Quercus lobata. 

RANGE:
California: Probably confined to a limited area including parts of Mendocino Lake, and northern Sonoma Counties, and rimming at least the northern part of the Sacramento Valley. The remarkable gall of this species is, apparently, not rare in the Mendocino-Lake County area, but it is as far as we know confined to that part of California. While often occurring on the twigs of the Valley oak, one of the two instances of a true Cynips gall in this location, it is by no means confined to the twigs as Weld suggests, for I have specimens from petioles and from leaf veins, both on the upper and under surfaces of the leaf. 

The remarkable gall of this species is, apparently, not rare in the Mendocino-Lake County area, but it is as far as we know confined to that part of California. While often occur- ring on the twigs of the Valley oak, one of the two instances of a true Cynips gall in this location, it is by no means con- fined to the twigs as Weld suggests, for I have specimens from petioles and from leaf veins, both on the upper and under surfaces of the leaf. Oviposition is evidently not so unlike oviposition with other agamic forms of Cynips. Weld records fully grown galls containing immature insects as early as August. Thru the courtesy of Miss Schulthess and Mr. Leach, I have series of the galls and of the adults from a number of localities. Galls collected at Kelseyville on September 25 (in 1926) were fully grown but contained very small larvae. The galls collected at the same locality on October 4 (in 1925) contained larvae that were still not mature. The galls collected at Yorkville on October 22 and at Clover dale on October 26 (both in 1922) contained live adults in November. Adults emerged from Kelseyville material, bred at Bloomington, Indiana, from January 22 to February 5, 1926, at a season when we were having freezing tempera- tures every night and a zero temperature on one occasion. Some of the adults had already escaped from the material collected at Cottonwood by Gambs in January, but others emerged on January 23. Old galls of the previous year’s growth may be found as late as October 25 (as at Clear Lake in 1923), together with the fully-grown galls of the new generation. From one small gall I cut an adult 1.5 mm. in length. It is a miniature of the normal insect except in having fewer spots in the wings, and in being generally less hairy. Heldae has usually been considered one of the most distinct species on our Pacific Coast, and this judgment is well applied to the gall produced by this insect. But the insect, as far as I can see, and as Fullaway pointed out in his original description, is practically identical with C. multipunctata conspicua. The two occur on the same species of oak, and they alone, among all species of Cynips, produce galls on the petioles and young twigs as well as on the leaves of the oak. Conspicua is unknown and there is no other Besbicus except heldae on Q. lobata in the Mendocino-Lake County area. The galls of conspicua and heldae (see figs. 195 and 198) are identical in plan, altho they do differ so much in superficial form. These several considerations lead me to believe that heldae and conspicua are derived from the same stock, and represent the closest of existent relatives of that stock. 

Cynips mirabilis
agamic form

Gall. Large, spherical, spotted, thin-shelled leaf galls with dense, silky, radiating fibers. Up to 33.0 mm., averaging nearer 20.0 mm. in diameter, symmetrical, normally spherical except where drawn out a bit at the attachment; very finely pubescent when young, largely smooth and naked when old, only microscopically shagreened, greenish or with a reddish tinge when young, brownish-yellow or more russet when old, more or less closely set with small, reddish-purple spots. Rather thin-shelled, the larval cell central, supported by numerous, fine, silky, radiating, branched fibers, the cell large, broadly oval, up to 5.0 mm. in length. Attached to the larger veins, singly or near each other, on the under surfaces of leaves of Quercus garryana. 

RANGE. — Probably found wherever Q. garryana or any of its forms occur, from British Columbia into northern California and further south along the Sierras. Figure 34. FIG. 34.

[Kinsey goes on to describe 2 varieties of this gall; see paper for details]

Cynips fulvicollis variety fulvicollis
bisexual form pallipes (Bassett)

Dryophanta pallipes
Diplolepis pallipes

GALL. — A rather elongate but still broadly egg-shaped cell with one end (the base) truncate, flat; reddish-brown and dark brown in color; the surface all but microscopically smooth ; the cell wall thin, hard and brittle, occasionally with more than one layer (as several bud scales become involved in the gall); entirely empty, without a separate larval cell; up to 3.5 mm. in length. Buried inside otherwise unmodified buds, or deforming and dwarfing clusters of leaves which become slender and thread-like about the gall; on Quercus alba. Figure 231.

RANGE. — Probably the same as that of the corresponding agamic form [this seems to be what we know as Philonix nigra] which occurs everywhere on Q. alba, from Massachusetts to Iowa and the Ohio River Valley (fig. 41). The bisexual form known definitely only from: Connecticut: Waterbury (Bassett, types). New York: state (acc. Beutenmiiller 1911). New Jersey: state (acc. Beutenmiiller 1911). Pennsylvania: state (acc. Beutenmiiller 1911). Indiana: Charlestown (E. W. Spieth in Kinsey collection).

These bisexual insects have been previously known only from Bassett’s very scant collection and from Beutenmuller’s records for which I have not seen material. I have secured 16 insects and numerous galls from southern Indiana by bagging large numbers of unopened buds of Q. alba in the early spring.

The date of appearance of the adult insect will vary as the latitude and season may affect the development of the buds of the oaks in the region. In southern Indiana adults were emerging from the galls on April 22 and May 1 in 1927. Bassett did not find the galls appearing in Connecticut until early May, while the insects emerged later in May. This insect has gone uninterpreted since Bassett's original discovery, but the following considerations now seem to lead to our conclusion that this is the bisexual form of no less common an insect than Cynips fulvicollis fulvicollis.

Pallipes is undoubtedly the alternate of some eastern American species of agamic Cynips for which at least some variety is already described. No Eastern species of Cynips (as categories are used in this paper) has been added to our list since 1882, altho new varieties are still being discovered. 

The only eastern American Cynips left for consideration as a possible alternate of pallipes is some variety of fulvicollis. Fulvicollis and pezomachoides are the only Cynips common on Q. alba in the northeastern quarter of the United States. 

[Cynips gemmula (link in text)](https://www.biodiversitylibrary.org/page/53516882#page/363/mode/1up)
bisexual forms

GALL OF BISEXUAL FORMS.— A thin-walled, egg-shaped cell, occurring singly in the buds of the chestnut oaks. Known for only a single variety, so see the description for variety gemmula bisexual form gemmula. 

[Cynips gemmula variety gemmula](https://www.biodiversitylibrary.org/page/53516882#page/371/mode/1up)
bisexual form gemmula

Cynips gemula
Dryophanta gemula
Diplolepis gemmula

GALL. — A small, thin-walled, elongate, egg-shaped cell in a bud. Monothalamous, about 2.0 to 3.0 mm. long, regular in form, nearly egg-shaped but somewhat more elongate, pointed apically, truncate at base; microscopically roughened; older galls dark brown to blackish. Very thin-walled, brittle when dry, without a distinct larval cell. Singly, deep and nearly hidden in the leaf or flower buds, or at the apices of young stems or on the main stems of the staminate flowers and surrounded by a few deformed bracts; on Quercus prinoides. Figures 305, 306, 335. 

RANGE. — Probably as given for the agamic form prinoides (fig. 62) ; probably restricted to an area on or near the Atlantic Coastal Plain. This bisexual form known only from: Massachusetts: eastern part (M. T. Thompson coll. No. 166). Connecticut: Waterbury (types, Bassett). New Jersey: Lakehurst (acc. Beutenmuller 1911). 

Altho not represented by many collections, the good-sized series of this insect in the Bassett and Thompson collections indicate that proper search should reveal this as a very com- mon even if obscure form in the leaf and flower buds of the chinquapin oak, Q. prinoides. Bassett described the galls as fully developed when the staminate flowers of the oak are in bloom. He remarked that “When the gall happens to be in a leaf bud, it is often found at the summit of a young branch one or two inches long, so rapid is the growth of the tree at this season. . . . The insects appear in both sexes about the middle of May ...” [at Waterbury, Connecticut] . A Bassett specimen in the American Museum is labelled as in coitu on May 10 (1879) . Thompson found the galls late in May in eastern Massachusetts, and Beutenmuller recorded (in Smith 1910 :599) galls occurring in May and June— the cells probably being empty for some time before they disappear from the dried-up flower buds. 

Bassett originally spelled this name gemula, a feminine adjective meaning moaning or complaining. It seems evident that Bassett intended to write gemmula, a noun meaning a small bud. This would aptly describe the gall of this insect. Perhaps it is not certain that Article 19 of the International Rules allows an emendation of this original form as a lapsus calami; but perhaps we need common sense rather than rules in this case. Gemmula is here taken to be the bisexual form alternating with the agamic prinoides. My reasons for this opinion are as follows : 

The insect shows the generic characters of a bisexual Cynips as established by the life histories experimentally determined for varieties of folii and divisa, and for erinacei. The insect belongs to the subgenus Acraspis, as evidenced by the character of the hypopygial spine, the wing-body ratio, and the blotches (even tho obscure) in the cubital cell. 

The relation to an agamic Acraspis is further proved by the close resemblance of gemmuia to bicolens, the experimentally demonstrated bisexual form of Cynips pezornachoides erinacei. 

That gemmuia is not also a form of pezornachoides is shown by the host on which it occurs, namely Quercus prinoides, one of the chestnut oaks. Pezornachoides occurs only on the true white oaks of the Q. alba group, and is rigidly excluded from the chestnut oaks. 

Cynips hirta and prinoides are the only agamic forms of Acraspis known from the chestnut oaks, and prinoides is the only Acraspis known from Quercus prinoides. 

Prinoides , then, seems to be the only agamic Cynips of which gemmuia might be the alternate, and none of the data are in conflict with the conclusion that these two are really alternates. 

Cynips echinus agamic form

Holcaspis douglasii
Dryophanta douglasii
Diplolepis douglasii
Disholcaspis douglasii

Cynips quercus echinus
Cynips echinus
Andricus speciosus
Dryophanta speciosa
Diplolepis speciosa
Dryophanta echinus
Diplolepis echina
Dryophanta echina

Gall: Spherical to squash-shaped, hard and crystalline leaf galls, in most varieties bearing short and blunt projecting tips. Strictly or irregularly spherical, or a regular or irregular, inverted, truncate cone; up to 9.0 mm. in diameter, in most varieties bearing short, stout, often crooked and irregular, spiny projections which are in length not half the diameter of the body of the gall; the surfaces of the galls naked or with a puberulence, whitish or light green or pinkish when young, soon becoming (in most varieties) rose violet to coral red in color; internally filled with a compact mass of crystalline material containing a very few branched fibers, the galls soft as rubber when moist, hard and brittle as glass when dry; a more or less complete cavity in the center of the gall below the larval cell and near the very base of the gall, this cavity with the larval cell making the gall appear, superficially, bithalamous ; the larval cell in- separable, embedded in the compact material of the gall or rarely (particularly in immature galls) held in place by fine, silky, radiating fibers; irregular in shape, up to 4.0 mm. in length, sometimes central in the gall, usually in the upper part of the gall directly under the epidermis, the cell centrally or asymmetrically placed, sometimes extending in part into one of the spiny projections. The gall wholly separable, attached by a point to a vein, usually on the under surfaces, sometimes on the upper surfaces of leaves of Pacific Coast white oaks, Quercus lobata , Q. douglasii , Q. dumosa, Q. turbinella, and Q. durata. 

RANGE .—California, Shasta County to the Mexican border, probably in Lower California. 

There are few cynipid galls more abundant than those produced by the agamic forms of echinus in the autumn in California. These galls are confined to white oaks, and do not occur north of Shasta County. 

The young galls of the agamic forms first appear, at least north of the Sierra Madre, early in the summer; the larvae are mature within a couple of months, and the adults mature at various times from then on, emergence occurring in December or January. It should be noted that in mid-winter, in most of the areas involved, freezing temperatures are rare (varieties echinus and douglasii ) , while the colder days and snow storms of other parts of the region are intermingled with warm days when emerging insects might easily become active. 

[Kinsey goes on to describe 6 varieties of the agamic form of this species; see paper for details]

Cynips echinus bisexual form

Dryophanta lobata
Andricus ribes

Gall: Spherical, berry-like, very succulent, pale green and more or less translucent when fresh, bright red when very young; the surface smooth or, in some varieties, pebbled and indefinitely marked with low ridges which bear soft, projecting points. The galls 4 to 7 mm. in diameter, shrivelling greatly upon drying, then becoming blackened; when dried upon the twig they become more or less obconical in shape, remaining lighter straw-brown in color. The wall of the gall moderately thin, thinnest apically; entirely hollow in- side, without a distinct larval cell. Attached very insecurely by a single, slightly elongated point, on the twigs (bud galls) of probably all the oaks on which the agamic generation occurs; definitely known from only Quercus lobata, Q. Douglasii, and Q. durata. 

RANGE .—California, Shasta County to the Mexican border, probably in Lower California. 

The bisexual forms of echinus, on the other hand, are known from hardly more than a half dozen recent collections, representing, however, four of the six known varieties of the species. These bisexual galls are probably not rare, but they are small, inconspicuously green in color, and so ephemeral that the chances of their collection are much reduced. They are bud galls that appear with the bursting of the leaf buds the first thing in the spring; they reach full size within a few days, and are then so loosely attached to the young, devel- oping stems that they fall to the ground in a heavy wind or at the slightest touch. The larval insects reach full size rap- idly and, transforming into adults, emerge within perhaps three weeks after the galls first appear. Emergence thus varies with the season, latitude, and altitude, the limits of our scant records being March 23 further south (at Three Rivers for form vibes) to May 13 further north (at Kelsey- ville for form atrata ) . Most of the emergence is probably in April. The males are not rare in our collections and they are probably produced in equal abundance with the females. Ovi- position has not been observed, but it must be in the veins of the then young but unfolded leaves, for the agamic galls appear within a month and a half to two months of that time. The bisexual galls shrivel very greatly when collected and dried, and decay very rapidly when moist, the latter being the explanation of our complete failure to find them except in the short season before the maturity of the insect. 

The insects and the galls of the bisexual and agamic forms here described have not hitherto been recognized as successive generations, for they are superficially quite distinct. They are not yet connected by the experimental data to which our conclusions must always be subject; but a closer examination shows so many points of similarity and so much confirmatory host, distribution, and life history data that we may be justified in our present interpretation. 

In the first place, the bisexual insects clearly belong to the genus Cynips and to the subgenus Antron. 

They differ no more from the agamic forms than the bi-sexual forms of European Cynips differ from their experimentally connected agamic forms.

There are only three species recognized among the agamic forms of Antron, and the present bisexual forms appear to be connected with the agamic echinus rather than with teres or guadaloupensis, for the following reasons: They have the mesonotum reticulated; they have blotched but not spotted wings; there is another bisexual insect (Beutenmtiller’s pulchella) which shows the spotted wings of the agamic teres. 

The present bisexual forms are represented by varieties on Q. lobata , Q. Douglasii, and Q. dumosa, on all of which hosts we find the agamic echinus but not teres or guadaloupensis ; teres is unknown from Q. Douglasii on which the agamic echinus is common. The bisexual form is represented in the Central Valleys of California, an area in which echinus is common but teres un- known. The bisexual galls are similar to the galls of the agamic echinus in being fundamentally spherical with short, blunt, projections; and the bisexual gall is most nearly spherical and most nearly spineless on the host that bears the agamic echinus galls of this nature (variety schulthessae on Q. durata). 

There is nothing in the structural, host, and distributional data which would preclude the connection of these bisexual forms with echinus, while there would be the several inconsistencies noted above in connecting the forms with teres. 

In spite of the difficulty of collecting the spring galls, they are known to represent four varieties and therefore probably belong to a species that has at least as many common agamic forms known. The agamic echinus again qualifies on this count. 

This species contributes materially to our comparison of morphologic and physiologic data in taxonomy. The agamic insects of the several varieties are all similar, and in several cases practically identical; the bisexual insects are as nearly identical; the bisexual galls are indistinguishable in three varieties, but the agamic galls are so distinct that most workers would accept them as the work of different species. Were we dealing with a group of insects in which we had no such physiologic data as the galls, it is quite certain that no one would arrive at our present classification, and the accompanying host and distributional data would then be largely meaningless. How much insect taxonomy is inadequate and misleading is a matter for serious contemplation. 

That the differences in these agamic galls are not due to the qualities of the various oaks on which they occur, is proved by the occurrence of three types of galls, the work of three different varieties of echinus, on Quercus dumosa in Southern California. 

Attention should be drawn to the peculiar, crystalline materials of which all these agamic galls are built. Between the thin epidermis and the wall of the larval cell there are solid masses of microscopic, deformed crystals intermingled with a few fibers of similar material. This substance is commonly reputed to be gallic acid (as in Fullaway 1911), and if this is so, these galls must have a very high percentage of the material. It would be interesting to know how the gall in its development segregates or stimulates the manufacture of this substance in the normal oak leaf. When the galls are moist (whether young or old) they are as soft as rubber. When dried, they quickly become as hard to cut with a knife or drill as tho they were made of so much compacted, ground glass. In order to mount such specimens on insect pins for preservation in our collections we soak the galls in water for a few hours, or place them out-of-doors in a damp location for a few days, after which they are readily penetrated by the pins. 

In addition to the six varieties now known in this species, there are probably two or three additional varieties still to be described from California, but probably none to be expected elsewhere in this country. 

[Kinsey goes on to describe 4 varieties of the bisexual form of this species; see paper for details]

Cynips guadaloupensis var guadaloupensis
agamic form

Callirhytis guadaloupensis
Acraspis guadaloupensis

GALL. — Much flattened, circular, disk-like, up to 7.0 mm. in diameter and 1.2 mm. in height; the walls thick and the larval cell the only internal cavity; on. leaves of Quercus chrysolepis. Figure 146.

RANGE. — California: Guadaloupe (R. W. Patterson; types). Santa Lucia Mts. (gall, acc. Weld 1926). Probably occurring thruout the range of Q. chrysolepis in a more southern Coast Range area of California; perhaps extending thruout the mountains of California from the Sierra Madre north to the Oregon boundary. Figure 27.

Cynips guadaloupensis var insolens
agamic form

Acraspis insolens

Gall. Peculiarly pouch- or wedge-shaped; drawn to a blunt, cylindrical tip at the base, much compressed and concave on two broad sides; the sides with distinct rims, approaching each other basally but well separated apically, well rounded at the apex; up to 10.0 mm. in length, the broad sides up to 5.0 mm. wide and up to 4.0 mm. apart at the apex; the dimensions usually smaller than this; with the larval cell located near the apex of the gall and a large, elongate, cylindrical cavity lying between the cell and the base of the gall; on the under surfaces of leaves of Quercus chrysolepis. Figures 140-141. 

RANGE. — California: Idyllwild (Weld; types). San Jacinto Mt. (galls, Kinsey). Not known beyond a single locality in the San Jacinto Mountains, but possibly occurring thruout a southern Californian and Lower Californian area that reaches its northern limit in the San Jacinto Mountains. Figure 27. 

Cynips guadaloupensis var patelloides
agamic form

Gall. Deep bowl- shaped, somewhat flattened basally, the sides steep but slightly flaring, the rim more or less circular, the top of the bowl solid, flat to concave; up to 13.0 mm. in diameter by 7.0 mm. high, usually smaller than this; with the larval cell centrally located at the top of the gall, the rest of the gall more or less spongy crystalline; on the under surfaces of leaves of Quercus chrysolepis. Figures 144-145, 192. 

RANGE. — California: Idyllwild (Weld in U.S. Nat. Mus. and Kinsey coll.). San Jacinto Mountains and Pasadena (galls, Kinsey coll.). Camp Baldy in San Gabriel Mountains (gall, acc. Weld 1926). Mt. Lowe (gall, Silvestri in Kinsey coll.). Kern County (U.S.D.A. 5527, acc. Weld 1926). Sequoia National Park (types, Weld coll.). Yosemite (gall, Silvestri acc. Trotter 1910) . 

Cynips teres
agamic forms

GALL. — Small, slender, stalked, swollen at the apex, resembling an inverted Indian club in shape. Up to 8.0 mm. in length and 2.7 mm. in width (near the apex). The stalk slender, strictly cylindrical except for a slight flare at the base, straight or bent, averaging 0.7 mm. in diameter, expanding gradually or abruptly into the swollen, spindle- shaped, or more spheroid body of the gall; the surface of the entire gall minutely granular, covered with a short, crystalline pubescence which more or less disappears with age; the young galls light, creamy brown, becoming light reddish brown and finally dark brown. Internally rather hard, crystalline, the outer wall solid, rather thin, the stalk filled with a less solid crystalline mass, the body of the gall occupied mostly by the larval cell which has a distinct cell wall but is imbedded closely against the outer wall of the gall. Attached more or less at right angles; occurring singly, or a few scattered over the leaf, on the under surfaces of the leaves of Quercus lobata, Q. garryana, Q. dumosa, and Q. durata. 

RANGE. — Probably from British Columbia to California, now known from western Oregon and more northern California southward in the mountains and the foothills to the southern end of the San Joaquin Valley. Figures 28, 29. 

Cynips teres var hidebrandae, new variety
agamic form

Gall. Rather short, up to 6.5 mm. in length, the stalk rather slender but not long, the swollen portion spherical, rounded at the apex; on leaves of Quercus dumosa and Q. durata. Figure 196.

Cynips teres var teres 
agamic form

Gall.  Shorter, more robust, up to 6. mm. in length, the stalk hardly longer than the swollen portion of the gall, the swollen portion quite spherical, well rounded at the apex; on leaves of Quercus garryana and its varieties. 

RANGE. — Oregon: Salem, Cottage Grove, Oakland, Wolf Creek, McLeod, and Siskiyou (galls, acc. Weld 1926). California: Scott Bar and Fort Jones (galls, acc. Weld 1926). Yreka (galls, Kinsey coll.). Pit River, and Cow Creek near Millville (galls, Leach in Kinsey coll.). At about 4000 ft. in Sequoia National Forest (above Cedar Creek Checking Station; types, Weld coll.). Probably confined to higher elevations in Central California, and to more northern California and adjacent Oregon, wherever Q. garryana and its varieties occur. The Oregon records north of the Rogue River may represent a distinct variety. Figure 28. 

Cynips pulchella 
bisexual form

Dryophanta pulchella

No material of this insect has been available for the present study. From the original description pulchella would ap- pear to be a true Cynips. The naked abdomen, blotched and spotted wings, and Pacific Coast distribution suggest that the subgeneric position is in Antron. All of the characters of the insect and the June collection would identify this as a bisexual form, and if it is a true Antron this must be the bisexual generation of some variety of C. teres or C. guadaloupensis. The bisexual forms of C. echinus are known to differ in detail from pulchella. The naive admission that the Oregon specimen has a black thorax while the Catalina Island specimen has a red thorax is, of course, in accord with our knowledge that southern Californian cynipids may be represented by closely related but not identical varieties in northern Oregon. Until someone obtains access to the Beutenmuller collection, it will be impossible to determine which of the two type specimens deserves the name pulchella and which represents the unnamed variety. Pulchella has not been counted in the statistics given in this paper, for if it proves the bisexual generation of some described, agamic form of Antron , it will not alter the total number of varieties known in the genus. 

Cynips multipunctata var multipunctata
agamic form

Dryophanta multipunctata

Gall. Indistinguishable from that of variety indicta, unless aver- aging larger. Green when young, becoming light and then dirty gray in color; appearing finely pubescent, in reality microscopically set with fine papillae and narrow, reticulated ridges, with only a scant pubescence of stellate and longer hairs, these wearing off in time; internally with finer, more silky, less compacted fibers between the outside wall and the larval cell, the cell in consequence easily broken from its position ; occurring singly, rarely in small, not compacted clusters, fastened to the veins, on the under surfaces of leaves of Quercus Douglasii. Figures 203-204. 

RANGE. — California: Kern County (types in U.S. Nat. Mus.). Diablo and Napa (F. A. Leach in Kinsey coll.). Probably thruout the more southern range of Q. Douglasii. Figure 31. 

This variety is restricted to the blue oak, Q. Douglasii, and has thus far been found only on the leaves. Beutenmuller’s original host record was a “species of oak/’ changed in his revision of the genus to “a species of white oak.” Without further data, Felt took this to mean Q. lobata, and McCracken and Egbert give this as the only host. The leaf with the type gall is, however, distinctly that of Q. Douglasii, and I have never seen specimens from any other oak and feel sure that the lobata records are unfounded. Leach collected galls in the latter half of July at Diablo. All of the galls collected after the first week in December have shown the emergence holes of the gall maker, and have never given me further adults. It is, however, to be questioned whether emergence is normally as early as this since the other varieties of the species emerge later in December and in January. The holotype female of multipunctata at the U.S. National Museum is a small, broken, and not deeply pigmented specimen that looks as if it might have been cut from an old gall. It is smaller, with the body lighter in color and the wings less heavily spotted than specimens I have from Diablo ; but these differences might well be due to the immature character of the holotype female, and certainly it is closer to the Diablo material than to indicta, the other variety known from Q. Douglasii.

Cynips multipunctata var indicta, new variety

Gall. Indistinguishable from that of variety multipunctata (q.v.), unless averaging smaller; mature galls becoming light and then dirty gray in color; internally with rather fine, silky fibers; on the under sur- faces of leaves of Quercus Douglasii. Figures 198, 201-202. 

RANGE. — California: South Cow Creek in Shasta County, and Inskip (galls, F. A. Leach in Kinsey coll.). Redding (Kinsey coll, and F. A. Leach in Kinsey coll.). Battle Creek at 2000 ft. (F. A. Leach in Kinsey coll.). Colfax (E. R. Leach in Kinsey coll.). Kelseyville (types, P. Schulthess in Kinsey coll.). 7 miles southeast of Kelseyville; Lower Lake (D. Hildebrand in Kinsey coll.). Cobb Mountain and the south side of Bartlett Mountain in Lake County (P. Schulthess in Kinsey coll.). 

Probably thruout the more northern range, or the altitudinally higher range of Q. Douglasii. Figure 31. 

This is the more northern variety on the blue oak, Q. Douglasii. The gall is hardly distinguishable from that of multipunctata, the more southern variety on the same oak, but the insect is quite distinct. The material from Kelseyville and points north of there is typical indicta; the material from seven miles south of Kelseyville and from other points in the southern portion of Lake County is intermediate between indicta and multipunctata, and not determinable as one or the other variety. The material from Napa County clearly represents variety multipunctata. Miss Schulthess, who collected the type material, found full-sized but not yet fully developed galls as early as July 15 (1925 at Kelseyville) . Mr. Leach has collected more mature galls at later dates in July, but none of the collections made before the first of September contained larvae large enough to be bred. I have bred adults (out-of-doors at Bloomington, Indiana) on December 23 and January 8, 12, and 20, and still later in January (1926-28). From one lot of 52 galls, 7 contained gall makers, and all the others were parasitized. 

Cynips multipunctata var conspicua, new variety
agamic form

Green when young, becoming light and then dark brown in color, appearing mostly smooth and naked, with only a sparse, microscopic puberulence, the microscopic ridges relatively broad; internally more or less solid with compacted, crystalline fibers; occurring singly or often in more or less compacted clusters of up to a dozen galls, on the veins on the under or upper surfaces of the leaves, on the petioles, or on the young twigs of Quercus lobata. Figures 203-204. 

RANGE. — California: Palo Alto (Wiltz coll, in Stanford Univ.). Cupertino (acc. Fullaway 1911). Diablo (F. A. Leach in Kinsey coll.). Napa (types, F. A. Leach in Kinsey coll.). Redwood City (C. T. Dodds in Kinsey coll.). Contra Costa County (galls, E. C. Van Dyke in Kinsey coll.). Klink (galls, L. H. Powell in Kinsey coll.). Probably wherever Q. lobata occurs thruout the Great Valley of California. Replaced in the Mendocino-Lake County area by variety heldae. Figure 32. 

This variety is restricted to the valley white oak, Q. lobata . The galls occur in about equal abundance on the leaves, the petioles, and the young twigs. They probably appear sometime in July. Mr. Leach has collected full-sized but succulent galls at Diablo as early as August 13 and 26 (1926), and more mature galls on September 16 (in 1924). On November 26 (in 1922 at Diablo) there were large larvae in the galls. Galls collected at the same locality as late as December 18 still contained mature adults which emerged soon after that date, tho other records indicate that emergence sometimes occurs late in November or earlier in December. Fullaway recorded two species of Synergus bred from this gall (see above) . Of all the material I have handled, perhaps 80 per cent of the galls were inhabited by parasites or inquilines, the latter not always destructive to the gall maker larvae. Fullaway first described this variety from Stanford University material which Beutenmuller had determined as multipunctata. Fullaway noted the distinctive character of the gall, its location and host, the characteristic spotting of the wings, and the size of the insect, concluding that Beutenmuller’s “specimens are obviously different from mine,” but neglecting to give this insect a new name. Felt listed the two as distinct, but under one name. McCracken and Egbert repeated Fullaway’s suggestion without giving separate de- scriptions of the two. Dr. McCracken has kindly forwarded the Stanford material for my examination, and there is no question that it represents the present variety. 

Cynips maculosa
agamic forms

GALL. — Moderate sized, hard, spherical to flask-shaped leaf gall, up to 9.0 mm. in length and 7.0 mm. in greatest diameter. Regular, spherical to bulb-shaped, attached by a moderately stout, blunt tip which may be up to 1.0 mm. in diameter, evenly but rather abruptly enlarged into a swollen, ovoid body which is well rounded at the tip; essentially smooth and naked, but the epidermis thin, becoming papery, broken, and more or less dehiscent when mature and dry, making the gall appear ragged or (after the epidermis has dropped) leaving it very smooth and polished; the young galls greenish, mottled with white, becoming more yellowish, light orange, pinkish, or reddish brown, finally becoming light to dark brown or blackish. The outer wall of the gall of moderate thickness, hard, compact crystalline, somewhat brittle; the larval cell central in the swollen portion of the gall, at first small with a distinct cell wall, held in place by rather dense, irregular, tangled, crystalline fibers; the larval cell finally up to 3.0 mm. in length, then filling a large part of the gall, the fibers being then compressed into a more solid mass. Attached usually to the veins, on the under side of the leaf, sometimes on the upper side or on the petiole. On Quercus dumosa and Q. durata. 

RANGE. — Known from Mendocino and Lake Counties southward in the mountains along the coast and in the Sierras as far as the Sequoia National Park. Figure 33. 

The interesting gall of this species is apparently not rare in an area including Mendocino and Lake Counties, but there are no records for what is technically Southern California even tho one of the hosts, Q. dumosa , is one of the common oaks in that part of the state. The gall of maculosa is not to be confused with that of any other cynipid, but the insect very closely resembles Cynips mirabilis, differing chiefly in having a shorter fourth antennal segment, less heavily clouded patches on the wings, usually several spots in the discoidal cell, and a smaller size. It is instructive to compare the descriptions of the two insects and note how dependent we are on the physiologic measure (the gall) for separating them. 

[Kinsey goes on to describe 2 varieties of the agamic gall of this species; see paper for details]

Cynips (Philonix) plumbea
agamic form

GALL. — Spherical, thin-walled, brownish, naked, the centrally- placed larval cell imbedded in spongy or more compact tissue. Monothalamous; up to 11.0 mm., averaging nearer 8.0 mm. in diameter. Strictly spherical except for the broadly flattened base; not at all shrivelled when old; the gall greenish or rose-tinged when young, becoming lead color with a bluish gray bloom that is easily wiped off and largely wears off the older galls, exposing the brownish tan color of the older galls; the surface all but microscopically smooth and naked except for the deciduous bloom. The outer wall of the gall moderately thin but firm because it is inseparable from the rest of the gall; the gall internally more or less compact-crystalline, in some cases indicating a more fibrous structure, the larval cell central, closely embedded and entirely inseparable from the compact material of the gall. Attached by a fine point, singly on the main veins, usually underneath the leaves, on Quercus oblongifolia and Q. arizonica (and related southwestern oaks?). Figures 200, 225-226. 

RANGE. — Arizona: Globe (var.? galls, Q. arizonica, Kinsey coll.). Whetstone (galls, Q. arizonica, Kinsey coll.). Oracle (Q. arizonica, Weld and Kinsey coll.). Santa Catalina Mountains (M. Chrisman acc. Weld 1926). Esperara Canyon in Santa Catalina Mountains (types; Q. oblongifolia; Hofer and Edmonston coll.). Sabino Trail in Santa Catalina Mountains ( Q . oblongifolia, Q. arizonica; Kinsey coll.). Santa Rita Mountains ( Q . oblongifolia, Weld and Kinsey coll.). Fort Hua- chuca ( Q . arizonica, Kinsey coll.; gall, acc. Weld 1926). Courtland ( Q . arizonica, Kinsey coll.). Chiricahua Mountains ( Q . arizonica, acc. Weld 1926). Nogales, Patagonia Mountains, and Tumacacori Mountains (galls, acc. Weld 1926). This variety probably restricted to more southern Arizona and southwestern New Mexico, altho related varieties may be expected elsewhere in the Southwest and in Mexico. Figure 36. 

The gall of this species is not uncommon in the mountains of southern Arizona, altho we have nowhere seen it as abundant as fulvicollis is in the East. The species occurs on both Q. oblongifolia and Q. arizoniea thruout the forested elevations below six or seven thousand feet, and the insects agree with many other Cynipidae in showing no appreciable variation on these two oaks. The galls of plumbea seem more firmly attached to the leaves than the galls of the easily deciduous fulvicollis. Plumbea galls probably begin development at some time in June. Weld found pupae in some galls on December 7 (1921), but a mature adult in another gall as early as November 24. This adult, when cut from the gall, lived in a pill box until January 1, just as it would have lived for some time as an adult before emerging from the gall. In this delayed emergence plumbea is in accord with the other species of Cynips . As dates of emergence of the adult, Weld records November 22, December 14 and 30, January 8, 12, 13, 24, and 25, and February 3, 4, and 6. I bred insects on January 6, on other dates in the first half of January? and at some date (unrecorded) after January 16 (gall in 1920). By analogy with Cynips fulvicollis of the present subgenus, the bisexual form of plumbea may be expected in a small, seed-like gall occurring in the buds of the oaks early in the spring. 

Cynips fulvicollis var gigas
agamic form

Philonix gigas 
Philonix nigra err det Weld

GALL. — As described for the species, larger than other varieties; prominently pubescent; up to 20.0 mm. in diameter; on the leaves of Quercus lyrata, Q. Michauxii, and Q, Muhlenbergii. Figure 229. 

RANGE. — Arkansas: Hoxie (Weld, Q. lyrata, types). Missouri: Poplar Bluff ( Q . lyrata, acc. Weld 1922). Kansas: Manhattan and Holton ( Q . Muhlenbergii acc. Weld, 1926). Riley County (Marlatt in Kans. Agric. College and Kinsey coll.). Arkansas City ( Q . Michauxii, R. Voris in Kinsey coll.). Winfield, Cedarvale, Silverdale, and Dexter ( Q . Michauxii, Q. Miihlenbergii, R. Voris in Kinsey coll.). Probably confined to Q. lyrata and to chestnut oaks, in the Ozark area and its immediate extensions. Figure 40. 

This is an Ozark variety which seems to be isolated from varieties major and vorisi of the same region by its host relations. The types of gigas came from Q. lyrata and I have many insects that appear to represent the same variety from Q. Muhlenbergii and Q. Michauxii. Gigas is closely related to major and vorisi, and occasionally the host restrictions fail and the three hybridize in eastern Kansas. Weld was correct in recognizing lanaeglobuli, from Q. bicolor in Florida, as another relative of gigas. 

Weld found pupae in the galls on October 10 (in 1917) ; he found live adults in the galls on November 16. The Marlatt material from Riley County, Kansas, was bred in November and January. Weld’s material bred out-of-doors at Evanston gave eight adults on December 1 and three more on December 18, the thermometer having registered — 14° F. between those dates. He found other adults emerging in the next spring, and suggests that “emergence must be distributed over at least two or three seasons, for normal larvae were found when the last of the galls were cut open December 2”, more than two years after collecting. Almost all of the insects I have bred emerged in the first winter after collection, only two emerging the second year in my breeding bags. My records for emergence are December 17, 20, 22, 28, and 24 ; January 4, 5, 7, and 9; and February 6. 

Cynips fulvicollis var lanaeglobuli
agamic form

Acraspis lanaeglobuli
Philonix lanaeglobuli

GALL. — As described for the species, up to 8.7 mm. in diameter, averaging a little larger than other varieties except gigas and vorisi; on the leaves of Quercus bicolor. 

RANGE. — Eastern Florida (Ashmead; types). The data are insufficient to suggest further extensions of this range. 

Nothing is known about this variety beyond the data with the type collection. The insect is distinct, and nigra and gillettei should not be made synonyms of lanaeglobuli as Beutenmuller would have them (1918). Q. bicolor galls which I have from New York (Staten Island) and New Jersey (Broadway) average near the size of the present variety, but until we can study insects from these localities we cannot suggest whether lanaeglobuli occurs thruout the Atlantic Coastal Plain or is confined to an area in Florida. The nearest relatives of lanaeglobuli are the Middle-Western gigas on Q. lyrata and the chestnut oaks, and vorisi on Q, macrocarpa and Q. bicolor. 

Cynips dugesi var simulatrix, new variety
agamic form

Cynips dugesi var C Kinsey

GALL. — As described for the species. [Thin-shelled, spherical, sessile, dull brown, unspotted, averaging 16.0 mm. in diameter, with the larval cell held centrally by fine, radiating fibers.] Apparently not to be distinguished from other varieties of dugesi, nor from Cynips bella bella which occurs in the same region. On leaves of Quercus undulata, Q. grisea, Q. oblongifolia, Q. arizonica , Q. Gambelii, and probably related oaks. Figure 263. 

RANGE. — New Mexico: Magdalena ( Q . grisea, acc. Weld in Kinsey coll.). Near Alamogordo at 7000 ft. (galls, Q. undulata and Q. arizonica, Kinsey coll.). Highrolls (galls, Q. undulata, Kinsey coll.). Hillsboro (types, Q. grisea, Kinsey coll.). Kingston (galls, Q. grisea and Q. Gambelii, Kinsey coll.). Soledad Canyon in Organ Mts. (galls, Q. grisea, L. H. Bridewell in Kinsey coll.). Arizona: Prescott ( Q . grisea, Kinsey coll.). Safford, Oracle, and Bisbee (galls, Q. arizonica, Kinsey coll.). Sabino Trail in Santa Cata- lina Mts. and Santa Rita Mts. (Q. oblongifolia and Q. arizonica, Kinsey coll.). Courtland and Fort Huachuca (Q. arizonica, Kinsey coll.). Probably confined to the desert mountain ranges of New Mexico and Arizona (and adjacent Mexico?). Records based on galls alone may involve confusion with Cynips bella bella. Figure 44. Hillsboro (types, Q. grisea, Kinsey coll.). Kingston (galls, Q. grisea and Q. Gambelii, Kinsey coll.). Soledad Canyon in Organ Mts. (galls, Q. grisea, L. H. Bridewell in Kinsey coll.). Arizona: Prescott ( Q . grisea, Kinsey coll.). Safford, Oracle, and Bisbee (galls, Q. arizonica, Kinsey coll.). Sabino Trail in Santa Catalina Mts. and Santa Rita Mts. (Q. oblongifolia and Q. arizonica, Kinsey coll.). Courtland and Fort Huachuca (Q. arizonica, Kinsey coll.). Probably confined to the desert mountain ranges of New Mexico and Arizona (and adjacent Mexico?). Records based on galls alone may involve confusion with Cynips bella bella. Figure 44. 

This variety, with Cynips bella variety bella , produces the common oak apple of our Southwest. Most of the insects had emerged before I collected the galls in New Mexico late in December (1919) and in Arizona in January (1920), but one female emerged at Prescott, Arizona, as late as January 23.

The available insect material represents a wide geographic range and a surprising number of hosts, apparently without the isolation of distinct varieties. 

The single, broken insect which I have of this species from Globe, Arizona, does not agree with typical simulatrix, but the specimen does not warrant description. On the basis of data from other cynipids, we might expect a distinct variety of dugesi in the Pinal Mountains near Globe. 

Cynips dugesi variety dugesi
agamic form

Dryophanta dugesi
Diplolepis dugesi

GALL. — As described for the species, galls thus far collected being rather small (up to 15.0 mm. in diameter), with dull surfaces; on leaves of Quercus mexicana and other white oaks. 

RANGE. — Mexico: San Luis Potosi (Ed. Palmer, in M. C. Z. and Kinsey coll.). Mts. of Guana juanto (Duges in U.S.N.M.). Probably confined to an area in more central Mexico. Figure 44. 

Mayr gave Q. mexicana as the host of the type material, and his illustration more nearly matches mexicana than the black oak, hypoleuca, to which Beutenmiiller referred it. My specimens were on an oak more closely related to Q. oblongi- folia (which it is not). If Mayr’s host reference is correct, then mexicana is a white oak arid not a black oak as Trelease lists it (Nat. Acad. Sci. Mem. 20 :173) . 

Cynips dugesi variety brevipennata
agamic form

Holcaspis brevipennata
Disholcaspis brevipennata
Andricus pellucidus
Diplolepis brevipennata
Cynips dugesi var A Kinsey

GALL. — As described for the species. Apparently not to be distinguished from other varieties of dugesi unless averaging smaller (types up to 15.0 mm. in diameter), with a thinner and more translucent shell. On leaves of the several forms of Quercus Gambelii (= Q. undulata Gillette, not recent authors), Q. fendleri (acc. Weld), and Q. grisea (acc. Weld). Figure 262. 

RANGE. — Colorado: Manitou (Gillette; brevipennata types). Colorado Springs (Carpenter, pellucidus types). Trinidad and Morley (acc. Weld 1922). Wetmore, West Cliff, La Yeta, and Spanish Peaks (acc. Weld 1926). New Mexico: Raton, Wagon Mound, Shoemaker, Glorieta, and Tijeras (acc. Weld 1926). Las Vegas Hot Springs (acc. Weld 1922, inch Kinsey coll.). 28 miles east of Raton (C. Schwachheim in Kinsey coll.). Probably confined to a limited area in the mountain region of southern Colorado and more northern New Mexico. Figure 44. 

INQUILINES . — Synergus similis Gillette (acc. Gillette 1896). S. atripes Gillette (acc. Gillette 1896). S. oneratus (Harris) (acc. Dalla Torre and Kieffer 1910). 

Weld found both pupae and adults in galls collected at Manitou as early as October 4 (in 1921), and adults in the galls “ready to emerge” [?] at Wetmore on October 6. Gillette cut live and active females from the galls of the type material on October 31. Weld cut a live adult out of a gall on November 9 (1917); and specimens in the U. S. National Museum (Hopkins coll.) were bred October 31, November 15, November 20 (1918) and January 10. I have seven adults from east of Raton which were bred on November 20 (1927).

I agree with Weld that my pellucidus is the same as Gillette’s brevipennata. The two insects came from probably identical localities, the Garden of the Gods, midway between Manitou and Colorado Springs. A direct comparison of types shows them to agree, even tho brevipennata has always figured in the genus Disholcaspis where it should never have been placed, while I placed pellucidus in the meaningless assemblage called Andricus pending the present revision of Cynips. My 1920 description of pellucidus contained numerous inaccuracies which must give way to the present re-descriptions of the insects. 

Considering our present difficulty in separating galls of the species bella and dugesi, much less of the several varieties of each species, records for brevipennata based on galls alone are of little value. 

Cynips dugesi variety pupoides, new variety
agamic form

Cynips dugesi var. B Kinsey

GALL. — As described for the species. Apparently not to be distinguished from other varieties of dugesi nor from Cynips bella congesta which occurs in the same region. On leaves of Quercus grisea. 

RANGE. — Texas: Alpine and Fort Davis (Kinsey coll.) . Probably confined to the mountain ranges of West Texas (and adjacent New Mexico and Mexico?). Figure 44. 

Insects were still emerging from the galls at Alpine, Texas, on December 14 (in 1919), and two days later at Fort Davis; but most of the emergence had occurred before that date. The two short-winged varieties of dugesi, namely pupoides and brevipennata , differ in head and thoracic color, in the angle of the parapsidal grooves at the scutellum, and in details of other structural characters. Pupoides is distinctly smaller than brevipennata. The two insects nearly agree in wing length and in several points of venation; and since they approach each other in their geographic location we may consider that they represent a distinct line of evolution from dugesi and its more typical varieties. The cynipid fauna of the mountains of West Texas is usually distinct from that of New Mexico and Arizona. 

Cynips bella variety bella
agamic form

Cynips bella
Dryophanta bella
Andricus bella
Holcaspis maculipennis
Diplolepis bella
Disholcaspis maculipennis
Cynips maculipennis
Dryophanta maculipennis

GALL. — As described for the species, apparently indistinguishable from the galls of other varieties of this species and of Cynips dugesi. On Quercus grisea, Q. undulata, Q. undulata var. pungens ( = Q. wrightii) , Q. Gambelii, Q. Toumeyi (acc. Weld), Q. oblongifolia, Q. arizonica, Q. diversicolor (acc. Weld; =Q. reticulata), and probably related oaks. Figure 263. 

RANGE. — New Mexico: Magdalena ( Q . grisea, Weld inch Kinsey coll.). Blue Canyon near Socorro, Abo Pass (in Sandia Mts.?), Nogal Canyon, Burro Mountains, Fierro, and Tijeras (galls, acc. Weld 1926). Near Alamogordo at 7000 ft. ( Q . arizonica, also galls on Q. undulata, Kinsey coll.). Highrolls (galls, Q. undulata, Kinsey coll.). Mountain Park (galls, Q. undulata var., L. H. Bridewell in Kinsey coll.). Hills- boro ( Q . grisea, Kinsey coll.). Kingston (galls, Q. grisea and Q. Gam- belii, Kinsey coll.). Organ Mountains (R. R. Larkin, Q. Wrightii; maculipennis types). Soledad Canyon in Organ Mountains ( Q . arizonica, L. H. Bridewell in Kinsey coll.). Arizona: Prescott ( Q . grisea, Kinsey coll.). Safford, Oracle, Courtland, Fort Huachuca, and Bisbee (galls, Q. arizonica, Kinsey coll.). Tucson (= Santa Catalina Mountains? Cox coll., bella types). Sabino Trail in Santa Catalina Mountains and Santa Rita Mountains ( Q . ob- longifolia, Q. arizonica, Kinsey coll.). Tumacacori Mountains, Patagonia Mountains, Chiricahua Mountains, Mule Mountains, and Nogales (galls, acc. Weld 1926). Probably confined to the desert mountain ranges of more southern New Mexico and Arizona (and adjacent Mexico?). Figure 45. 

This is the most widespread variety of the species. The oak-inhabited mountain ranges of this part of the Southwest are isolated by broad deserts, but the available insect material does not show variation enough (unless in the size of the areolet) to warrant a distinction of varieties in this whole area. Weld, however, records galls from Hackberry, Ashfork, and Williams in Arizona, and these more northern localities may represent some other, as yet undescribed variety if the distribution of this insect agrees with that of most of the other Cynipidae of Arizona. The material from the Apache Trail country of Arizona represents a distinct variety, vanescens. 

It is surprising to find so many oaks serving as hosts of this one variety, but as far as I can determine they do not effect the isolation of distinct varieties. In this respect bella parallels Cynips dugesi simulatrix. 

Weld found nearly full-grown galls in southern Arizona (Huachuca Mts.) as early as July 9, and I have fully grown galls dated August 16 and 22. Weld found pupae in galls at Hackberry (representing another variety?) on October 6. He found mature adults emerging from the galls as early as November 1 (1921), and sent me a wasp which was alive for almost three months (until nearly the end of January) after thus coming out of the gall. There were a few live adults left in the galls I collected in New Mexico at Alamogordo on December 21, and at Hillsboro on December 26 (1919), and in Arizona in the Santa Catalina Mts. on January 1 (1920); and these insects emerged soon after collecting. Last winter (1927-28 I had adults emerge from Organ Mountain material on January 2. Weld secured one lot of galls at Tijeras, New Mexico, which were only scantily parasitized, but records only 6 gall makers left in 65 galls collected in the Chiricahuas, and one in fifty in Santa Rita specimens. In my own experience I have never found very high parasitism in this species, tho it might well be expected locally.

I have examined the types of both bella and maculipennis, and checked them with series representing the wide range and hosts of the present insect. I had already concluded that the two names were synonyms when Weld published the same opinion. The types of bella and of maculipennis were collected at points 270 miles apart but well within the limits of the range of the present insect. The hosts of the types were not the same, but both are represented in our recent collections. As we now understand it, the only insect of the region with which bella can be confused is Cynips dugesi simulatrix, but altho the galls of these two are absolutely indistinguishable, the insects are abundantly distinct. 

Cynips bella variety vanescens, new variety
agamic form

GALL. — As described for the species, apparently indistinguishable from the galls of the other varieties of bella and probably from an as yet undescribed variety of dugesi that may occur in the same region. On leaves of Quercus grisea.

RANGE. — Arizona: Globe (Kinsey coll.). Apache Trail at Fish Creek Probably restricted to a limited area that lies east of Phoenix in Arizona. Figure 45.

Most of the insects of this variety had emerged at Globe, Arizona, before January 20 (in 1920), but the type female emerged after that. It is never good practice to publish a new species on the basis of a lone specimen, but it is perhaps permissible when the new form is clearly differentiated from a group of related things whose variation, range, and host relations we already understand.

Cynips (Atrusca) cava
agamic form

Diplolepis cava

GALL. — A large, gourd-shaped leaf gall with a spherical body, with fine, dense fibers supporting the central larval cell. Monothalamous. Most of the gall strictly spherical, up to 22.0 mm. in diameter, drawn out into a short, rather stout, more or less cylindrical base which may be as much as 3.0 mm. in length and 3.0 mm. in diameter; the surface essentially smooth but not very shining, microscopically shagreened; young galls light yellow with a rose tinge; older galls becoming light russet brown, unspotted; the outer shell rather thin but hard and brittle when dry. Internally filled with silky, moderately dense (but not packed), branching fibers which hold the larval cell centrally; some short, incomplete fibers also on the inner wall of the outer shell and on the larval cell; the cell up to 3.5 mm. in diameter. Attached by a fine point at the bottom of the drawn-out base; on the upper or ('more often) the under surfaces of leaves of Quercus breviloba and its form laceyi. Figures 260-261. 

RANGE. — Texas: Austin (type, Weld coll.; also Patterson in Kin- sey coll.). Round Rock and Leander (Kinsey coll.). Boerne and Kerr- ville (acc. Weld 1926). , Probably restricted to a limited area in Texas, extending mostly west from Austin. Figure 46. 

This gall closely resembles the galls of the related species bella and dugesi, but differs in having a drawn-out, blunt base. Patterson found young galls as early as June 7 (in 1921). The galls develop thru the summer and fall, the adults maturing and chewing out of the larval cells sometime before emergence from the outer walls of the galls. On November 10 and 20 Weld found pupae which transformed on November 26 and later. He found adults among the radiating fibers in the galls by December 4. Patterson found live adults in this condition on November 25 (in 1921). By early De- cember most of the galls have fallen from the trees, in many cases still attached to the dead leaves, and some of the adults begin emerging then, altho my collections made on December 6 and 8 (in 1919) still contained many adults which did not emerge until later. 

Several of the cynipid species which occur just northwest of Austin are confined to the Texan post oak, Quercus breviloba, and it is possible that we will not find cava on any other host. 

The insect is close to Cynips bella and dugesi. Cava might be considered a variety of dugesi , but its median depression on the scutellum, its nearly unspotted wings, and more particularly its distinctive gall leads to my conclusion that it is a distinct species. Cynips cava and Cynips dugesi are the only insects in this genus which have shortened wings in which the reduction is less than forty per cent of the wing length normal for the subgenus. 

Cynips (Atrusca) centricola
agamic forms

GALL. — A large, spherical, spotted or unspotted, thin-shelled leaf gall with rather sparse, radiating fibers. Monothalamous. Strictly spherical except where flattened a bit at the point of attachment; ivory to apricot or more pinkish in color, some galls well covered with purplish spots; smooth and often shining, only very microscopically roughened, young galls with a microscopic, stellate pubescence; up to 26.0 mm., averaging under 20.0 mm. in diameter. Internally hollow except for the thick-walled larval cell which is up to 4.0 mm. in length, for the relatively few, silky, radiating, branched fibers that hold the larval cell in position, and for a few shorter, incomplete, branched fibers also located on the inner wall and on the larval cell. Attached by a minute point to a vein; usually on the under surface of a leaf, on Quercus stellata. Figure 197. 

RANGE. — Known from New York to Florida, Missouri, and Texas; probably not extending much further than this. Figures 47, 48. 

This is the oldest and most commonly known American species of Cynips, but it is poorly represented in most collections. None of the published data, nor most of the collections I have made would indicate that it is ever abundant in any part of its range, but 1 have on several occasions found individual trees bearing more than fifty galls each. These galls are confined to the post oak, Quercus stellata, on the leaves of which they are strikingly beautiful objects. There are four known varieties, centricola from the Coastal Plain and its inland extensions, clivorum from the Southern Appalachians, strians from the Ozark area, and Karsch’s rubrae from Texas. The varieties are separable on only a few insect characters, and on the presence or absence of spots on the galls, and altho typical material from the center of each range is definitely determinable, hybrid material from areas between two of these ranges is not easily analyzed. The slow evolution in this species is in contrast to the more rapid evolution of such a species as Cynips mellea in the same range. 

The presence or absence of spots on the galls of this species was taken by Osten Sacken, in the original description of the typical variety, to vary with the age of the gall. It is possible that the amount of weathering has something to do with this, but I have unspotted galls of the normally unspotted variety clivorum collected on various date from October 24 to November 16, and spotted galls of the normally spotted varieties rubrae and strians from September 9 to January 21. Only with variety centricola do we regularly find galls of both types, and the data do not seem to show whether hereditary or environmental factors are concerned in this instance. 

The galls of centricola superficially resemble those of Amphibolips inanis, a black-oak species of the eastern United States; but inanis galls are distinguishable by their host and by their fewer, coarser fibers, and by numerous other points. The galls of centricola even more closely resemble those of the Pacific Coast species, Cynips mirabilis, but the galls of mirabilis have denser fibers inside. Mirabilis and centricola belong to distinct subgenera. Some of the unspotted galls of Cynips centricola clivorum also approach near identity with the galls of Cynips bella and Cynips dugesi.

Centricola is in several respects distinct from the other species of the subgenus Atrusca, but these differences are not fundamental enough to warrant the establishment of another subgenus for one species. 

[Kinsey goes on to describe four varieties; see paper for details]

Cynips (Acraspis) arida, new species
agamic form

GALL. — Small, globular, thick-walled leaf gall with the larval cell filling a large part of the interior. Monothalamous, up to 7.0 mm., averaging under 5.0 mm. in diameter. Strictly spherical when fresh, often shrinking irregularly when dried; the gall sometimes drawn out at the point of attachment; the surface of the gall finely, irregularly shagreened, covered with a stellate pubescence when young, the older galls becoming more naked but still with a whitish puberulence; the younger galls probably whitish or pinkish, the older galls rather dark purplish brown. The thin outer shell re-enforced with a rather thick mass of not very hard, crystalline material, the larval cell highly variable in size, often occupying most of the interior of the gall, the wall of the cell without a distinct cell wall; on leaves of Quercus grisea. Figures 290, 291, 333. 

RANGE. — Texas: Alpine (types); Fort Davis. Probably confined to the mountain ranges of western Texas and adjacent parts of New Mexico; related varieties known to occur in more southern New Mexico and Arizona; probably in Mexico. Figure 57. 

This is a very common cynipid in western Texas. I have New Mexico and Arizona material that represents at least two other varieties of this species, but none of this material warrants description at this time. 

A few of the insects had emerged from the galls which I collected in West Texas in the middle of December (1919). Most of the insects emerged soon after collecting. 

It is largely an academic question whether this insect should be considered a variety of Cynips mellea, a variety of Cynips nubila, or a distinct species showing affinities to both mellea and nubila. In general appearance the insect is more like nubila, the smoky patches on the wing giving it a striking if superficial resemblance to that insect; and in many details of structure it shows more than subgeneric relation to nubila . The galls, on the other hand, are so unlike those of nuhila, and so similar to those of mellea that our first decision was to consider this a Southwestern representative of mellea. There are, however, few cases of specific identities among the Cynipidae of more eastern and more western Texas. 

Cynips mellea variety rydbergiana
agamic form

Dryophanta rydbergiana
Diplolepis rydbergiana
Cynips mellea var D Kinsey

GALL. — Finally naked, flesh-colored to light brown, with little solid material; rounded or flattened basally, occurring singly on the leaves of Quercus undulata varieties (Q. fendleri, Q. rydbergiana , etc.). 

RANGE. — New Mexico: Las Vegas Hot Springs (acc. Cockerell 1903; also acc. Weld 1926). Wagon Mound and Shoemaker (acc. Weld 1926). Figure 51. 

Probably confined to a southern Rocky Mountain area in southern Colorado and New Mexico, probably not extending south of the Sandia Mountains in New Mexico.

This is the southern Rocky Mountain variety of mellea. It occurs on at least two and probably several of the other oaks which are hardly more than varieties of Q. undulata. 

Weld found mature galls in October and obtained emerging adults on December 31 (1921), January 4 and 20, and February 1 and 6 (in 1922, from galls collected in the fall of 1921), most of the emergence occurring at the later dates. From galls collected in the fall of 1922, Weld secured adults on February 16, 1923. Cockerell cut the type female (dead or alive?) out of a gall on March 21. 

Cockerell and later workers have placed this insect in our present genus (known to them as Dryophanta or Diplolepis), but with the remark that “it does not exactly agree with any described American genus.” Cockerell added that “In Mayr’s table, given by Cresson, it runs to Biorhiza , but it does not belong there. The gall is like that of Amphibolips.” 

The mesonotum is more shining in rydbergiana than in any other variety of mellea, and this and the nearly simple tarsal claw provide the superficial resemblance to Biorhiza. The gall, as I understand our cynipid genera, does not have the slightest resemblance to an Amphibolips gall. A direct com- parison of rydbergiana with other varieties of mellea, especially with unica and anceps which are its geographically closest relatives, gives convincing proof of the relationships of all these insects. The gall of rydbergiana is quite indistinguishable from the galls of anceps and compta. While recognizing the extreme position of mellea in the subgenus, I have given my reasons, under the subgeneric description, for keeping this species in Acraspis. 

We owe the recognition of this insect to Weld’s collection of a good series from the type locality. Rydbergiana would otherwise have to pass in this paper as an unrecognizable name, for the type is a broken and immature specimen that as recently as the summer of 1926 did not convince me of its relations. Thru Mr. Weld’s courtesy I have been able to study some of his material and have reached my present conclusions. Altho Cockerell described this insect with 12-segmented antennae, better material shows the 14 segments common to the other varieties of the species. 

Cynips mellea variety unica
agamic form

Diplolepis unica
Cynips mellea var E Kinsey

GALL. — As described for the species, becoming white and naked when mature; usually rounded or flattened basally; occurring singly or clustered, on the leaves of Quercus stellata. 

RANGE. — Illinois: America, Olney, and Christopher (Kinsey coll.). Missouri: Ironton (Weld coll., types). Arcadia and 6 miles west of Dexter (Kinsey coll.). St. Louis (gall in U.S. Nat. Mus.). Poplar Bluff (acc. Weld 1926). 10 miles southeast of Springfield (R. Voris in Kinsey coll.). Arkansas: Hoxie (galls, acc. Weld 1926). Kentucky: Dawson Springs (Kinsey coll.). Ohio: Chillicothe (Kinsey coll.). Probably centering in the Ozarks of Missouri and more northern Arkansas, but extending eastward into the hills of southern Illinois, Kentucky, and southeastern Ohio. Figure 51. 

I found not fully mature galls of unica, which, however, were already dropping from the leaves of the post oaks, in southern Illinois during the second week of October (1927), and in southeastern Missouri and southern Ohio during the last week in October (1926). Live adults were in galls from western Missouri on November 21 (1928) . I have bred adults on December 12, on March 22, during the last week of March, and an April 10. From the galls which Weld collected in Missouri in October, he bred adults at Chicago on May 15, 1918. 

The distribution data for this variety are typical for an Ozark cynipid. The hills of western Kentucky and southern Illinois, which are geologically of the same origin as the body of the Ozarks in Missouri and northern Arkansas, contain an interesting extension of the Ozark fauna. The Mississippi Valley in this region has a Coastal Plain fauna. Two of my forty insects of this species from Dexter, Missouri, represent unica , the other 38 being the typical Coastal Plain Carolina. Dexter stands at the boundary between the lowlands of the Mississippi and the uplands of southern Missouri. The single insect from my Poplar Bluff collection has an interesting combination of unica, rydbergiana, and compta characters. Weld’s records (1926) for unica in southern Arkansas and in Texas, based on galls only, probably apply to anceps, and his Florida record to variety mellea. I cannot interpret his Virginia record (Pergande collection), but I question the occurrence of unica that far east. 

Cynips teres variety clavuloides, new name
agamic form

Dryophanta clavula
Diplolepis clavula

GALLS. — Slender, long, up to 8.0 mm. in length, the stalk nearly twice the length of the swollen portion of the gall, the swollen portion slender to broad spindle-shaped, drawn out into a slender point apically; on leaves of Quercus lobata. Figures 142-143. 

RANGE. — California: Yorkville and Clear Lake (galls, Leach in Kinsey coll.). Kelseyville (P. Schulthess in Kinsey coll.). Sonoma and Napa Counties (types; Koebele coll.). Napa (galls, Leach in Kinsey coll.; also J. C. Bradley acc. Weld 1926). Sacramento (in Mus. Comp. Zook). Chico, Calistoga, Stockton, Kaweah, and Levee (galls, acc. Weld 1926). Concord (C. T. Dobbs in Kinsey coll.). Mt. Diablo (Leach in Kinsey coll.). Palo Alto (Wiltz in Stanf. Univ.). Sierra foothills east of Dinuba, and St. Johns River near Klink (L. H. Powell in Kinsey coll.). Confined perhaps to a narrow area rimming the Central Valleys of California, probably limited to Q. lobata. Figure 28. 

This distinctive gall is never abundant, but it is usually included in collections made in the autumn on the Valley white oak of the foothills rimming the Central Valleys of California. The young galls begin development in August (acc. Weld), and are fully grown by the middle of September (September 12 in 1921 at Concord). Galls collected as early as October 4 (in 1925 at Kelseyville) had very large larvae in them. Most of the insects 1 have bred have emerged during the first half of January, but the Kelseyville material emerged between January 29 and February 5, 1926. McCracken and Egbert record January emergence for indoor breeding. 

All of the material I have examined has come from Q. lobata , and if this or a related variety ever occurs on Q. Douglasii it would appear to be rare. In the original description Beutenmiiller’s host record ran “a species of white oak (prob- ably Quercus douglasi).” He corrected this to “Quercus lobata ” in his subsequent treatment of the genus Dryophanta. I do not know the basis of the McCracken and Egbert record for Q. Douglasii. 

While we may yet find this variety in the Central Valley proper, it is noteworthy that up to date the only records are from the foothills rimming the Valleys from their northern to their southern limits, a peculiarly outlined area that has other insects and plants confined to it. 

The figure in Kellogg’s American Insects, and galls in my collection taken east of Dinuba, show larger, more swollen apices than any of the galls from other parts of California. The insects from the Dinuba material similarly average larger. 

Upon transferring this insect to the genus Cynips , Beutenmuller’s name of clavula, being preoccupied, gives way to teres for the species and a new name for the variety. 

There are galls of this species in the Ashmead collection at the National Museum labelled types of Andricus claviger Ashmead. This is a confusion of material, for the holotype insect of claviger represents a Florida species of an entirely different genus. 

Cynips echinus variety douglasii
agamic form douglasii

Holcaspis douglasii
Dryophanta douglasii
Holcaspis douglasi
Diplolepis douglasi
Disholcaspis douglasii
Dryophanta douglasi

GALL. — Squash-shaped, an irregular, truncated cone attached by its smaller end, flaring distally, with 5 to 11 short, spiny projections from the rim of the distal end of the gall; mature galls pale pink, often with a puberulence which makes them more violet in color. On leaves of (usually) Quercus lobata , rarely on Q. Douglasii. Figures 154-159. 

RANGE. — California: 5 miles north of Upper Lake (galls, Hilde- brand and Schulthess in Kinsey coll.). Konocti Bay in Lake County, and Napa County 12 miles southeast of Middletown (Schulthess and Hildebrand in Kinsey coll.). Inskip and Vina (galls, Leach in Kinsey coll.). Scott Valley in Lake County (gall, P. Schulthess in Kinsey coll.). 7 miles southeast of Kelseyville (Hildebrand in Kinsey coll.). Kelseyville (P. Schulthess in Kinsey coll.). Clear Lake, Napa, and Diablo (Leach in Kinsey coll.). Sonoma County (in U.S. Nat. Mus.). Sacramento (gall, in Kinsey coll.). Marin County (Koebele, types). Walnut Creek (C. T. Dodds in Kinsey coll.). Woodland (gall, L. Ewart in Kinsey coll.). Stanford University (McCracken in Mus. Comp. Zool.). Morgan Hill (gall, G. Reed & Z. Cunningham in Kinsey coll.). San Jose (Patterson acc. Fullaway 1911). Byron, Paso Robles and Three Rivers (Kinsey coll.). El Portal (gall, Silvestri in Kinsey coll.). Klink on St. Johns River, and Dinuba (L. H. Powell in Kinsey coll.). Han- ford (E. 0. Essig in Kinsey coll.). Tulare (0. E. Brown in Kinsey coll.). Probably thruout the Great Valley of California, wherever Q. lobata occurs. Figure 23. 

The gall of this agamic form is common on the valley white oak, Quercus lobata , thruout the Great Valley of California. The young galls appear late in June (June 28, 1927, near Kelseyville, and July 6 at Diablo in 1922), and many of them are fully grown by the middle of August. Galls collected on September 10 (at Inskip in 1925) were fully grown, but the insects were still so young they could not be bred after collecting. The larvae mature sometime after the first of Oc-tober (before October 4 at Kelseyville and October 16 at Di- ablo in 1925) ; and adults may be found in the galls soon afterwards. The pupal period is short, and for a time in October and November mature larvae, pupae, and immature adults may be obtained from a single collection.

Adults, not yet fully pigmented, were in the galls at Kelseyville on October 30, and on October 29 at Diablo in 1925. These adults do not begin emerging until the end of November (November 26 from Diablo material in 1925, and November 28 acc. McCracken coll.), or after the first of December (December 19-24 acc. Ashmead, and December 18-29 from Diablo material in 1922). Kelseyville material bred here at Bloom- ington, Indiana, emerged out-of-doors early in January (in 1926), and a few of the insects emerged as late as Febru- ary 1 (in 1927). In the field the galls collected by the last of January (at Diablo) are usually empty of insects except for the parasites and inquilines which emerge at later dates. One very large collection made at Diablo as early as December 18 (in 1922) was already empty of gall-making Cynipidae. The records of indoor breedings are usually later: January and February acc. McCracken, and January in my own experience; but it should be noted again that increased temperatures appear to delay emergence in many Cynipidae. The succeeding generation, a form named lobata, appears with the bursting of the buds in March or April. 

In November the agamic galls fall to the ground either attached to or separated from the leaves, or the galls remain attached to the leaves which hang on the trees over winter. The galls on the ground are decayed soon after the first of the year, but good specimens may still be found on the trees in the spring (as late as March 7 at Paso Robles in 1920). 

Ashmead accredited this insect to Quercus Douglasii, but the fragments of leaves with the type galls are clearly those of Q. lobata. The Douglasii record is copied in most of the literature, altho two students (Fullaway and McCracken) who have known this variety in the field have pointed out that the white oak, Q. lobata, is the real host. Of the several thousand galls which Mr. F. A. Leach collected for me from a number of localities and at all seasons over a period of four years, practically every one came from Q. lobata; but two meager collections, a total of 21 galls found in 1922 at Diablo, seem to have come from Q. Douglasii. My records show that for at least one of these collections I verified the host determination while noting its unique nature. It is, however, inconvenient that an insect that is all but confined to Q. lobata should carry the name douglasii, and the situation is the more unfortunate because the bisexual form of this same insect was named lobata by McCracken and Egbert. 

In distribution this variety for the most part parallels echinus, wherever the hosts of the two occur together. But while echinus is replaced by another variety at the higher elevations fringing the Great Valley, I would refer insect and gall material which I have from Lake County to typical douglasii. I have 46 fine insects from Kelseyville alone, and cannot find material differences between them and douglasii unless the Lake County material averages darker. In the consideration of the reality of life zones, perhaps this case should be emphasized as an instance where two, very closely related insects (douglasii and echinus) do not respond in the same way to the same geographic factors. 

If there are constant characters by which douglasii may be distinguished from echinus, no one has yet described them. Upon examining a large series of the insects, I find the color distinctions noted by Fullaway are highly variable. The distinctive form of the gall and the host seem to provide the only marks for recognizing this insect. It is surprising that no one has previously adjudged echinus and douglasii to be varieties of one species.

Cynips mellea 
agamic forms

GALL. — A small, globular, thick-walled leaf gall with the larval cell filling most of the interior. Monothalamous ; strictly spherical when fresh and sometimes when dried after maturity, often shrinking irregularly when dried; the gall sometimes drawn out at the point of attachment; the surface of the gall roughly, irregularly shagreened, densely covered with long, white or brown hairs when young (at least in some varieties), the hairs deciduous, leaving a stellate pubescence on older galls; the galls of all varieties finally naked (altho still appearing puberulent because of the shagreened surface; the naked galls whitish, flesh pink or pinkish brown in color; up to 7.0 mm., dried galls averaging about 4.0 mm. in diameter. The thin outer shell re-enforced with a more or less thin, solid mass of crystalline structure; the larval cell highly variable in size, often occupying most of the gall, the wall of the cell hardly differentiated from the solid, crystalline material. Attached to the veins, singly or in small clusters, on the upper or (less often) the under surfaces of leaves of white oaks of all groups. 

[Kinsey describes many varieties of this gall; rydbergiana, unica, and carolina are listed separately, following other literature. See paper for details on the others.]

Cynips mellea variety mellea
agamic form

Biorhiza mellea
Biorrhiza mellea
Sphaeroteras mellea
Diplolepis carolina err det Weld
Diplolepis unica err det Weld
Cynips mellea var C Kinsey

GALL. — Typical for the species, averaging small; on the leaves, often in small clusters, on Quercus stellata and its varieties, probably including the related Q. Chapmani and Q. Margaretta. 

RANGE. — Florida: Jacksonville (Ashmead; types). Greencove Springs, Ocala, Clearwater, St. Petersburg, and Daytona (galls, acc. Weld 1926). Probably restricted to Q. stellata and its varieties in the South- east, perhaps to Florida and adjacent parts of Georgia. Figure 51. 

We have no insects of this variety except those of Ashmead’s type collection. These adults were bred in February. His host record, Q. parvifolia , is ordinarily taken to mean Q. Chap- man i, an oak so closely related to Q. stellata that there is, as far as I know, no distinction in the cynipid faunas of the two. 

The galls of variety mellea cannot be separated from those of Carolina. All of the insects thus far bred from Coastal Plain material, from the Carolinas and north, have proved to be the long-winged, rufous and black variety Carolina , and Beutenmuller’s record of mellea in New Jersey, apparently based on galls alone, is probably a mis-determination. His published figures of mellea galls appear to have been based on this New Jersey material. 

The relationships of the short-winged mellea and the long-winged varieties here treated as belonging to the same species, and the position of mellea in our present subgenus, are discussed in the general treatment of Acraspis and in the introduction to our present species. In southern Georgia and Mississippi there is another short-winged variety, bifurca, which resembles mellea in many respects, altho it has a wing venation of such a distinctly different type that we must consider mellea and bifurca independent mutations from the long- winged stock of the species.

Cynips mellea variety carolina
agamic form

Dryophanta carolina
Biorhiza mellea err det Beutenmuller
Diplolepis carolina 
Cynips mellea var B Kinsey

GALL. — Woolly when young, becoming naked when mature, then light pinkish brown in color; averaging smaller than variety crassior; usually drawn out to a conical base; usually clustered, on the leaves of Quercus alba and Q. stellata. Figures 292-293, 334. 

RANGE. — New York: Nyack (Zabriskie in U.S. Nat. Mus.). New Jersey: Fort Lee and Lakehurst (acc. Beutenmiiller 1910). Richland and Carmel (Kinsey coll.). Virginia: Eastville and Cape Charles (Kinsey coll.). North Carolina: Asheville (Ashmead, types) . Maysville, Marshall, and Cherokee (Kinsey coll.). Kentucky: Dawson Springs (gall, Kinsey coll.). Tennessee: Newport (Kinsey coll.). Missouri: 6 miles west of Dexter (Kinsey coll.). Springfield (R. Voris in Kinsey coll.). Probably thruout the Coastal Plain of the northeastern United States, from southern New York to North Carolina and Tennessee and in the Mississippi Valley from southern Indiana to Missouri. Figure 54. 

This is the common eastern variety of the species, well known to all who have observed galls in the Atlantic Coastal Plain region. It occurs westward to the Great Smokies of the North Carolina-Tennessee boundary, and also at Newport, Tennessee, immediately west of the mountains. At higher elevations in the mountains themselves, from Ohio to northern Georgia and Alabama, variety litigans represents the species. In the heart of Georgia and Alabama, and at a few points further north, variety crassior, a close relative of Carolina, replaces the Coastal Plain variety. Extensive field work in Tennessee, Georgia, and Alabama failed to disclose Carolina occurring- west of the Appalachians until we found it in western Kentucky and southern Missouri where I have re- covered fine series of the insects. From west of Dexter, in the southeastern corner of Missouri, collections made on a small clump of Q. stellata scrub, located very near the fault that separates the river lowlands from the highlands, gave me forty insects, two of which represent the Ozark variety unica and 38 of which are typical Carolina. From Springfield, from near the southwestern corner of Missouri, I have 25 insects, 9 of which are unica and 16 of which are Carolina. The insect is so unique as to leave no possibility of misdetermination, and the galls of unica and Carolina are similarly distinct, so there can be no doubt of the occurrence of the Atlantic Coastal Plain variety all the way across the state of Missouri. 

Except for the type collections, all of the insect material I have examined has come from Q. stellata (the post oak), to which the variety seems largely confined. The type specimens recorded as from Q. alba seem no different from this Q. stellata material unless the insects average a little smaller in size. Beutenmuller’s record (in Smith, 1910, Ins. N.J. : 599) of similar galls on Q. prinoides (the chinquapin oak) , and specimens I have from this same oak from Roselle, N. J. (C. J. Long, Jr., coll.) will probably prove to belong to another variety. Weld’s record of Carolina from Q. stellata at Ironton in the Missouri Ozarks should be re-examined in connection with the numerous varieties now described. 

Beutenmuller records the galls as occurring in August, September, and October. I have found mature galls still attached to the leaves in the middle of October in southern New Jersey and Virginia, and at the end of October in the mountains of the western part of North Carolina and eastern Tennessee. The galls I collected at Dexter, Missouri, late in October (1926), were mature and dropping from the trees, but the insects did not emerge until the following March. I have bred other material early in January. Weld collected galls in October from which he reared adults in the following June, but this seems late emergence for a species in which the other known varieties emerge in the late winter or early spring. 

One of my insects from Marshall, North Carolina, seems identical with typical Carolina from the same locality except that it is almost entirely piceous in color. 

Cynips (Acraspis) conica, new species
agamic form

GALL. — A smooth, pointed, conical leaf gall with the broadened base bearing spiny projections. Regularly or irregularly conical, up to 7.0 mm. in height; or the tip rather sharply pointed (and sometimes curved) ; the base sometimes flaring, its edge entire or bearing up to 6 short, blunt, spiny projections; the base up to 6.0 mm. in diameter; entirely smooth and glossy, or in part finely puberulent, light pinkish brown, with a violet tinge due to the puberulence. Internally filled with a rather compact mass of branched fibers, the larval cell rounded, up to 2.5 mm. in diameter, located very near the base of the gall. Attached by a slightly projecting point on the middle of the broad base, on the veins of the under surfaces of leaves of Quercus grisea. Figures 288- 289, 336. 

RANGE — Arizona : Globe. Probably confined to a limited area including the country between Globe and Phoenix, Arizona. Figure 57. 

When the type material was collected at Globe on January 20 (in 1920), most of the galls showed the holes thru which the adult gall makers had previously emerged ; but three adults emerged a few days after that. Emergence thus occurs in mid-January, even tho the winter season is rather severe in the mountains near Globe. No further life-history data are available, but judging from the related species of the genus, the gall may be expected to appear early in the summer. Other varieties of conica may be expected in Arizona, New Mexico, and West Texas, and further south in Mexico. 

Cynips (Acraspis) nubila
agamic forms

GALL. — A mass of coarse hairs, containing a spherical, hard core. Monothalamous, the core averaging 4.0 to 7.0 mm., the entire gall averaging 10.0 to 15.0 mm., compacted clusters of several galls measuring up to 24.0 mm. in diameter. The central core strictly spherical except where drawn out basally at the point of attachment; the surface of the core crystalline, very rough, scurfy, covered with short, straw-colored, crystalline hairs among which are the close-set, long, fine, hair-like spines which form the tangled, woolly covering of the gall, these spines up to 8.0 mm. in length, unbranched, wavy. The central core hard, crystalline, brittle, with an outer wall fully 1.0 mm. in thickness; all of the rest of the core (up to 5.0 mm. in diameter) is hollow, the inner walls being rough, irregular, without any other larval cell. Occurring singly or, more often, in clusters of two to five galls, the clusters usually more or less hemispherical but sometimes more elongate when containing several galls; attached to the mid-veins, on the under surfaces of the leaves of Quercus arizonica, Q. oblongifolia, Q. Toumeyi, Q. reticulata, Q. glaucophylla) , and probably other related oaks. 

RANGE. — Known from Arizona and central Mexico. Probably widespread in Mexico, and to be expected in southwestern New Mexico. Figure 58. 

The gall of this species is one of the most attractive in the genus Cynips. The tangled mass of brilliantly colored, hair- like spines covering the spherical core marks the gall at some distance in the field. In collecting in January (1920) in the Santa Rita Mountains of southern Arizona, I camped at four thousand feet among trees on which nubila galls were common. A mountain storm had banked these evergreen oaks with wet snow, offering a background against which the wine-purple galls appeared too fantastic to be real anywhere but in the enchanted lands of our Southwest. 

The galls of the Arizona varieties probably appear early in the summer; they are full-grown by September, and the adults probably mature in another month or so. The insects, however, do not emerge until late in December or in January, during a season of cold weather and snow. Perhaps the more southern, Mexican variety, incompta, has a different life history. 

While nubila is common in Arizona south of Globe, it has not yet been recorded from north of there or from New Mexico. If must be very rare in these latter regions if it occurs in them at all. 

While the hairy gall of this species appears superficially different enough from the galls of most Cynips, it is nothing but a villosa gall with the spines much lengthened. The spines are inserted on the surfaces of the two in the same way, in two varieties of nubila they are enlarged basally, quite as they are in villosa, and in both species they are surrounded basally by peculiar, short, crystalline hairs. The resemblances are so striking that, once having seen them, one wonders why they are not apparent at first glance. 

[Kinsey describes three varieties of this species; only [Andricus incomptus](https://www.gallformers.org/gall/2909) is listed separately. See paper for details on the other two.]

Cynips nubila variety incompta
agamic form

Andricus incomptus
Diplolepis nubila err syn Weld

GALL. — The core yellowish to rich reddish russet in color, the spines for the most part yellowish white, the tips colored rich golden yellow; the bases of the spines not at all swollen; on leaves of what at least resembles Quercus reticulata and Q. glaucophylla. Figures 299-300. 

RANGE. — Mexico: San Luis Potosi (Palmer coll., types). Sierra de Zacopoaxtla (L. Diguet acc. Houard 1928). Probably confined to an area in more central Mexico. Figure 58. 

The galls collected in September were fully grown and the insects far enough along at that time to have matured after collecting. It is not impossible that the more southern varieties of Cynips have only one generation a year, as is certainly true of several of the southern California Cynipidae, and in such a case the growth of incompta may begin much earlier, tho proceeding more slowly, than is the case with the Arizona varieties of nubila. 

The leaves with the type material would now appear to represent Quercus reticulata (as restricted in Trelease’s mono- graph of the American Oaks) . I also have galls of what seems to be incompta collected by Bonansea in Mexico (without defi- nite locality) and sent me by Prof. Trotter of Portici, Italy. This material appears to be on Q. glaucophylla , but I cannot be certain of determinations based on so few leaves of such difficult Mexican oaks. 

Soon after the original publication of incompta , Beutenmuller, who had never seen the type material, wrote me that it was a synonym of nubila. Weld has recently published this synonomy with the following comments: “The writer has examined both [types] , comparing one directly with a Bassett type of nubila. As the galls were collected in September it is the writer's idea that at that time the nutritive layer had not been all used up and the larva had vitality to transform into an undersized adult but not enough to chew its way out of the hard gall. Never having been exposed to light and open air it is much paler than flies that emerge normally. Had it darkened up normally the two adjacent spots near the apex of the wing might have become connected into one double one as is the case in nubila , a series of which shows considerable difference in the amount of fusion that has taken place in the spots, due either to fluctuating variation or to the length of time that has elapsed or amount of exposure to light since emergence from the gall. If a nubila wing were bleached somewhat it would present the condition seen in incomptus.” 

My own distinction of nubila and incompta is herein embodied in comparative descriptions, an examination of which should show that it will take more than bleaching to turn nubila into incompta . While admitting the inadequacies of two specimens cut from old galls, and while admitting that fresh material may be larger and darker than the types, I cannot believe that the surface of the mesonotum, the length of the parapsidal grooves, the surfaces of the mesopleuron, the hairy areas on the abdomen, the size of the areolet, or other such distinctive characters are abnormal in the type material of incompta. The galls of incompta and nubila , while similar, are certainly distinct in color and, more significantly, in the shapes of the bases of the spines. 

The types of incompta come from a locality removed from the known range of nubila by eight hundred miles of Mexican desert and mountain country. This alone should have invited careful comparisons before it was concluded that the two names are synonyms. 

[Cynips villosa variety acraspiformis](https://www.biodiversitylibrary.org/page/53516882#page/353/mode/1up)
agamic form

Diplolepis acraspiformis

GALL. — Rosy red to rosy brown in color; up to 10.0 mm., averaging nearer 7.5 mm. in diameter, the spines up to 2.5 mm. in length, stiff, straight, stout, the bases up to 0.5 mm. in diameter, tapering rather gradually; the whole suggesting a sea urchin; on the leaves of Quercus undulata, Q. Toumeyi, Q. grisea, and Q. arizonica. Figures 304, 330. 

RANGE. — New Mexico: Blue Canyon west of Socorro (types, Weld coll.). Nogal Canyon south of Socorro (acc. Weld 1926). Kingston (Kinsey coll.). Hillsboro (galls, Kinsey coll.). Burro Mountains (galls, acc. Weld 1926). Arizona: Ashfork (galls, acc. Weld 1926). Prescott (Kinsey coll.; also W. W. Jones in Kinsey coll.). Safford (Kinsey coll.). Hackberry and Patagonia (acc. Weld 1926). Probably confined to the southern two-thirds of Arizona and New Mexico (and adjacent Mexico). Figure 59.

This variety has a wide range over the southern two-thirds of New Mexico and Arizona, and shows no segregation into distinct varieties on Q. undulata, Q. grisea, Q. arizonica, or Q. Toumeyi. 

Weld cut live adults out of galls collected near Socorro, New Mexico, on November 7 (in 1921), and had others emerge December 31 and January 13. One adult from Patagonia, Arizona, emerged December 13. From galls which I collected in Arizona at Safford on January 18 and at Prescott on January 23 (in 1920) all but the last adults had already emerged. The galls Mr. W. W. Jones collected in March (1924) were empty. 

The original spelling of the name of this variety should have been acraspidiformis, if we are to follow the declension of the Latin aspis, aspidis; but as the Rules stand, we accept the published spelling. 


[Cynips villosa variety alaria](https://www.biodiversitylibrary.org/page/53516882#page/357/mode/1up)
agamic form

Acraspis alaria

GALL. — Mature gall straw-yellow in color, up to 11.0 mm. in diameter, the spines up to 2.5 mm. in length, rather flexuous, slender, the whole gall appearing as a mass of coarse and tangled hairs; on the leaves of Quercus Gambelii, Q. submollis [synonym of gambelii], and probably related oaks. 

RANGE. — Colorado: Colorado Springs (Pollock in U.S. Nat. Mus.; types). New Mexico: Raton Pass near Raton (F. Cogshall in Kinsey coll.). 28 miles east of Raton (C. Schwachheim in Kinsey coll.). Probably restricted to a Rocky Mountain area in southern Colorado and northern New Mexico, probably not to be expected south of the Sandia Mountains in New Mexico. Figure 59.

This is the southern Rocky Mountain variety of the species, a short-winged insect replaced in the southern two-thirds of New Mexico and Arizona by the long-winged acraspiformis, but in Utah by the short-winged calvescens. The material collected by Frederick Cogshall near Raton, New Mexico, was full-sized late in July (1926), and the larva was large enough to complete development in spite of a prolonged drying to which its gall was subjected before reaching our laboratory. This insect emerged out-of-doors at Bloomington, Indiana, on December 18 (1926). The galls of the type material were collected at Colorado Springs on November 4 (1918), one adult emerging on November 11 while others were found alive in the breeding cage on the following January 15.

[Cynips villosa variety villosa](https://www.biodiversitylibrary.org/page/53516882#page/359/mode/1up)
agamic form

Acraspis villosus
Acraspis villosa
Philonix villosus
Philonix villosa

GALL. — Mature gall straw-yellow, staining browner in color; up to 13.0 mm. in diameter, the spines up to 2.0 mm. in length, rather flexuous, slender; the whole gall appearing as a dense mass of coarse and tangled hairs; on the leaves of Quercus macrocarpa. 

RANGE. — New York: Medina (acc. Weld 1926). Seneca Lake (Dudley in Cornell Univ., acc. Weld 1926). Ithaca (acc. Weld 1928). Michigan: Agricultural College (acc. Gillette 1889). Indiana: Crawfordsville (E. C. Stout in Kinsey coll.). Rogers in Pike County (gall, Kinsey coll.). Illinois: Evanston and Winnetka (acc. Weld 1926). Pana (Kin- sey coll.). Minnesota: Minneapolis (J. S. Benner in Kinsey coll.). Iowa: Ames (types, Gillette coll.). Kansas: Manhattan (thru C. V. Riley; in U.S. Nat. Mus.). Riley County (Marlatt in U.S. Nat. Mus., Kans. Agric. Coll., and Kinsey coll.) . Probably restricted to the range of Q. macrocarpa in the more northern Middle West, from northeastern Kansas into New York State. Figure 59. 

INQUILINE . — Synergus villosus Gillette (acc. Gillette 1890). 

This is the most eastern representative and the first-known variety of the species. It is usually not a common insect and it is not well represented in our collections. Superficially the gall appears as a tangled mass of coarse hairs, but the slender tips break off on handling and expose the broader bases of the spines, leaving the gall more like the agamic Cynips gemmula in appearance. The gall of villosa is hardly distinguishable from the galls of consocians and of alaria, from southern Kansas and the southern Rockies, respectively. Villosa insects are more rufous, with shorter wings than alaria and with a more slender thorax than consocians. 

Gillette recorded these galls as appearing in the mid-summer. Weld found pupae (in galls from Medina, N. Y.) on August 24 and September 4, and cut active adults out of galls early in October and (from northern Illinois material) on November 1. On October 23 (1926) the insect had already emerged from a gall I collected in southern Indiana. Gillette’s galls collected on October 20 gave adults on October 29. My material from Minneapolis emerged on November 10, from central Indiana on November 20, and from central Illinois on December 8. The Marlatt material from Riley County, Kansas, emerged in January (one specimen is labelled September!). 

[Cynips villosa variety consocians, new variety](https://www.biodiversitylibrary.org/page/53516882#page/361/mode/1up)
agamic form

GALL. — Mature gall straw-yellow, staining browner in color; up to 13.0 mm. in diameter, the spines up to 2.0 mm. in length, rather flexuous, slender; the whole gall appearing as a dense mass of coarse and tangled hairs; on the leaves of Quercus macrocarpa. 

RANGE. — Kansas : Winfield (types, R. Voris in Kinsey coll.) . Either an Ozark variety extending westward to eastern Kansas, or a Texas variety extending northward to southern Kansas; the material insufficient to allow more precise prediction. Figure 59. 

This is either the Ozark or the eastern Texas variety of the species. In addition to the type series from southeastern Kansas, I have some material without data other than the record that I collected it in 1919-20. It was therefore probably collected in the southern United States not further north than northern Texas or southern Mississippi. This suggests that consocians is an eastern Texas variety finding its northern limit in Kansas. The galls are identical with variety villosa, and the insects of consocians are very close to villosa and alaria. These three varieties would hardly be distinguished except for their distinct ranges and the fact that among these sub-apterous Cynipidae the simplification of structure has proceeded so far that one may use only a part of the characters available among long-winged insects. The type series of consocians is a fine collection large enough to insure constancy in the few characters on which the variety is established. 

Galls from the type locality were full-sized and contained very small larvae early in August (1927). By early September the larvae had grown enough to be suitable for collection for breeding. Adults emerged from these galls on November 20 and December 5 and 17 (1927).


[Cynips villosa variety apache, new variety](https://www.biodiversitylibrary.org/page/53516882#page/356/mode/1up)
agamic form

GALL. — Indistinguishable from that of variety acraspiformis (q.v.), the spines stiff, straight, stout, the whole suggesting a sea urchin; on leaves of Quercus grisea and Q. arizonica. Figure 303. 

RANGE. — Arizona: Globe (types, Kinsey coll.). Fish Creek on Apache Trail (gall, Kinsey coll.). Probably confined to a limited area east of Phoenix, Arizona. Figure 59. 

The holotype female seems to be a mature adult (unfortunately laden with glue in the mounting) ; the other female is small and not fully pigmented but otherwise agreeing with the holotype. The two specimens came from different oaks. The wings of both are shorter than those of Cynips dugesi brevipennata, tho not as short as those of a Philonix. The hypopygial spine is a little smaller than in the other two varieties of the species. 

The Apache Trail country of Arizona, the region in which apache is to be found, seems to have a cynipid fauna largely distinct from that of adjacent areas both north and south of Globe and Phoenix, but I have no data on the extent of this area. 

[Cynips villosa variety calvescens, new variety](https://www.biodiversitylibrary.org/page/53516882#page/358/mode/1up)
agamic form

GALL. — Mature gall straw-yellow in color, up to 14.0 mm. in diameter, the spines up to 2.5 mm. in length, rather flexuous, slender, the whole gall appearing as a mass of coarse and tangled hairs; on the leaves of Quercus utahensis [synonym of gambelii] (and related oaks?). Figures 307-308. 

RANGE. — Utah: Bountiful (types; B. and H. J. Pack in Kinsey coll.). Farmington, Santaquin, and Layton (B. and H. J. Pack in Kinsey coll.). Probably restricted to an area in more northeastern Utah. Figure 59. 

This is an apparently common insect in the mountains of northeastern Utah. Its closest relatives are alaria, which occurs in southern Colorado and northern New Mexico east of the Continental Divide, and the variety villosa , which ranges from Kansas into New York State. Calvescens is immediately distinguished by the large, naked area on the abdomen, the more eastern insects having the sides of their abdomens entirely hairy. The name calvescens (becoming bald) emphasizes the striking appearance of the new variety. This Utah insect also has a distinctly longer wing and averages shorter in body length than either of the other varieties. 

The Cynipidae of Utah, as far as I have studied them, are never the same as those of Colorado, and the present species serves to illustrate how distinct these faunas may be. We are indebted to Dr. H. J. Pack and his daughter, Bessie Pack, for all the material we have of this insect. The galls collected early in September (1927) seemed fully mature. Out- of-doors at Bloomington, Indiana, I bred 66 adults by December 5, another 76 by December 13, and another 9 at some later date (all in 1927). 

[Cynips villosa variety expositor, new variety](https://www.biodiversitylibrary.org/page/53516882#page/355/mode/1up)
agamic form

GALL. — Mature galls yellow-brown in color, up to 8.5 mm., aver- aging nearer 6.0 mm. in diameter; the spines up to 4.0 mm. in length, rather flexuous, slender, the bases up to 0.4 mm. in diameter, but the spine very slender right down to this base; the whole appearing as a tangled mass of coarse hairs; on leaves of Quercus grisea and Q. arizonica. Figures 301-302, 326. 

RANGE. — Texas: Alpine (types; Q. grisea , Kinsey coll.). New Mexico: Soledad Canyon in Organ Mts. (galls, Q. arizonica; L. H. Bridewell in Kinsey coll.). Probably restricted to the desert mountain areas of West Texas and immediately adjacent areas in New Mexico. Figure 59. 

The galls collected in the Organ Mts., New Mexico, on August 23 (1927) were too young to breed. The last adults were emerging from the galls which I collected at Alpine, Texas, on December 14 (1919). 

This insect is distinct from its widespread New Mexico and Arizona relatives, altho clearly related to them, and the gall is also readily distinguishable. Here is one more instance of the fact that the West Texas and adjacent New Mexico cynipid fauna does not extend further north or west in New Mexico. 

Cynips gemmula 
agamic forms [partial; excluding Acraspis prinoides. Tentatively assigned to Acraspis echini based on host.]

GALLS OF AGAMIC FORMS. — Moderately large, spherical, roughly faceted, with short spines. Usually monothalamous, often with two or three cells in variety cruenta; the normal galls strictly spherical, up to 17.0 mm. in diameter; the surfaces closely set with numerous, bluntly cone-shaped projections which usually terminate in short, stiff, and blunt spines; the cones plus the spines usually not over 1.5 mm. in length; the young galls light greenish or yellow tinged with red, the older galls light buff yellow to brown and blackish. Internally compact crystalline, the walls very thick, the entire gall consequently solid ex- cept for the centrally placed larval cell which is up to 2.2 mm. in diameter and without a distinct cell wall. Attached to either the upper or under surfaces, usually to the mid-rib but sometimes on lateral veins, on leaves of chestnut oaks (known from Q. prinoides, Q. Michauxii, Q. Muhlenbergii ) . 

[Kinsey goes on to describe several varieties on Quercus michauxii, with varying spine length. See paper for details.]

[Cynips pezomachoides variety echinoides, new name](https://www.biodiversitylibrary.org/page/53516882#page/416/mode/1up)
agamic form

Acraspis echini
Philonix echini

Gall.--Known only from the spiny form; on leaves of Quercus bicolor.

Range: FL

This insect is recognizable by its nearly uniformly bright rufous color, its more shortened parapsidal grooves, and its usually larger size. It is known only from the type material, and this is insufficient data to show whether the variety ranges on the swamp white oak, Q. bicolor, thruout the Atlantic Coastal Plain area, or along the Gulf into eastern Texas, or is confined to Florida. 

Ashmead described the galls as precisely similar to those of “erinacei,” meaning a spiny form of this species. The gall described and figured by Beutenmuller, perhaps from the type material in the Philadelphia Academy, is intermediate between the naked and spiny forms. There are no galls among the types in the National Museum.

Ashmead bred adults in November. He expressed surprise that he had found this species on Q. bicolor without finding it on Q. Prinus, for, he says, Q. bicolor “is considered by many botanists only a variety of Quercus prinus; galls found on one are very apt to be found on both, and insects are good botanists.” As a matter of fact, these insects are better judges of the relationships of the oaks than anyone who would consider Q. bicolor a variety of Q. Prinus. Long ago in the history of our insects there was a division on this very question, and Cynips gemmula and its varieties claimed the exclusive rights to Quercus Prinus, prinoides, Michauxii, and Muhlenbergii, while Cynips pezomachoides accepted Q. alba and bicolor. Modern botanists acknowledge this to be the correct grouping of these oaks (cf. Trelease, 1924, Nat. Acad. Sci. Mem. 20:102-111). Echinoides, as a variety of pezomachoides, is not departing from the ancestral traditions when it takes up its dwelling on Q. bicolor. 

Cynips gemmula variety gemmula
agamic form prinoides

Cynips prinoides
Philonix prinoides
Philonyx prinoides
Acraspis prinoides

GALL. — Averaging nearer 8.0 mm., altho up to 11.0 mm. in diameter; always monothalamous ; on leaves of Quercus prinoides, Q. Michauxii, Q. Muhlenbergii, and possibly other chestnut oaks. Figure 327. 

RANGE. — Massachusetts: Springfield (gall, acc. Stebbins 1910). Eastern part (bisex. form, Thompson in Bost. Soc. coll.). Connecticut: Waterbury (bisex. form, Bassett coll.). New York: Oriskany Falls (galls, J. A. Douglass acc. Felt 1912). Vicinity of New York City (acc. Beutenmiiller 1904). Staten Island (acc. Weld 1928). New Jersey: Lakehurst (Beutenmiiller and W. T. Davis in Kinsey coll.). Toms River (Beutenmiiller, types). Richland (Kinsey coll.). Virginia: Cape Charles (Kinsey coll.). Ohio: Cincinnati (gall, Braun in Amer. Mus.). Indiana: Bloomington (C. M. Kinsey coll.). Benham (G. F. Hyatt in Kinsey coll.). Aurora (Kinsey coll.). Kentucky: Lebanon (Kinsey coll.). Tennessee: Tazewell (gall, Kinsey coll.). Missouri: Ranken (Kinsey coll.). Occurring in the northeastern United States from Massachusetts to southern Indiana and Missouri, south in the mountains to Kentucky and Tennessee, confined to areas in which chestnut oaks are found. Figure 62. 

This is the common variety of the species over the northeastern quarter of the United States. I find no appreciable difference between the northern Coastal Plain material on Q. prinoides (the source of the type) and our Middle-Western series of the insect on Q. Michauxii and Q. Muhlenbergii. In this failure to differentiate a Coastal Plain variety gemmula is matched by Cynips fulvicollis. 

In common with all the other Cynips of this part of the country, gemmula shows an extreme degree of individual variation that may indicate a hybrid origin of the variety, possibly dating from the Pleistocene glaciation which projected northern varieties into the ranges of southern varieties and which ultimately, upon the glacial retreat, left a great territory where the hybrid individuals developed into hybrid populations deserving taxonomic recognition (see pp. 55 to 60) . The Ozark variety fuscata would appear to be the southern parent of gemmula, for gemmula intergrades into fuscata in southern Indiana (and probably elsewhere). Unlike the other species of Cynips involved in this area, the northern parent of gemmula is not known from any pure population in the north, possibly because most of the chestnut oaks, which are the hosts of our insect, do not range as far north as other white oaks. If they did extend further north in the former day, there may have been a northern parent of gemmula which does not now survive. Nevertheless, in the line series of gemmula which we have from Bloomington, in southern Indiana, there are a few individuals that are much smaller and darker, with smoother and more naked body surfaces. Since these are the very characters which distinguish the most northern varieties of Cynips pezomachoides, C. fulvicollis, C. hirta, C. folii, C. divisa , and C. longiventris, it is possible that these small insects of gemmula are Mendelian segregates from the hybrid stock, and that these represent the northern parent of gemmula. It is, on the other hand, possible that these small individuals represent local mutants. We have described them as variety suspecta. Out of the 606 insects we have bred from Bloomington material, we find 23 (= 3.8%) have this small and dark form, 22 (=3.6%) seem characteristic fuscata, and the remaining 92.6% are the variable hybrid series which is gemmula. 

The galls of the agamic gemmula are to be found late in August (acc. Beutenmuller) but they probably appear much earlier than this — perhaps late in June as with other species of Cynips. The galls were of full size at Lakehurst, New Jersey, on September 14 (1904, W. T. Davis in Kinsey coll.), and at Bloomington, Indiana, early in September (1928). The insects had not yet emerged from the galls I collected at Rich- land, New Jersey, on October 13, at Cape Charles, Virginia, on October 17 (both in 1919), and at Ranken, Missouri, on October 29 (in 1926). Material in the American Museum records unemerged adults in the galls on an October 26. Our emergence dates for this insect are December 1, 2, 4, 6, 8, 10, 16, 18, and 21. 

Cynips pezomachoides variety erinacei
bisexual form bicolens

Dryophanta erinacei
Cynips erinacei
Andricus fulvicollis form bicolens
Andricus erinacei bicolens
Acraspis erinacei
Diplolepis gemula err det Weld
Dryophanta erinacea

GALL. — As described for the species. A seed-like, egg-shaped, or more compressed, elongate cell in the buds of Quercus alba. 

RANGE. — Undoubtedly as indicated for the agamic form erinacei (fig. 68) ; in the northeastern quarter of the United States and southern Canada, from Maine and Minnesota to Maryland and central Missouri. Known definitely from: New York: Ithaca (acc. Triggerson 1914). D.C.: Washington (acc. Weld 1926). Ohio: Cincinnati (acc. Wieman 1915) . Indiana: Miller (acc. Weld 1926). Charlestown and Memphis (E. W. Spieth in Kinsey coll.). Clinton (Kinsey coll.). Illinois: Ravinia (types, Weld coll, in Mus. Comp. Zook). Glencoe, Winnetka, and Willow Springs (acc. Weld 1926). Kentucky: Paducah (variety? Kinsey coll.). 

While bicolens is unknown except from Triggerson’s and Wieman’s studies, Weld’s material, and my southern Indiana and Kentucky series, the insect is of course to be expected wherever the galls of the agamic generation occur in abun- dance. The galls of bicolens are not usually visible until after the buds have opened and, as Weld (1926), suggests, “The easiest way to rear the flies is to locate a tree in the fall well infested with the hedge hog gall and from this tree gather twigs in the spring just before the buds start, putting them in a bottle of water and setting the whole in a battery jar with a cloth over the top.” This bisexual insect is very close to gemmula , the bisexual form of prinoides; and the bisexual forms to be discovered for for the other eastern species and varieties of Acraspis will probably need careful comparisons in making determinations. Altho the agamic females have oviposited late in the fall or early winter, it is not until nearly six months later, after the middle of April, that the eggs hatch. Adults were mature and emerging from my southern Indiana material on April 22 and May 1, 8, and 9 (1927). Wieman (1915) secured adults at Cincinnati over a period of two weeks beginning April 23, 1914. Weld (1926) reports pupae near Chicago on April 28 (1913), and adults on May 1-20 (in 1913), on May 6 (1924 at Washington) , and May 17 (1909 near Chicago) . Triggerson, who first recognized on the basis of both laboratory and field observations, that this represents the bisexual generation of erinacei , did not find the eggs hatching to produce these bi- sexual insects at Ithaca, New York, until after May 8. Quot- ing from Triggerson : “On the twelfth of May a slight swelling, at the apex of which an empty egg shell was visible, appeared on the lower green portion of the scale . . . This proved to be a freshly formed gall, containing a young larva of Dryophanta erinacei. The gall at this stage was thin-walled, with a pebbled surface, greenish in color, and contained a watery fluid. The egg-shell remains attached to the apex of the gall until the latter has reached considerable size, when it dries up and disappears. These hypertrophies develop rapidly, as many as three ap- pearing on one scale. The wall of the gall has by this time changed to a yellowish brown color, and soon becomes quite dry and brittle. “Galls also develop on the apical portion of the leaf and flower buds . . . The terminal galls are of the same size as those on the scales, varying in number from one to four, and when mature are reddish brown . . . males and females similar in size and character emerge from the two galls. . . . The difference in color in the galls is due to the normal dif- ference of the tissue of which they are formed. “Shoots were brought into the laboratory, placed in water and covered with bell jars. Here about noon on the twenty- first of May the first male and female emerged. They were quite vigorous, and about four-thirty in the afternoon the fe- male was noticed actively moving along the midrib of the young leaf. Suddenly she stopped, and set up a rapid nodding motion which lasted thirty-five seconds, during which the ovi- positor was thrust into the tissue. The insect remained mo- tionless for a time, then withdrew the ovipositor, filling the passage with a yellow substance which, as in the agamic form, is probably a secretion poured forth by the accessory glands of the reproductive system. The process was repeated four times in succession without moving the body forward. Each time the ovipositor was inserted the body was curved slightly more than at the preceding puncture. The entire time oc- cupied by the four ovipositions was from four-thirty-four to four-fifty, or sixteen minutes, thus allowing four minutes to each oviposition of which a little over two minutes and a half was occupied by the passage of the egg. Many other observa- tions were made, and the time in all instances corresponded to the first recorded. “While the first observations of oviposition were made with- out having seen copulation occur, in all the following instances it was observed. The male strikes the female several times with the antennae after which the latter rests quiet. The male then clasps her thorax latero-caudad of the second pair of wings with the second pair of legs, while the first pair rest on the dorso cephalic portion of the thorax, and the third pair extend slightly latero-cephalad of the abdomen; copulation takes place, lasting for a few minutes. “The egg of the sexual form ... is oval, 160^. x450^. provided with a pedicel 750 in length, which is shorter than in the agamic form. It is always placed in the fibro-vas- cular bundles, and at an angle of about 80° to the axis of the leaf. The egg differs from that of the agamic form only in the elongate portion being shorter. “The larva is characteristic of the Cynipidae, having a slightly depressed head, fine needle-like mandibles, broad thorax, and reflexed pointed abdomen. During development the abdomen does not become as enlarged as in the agamic form. The thorax also continues prominent throughout all larval stages, which is not the case with the agamic form. . . . “In the open the adults did not emerge until the twenty-ninth of May, and continued to oviposit from that time until the fifth of June. Oviposition here was as observed in the labora- tory, the time occupied corresponding exactly to that already noted.” In this 1914 paper, Triggerson includes a small photograph of the female of this form, figures the young galls in the bud scales, details the mouth parts, wings, and antennae — the lat- ter showing sensory pits similar to those of the agamic female — and figures the egg of this insect. He also shows a longi- tudinal section thru a larva including a portion of the Mal- pighian tubule. Wieman (1915) has studied the spermatogenesis of the male of the bisexual erinacei. He admits his conclusions as some- what uncertain and perhaps open to question because of diffi- culty in seriation of his sections, and more especially fragmen- tary because the maturation stages of the eggs of this genera- tion and of the agamic female were not studied. In the male he found “but one true maturation division, namely that of the second spermatocyte. The first spermatocyte division is indi- cated by the pinching off of a small quantity of chromatin- free cytoplasm which forms the so-called polar body.” In the second division “it would seem that each chromosome is di- vided quantitatively by a longitudinal splitting; although it must be remembered that attempts at verifying this conclu- sion by studying the constituents of the daughter groups [were] not satisfactory owing to the tangled condition of the chromosomes.” Counts in the metaphase seemed to show 12 chromosomes as, presumably, the haploid number. Counts of somatic tissues of the male, although not conclusive, also seemed to give 12 as the normal number. Counts of female somatic tissues seemed near this same number, the best prep- arations, coming from follicle cells of the ovary, giving counts of 13, 14, and 13, respectively. Wieman’s findings are unexpected, for among many Hy- men optera the males are haploid because they are produced parthenogenetically from eggs in which reduction has oc- curred, but the females are diploid because, if they are pro- duced parthenogenetically, they come from eggs in which re- duction has not occurred. Doncaster’s work (1910-1916) seems to confirm this general statement for the cynipid species Neuroterus baccarum, and Patterson (1928) accepts this as the explanation of the demonstrated fact that agamic Cyn- ipidae are apparently of two sorts : those that produce females only (because no reduction occurs), and those that produce males only (because reduction does occur). If Wieman’s re- sults are acceptable, they suggest that the bisexual female is haploid as well as the male, for even if the accuracy of the somatic count on erinacei is open to question, the count falls short of the expected 2X number of 24 chromosomes. Wieman suggests that the slight difference between the male and female counts in this species is due to the presence of true sex chromosomes in the female. Either Doncaster or Wieman is wrong in the cytologic findings, or, if both are correct, there is more than one type of maturation involved with the alternation of generations of diverse species of Cynipidae. It is highly desirable that Wieman’s brief investigations be ex- tended by further work on erinacei. 

Cynips pezomachoides 
agamic forms

[Kinsey treats [Acraspis erinacei](https://www.gallformers.org/gall/838) as a variety of pezomachoides, and the presence or absence of spines is not diagnostic for most of his varieties. The description below may not apply fully to pezomachoides if erinacei is excluded.]

GALL OF AGAMIC FORMS. — Small to moderately large, spherical to ellipsoid, naked and faceted to spiny. Monothalamous or agglomerate or polythalamous with 2 to 8 cells. More nearly spherical galls usually smaller, occasionally larger, up to 7.5 mm. in diameter; these galls usually with one or at the most two cells. The polythalamous galls more ellipsoid, occasionally showing lines of fusion of two or more smaller galls; up to 20.0 mm. in length. The surfaces of the galls closely set with broad, polyhedral bodies, giving a faceted appearance, these without spiny tips or with short, spiny tips, or with long and flexuous tips, the galls in consequence varying from entirely naked to roughly faceted to densely spiny. Young galls light green and rose-tinged, the spines (if present) often brilliant purplish red; the older galls be- coming straw yellow to brown and darker, the spines (if present) be- coming dark purple. Internally compact crystalline; the walls thick, flexible when moist, very hard when dry; the larval cavity without a distinctive cell wall, the cavity central if single, the cavities side by side if there are two, the alignment somewhat irregular if there are more than two. Attached to the midveins or lateral veins or on the petioles; on either the upper or under sides of the leaves of the true white oaks (known from Quercus alba, Q. bicolor, and Q. Gambelii). 

[Kinsey goes on to describe 8 agamic varieties of pezomachoides: 

[cincturata](https://www.biodiversitylibrary.org/page/53516882#page/381/mode/1up), known from wasps collected in Colorado but no gall, tentatively associated with Quercus gambelii

[ozark](https://www.biodiversitylibrary.org/page/53516882#page/382/mode/1up), known from galls ranging from naked to spiny on Quercus alba in OH, WV, IN, IL, KY, MO, AR

[wheeleri](https://www.biodiversitylibrary.org/page/53516882#page/384/mode/1up), known from galls ranging from naked to spiny on Quercus alba in NH, MA, PA, OH, MD, WV, MI, IN, IL, MO, VA, KY, TN, NC, GA

[pezomachoides](https://www.biodiversitylibrary.org/page/53516882#page/390/mode/1up), known from galls lacking spines on Quercus alba in NH, MA, CT, NY, NJ, PA, MD, DC, VA,NC, SC, TN, GA, AL, FL

[derivatus](https://www.biodiversitylibrary.org/page/53516882#page/395/mode/1up), known from galls lacking spines on Quercus alba in PA, OH, WV, IN, IL, KY, TN, SC, GA, AL

[erinacei](https://www.biodiversitylibrary.org/page/53516882#page/397/mode/1up), known from galls either naked or spiny on Quercus alba, covering the northern midwest to New England

[advena](https://www.biodiversitylibrary.org/page/53516882#page/413/mode/1up), known from naked to spiny galls on Quercus alba in KY, TN, NC, SC, GA

[echinoides](https://www.biodiversitylibrary.org/page/53516882#page/416/mode/1up), known from spiny galls on Quercus bicolor--likely a mistake, see ID Notes for [Acraspis echini](https://www.gallformers.org/gall/931)]

Cynips bella var congesta, new variety
agamic form

GALL. — As described for the species, apparently indistinguishable from the galls of the other varieties of bella or from the gall of Cynips dugesi pupoides which occurs in the same region. On leaves of Quercus grisea.

RANGE. — Texas: Alpine and Fort Davis (Kinsey coll.). Probably confined to the mountain ranges of western Texas (and adjacent regions?). Figure 45.

These insects were emerging on December 14 (in 1919) at Alpine, Texas, and two days later at Fort Davis, Texas. Cynips dugesi pupoides, whose galls also occur at Alpine and Fort Davis, is a small, short-winged insect that cannot be confused with congesta.

## *Cynips centricola* variety *clivorum*, new variety  
**Agamic form**  
**Figures**: 48, 258–259, 269, 285  

### FEMALE  
Head, entire antenna, and thorax usually entirely black; median groove poorly indicated or absent. The tip of the second abscissa of the radius is enlarged but not triangulate. The marks in the cubital cell are usually much fused (Figs. 269, 285).

### GALL  
Always unspotted; occurs on leaves of *Quercus stellata* (Figs. 258–259).

### RANGE  
- **Ohio**: Coolville (types, Kinsey coll.), Chillicothe (Kinsey coll.)  
- **West Virginia**: Parkersburg (Kinsey coll.)  
- **Tennessee**: Tazewell, Clinch River near Maynardsville, and Charleston (galls, Kinsey coll.); Tellicoe River near Madisonville, and 9 miles northeast of Chattanooga (includes some hybrids with *centricola*, Kinsey coll.)  
- **North Carolina**: Hendersonville (includes some hybrids with *centricola*, Kinsey coll.)  
- **South Carolina**: Travellers Rest (Kinsey coll.)  
- **Georgia**: 6 miles north of Trion, and Rome (Kinsey coll.); Acworth (gall, Kinsey coll.)  
- **Alabama**: Athens (gall, Kinsey coll.)  

Probably confined to a limited area on the western side of the Southern Appalachians (Fig. 48).

### TYPES  
20 females and many galls in the Kinsey collection from Coolville, Ohio; galls collected October 24, 1928; insects emerged November 13 and December 1, 9, 16, and 20, 1928; *Quercus stellata*; Kinsey collector.

### NOTES  
Throughout the more southern portion of the Appalachians, from the foothills in southern Ohio southward into Georgia and Alabama, this is the variety of *centricola* to be expected wherever *Quercus stellata* occurs. This region is poorly known biologically, and much of its cynipid fauna remains undescribed.

At lower elevations both east and west of the Great Smokies, *clivorum* appears to hybridize with variety *centricola*, and such hybridization may be expected wherever the two overlap. *Clivorum* is distinguished by the insect characters described above and by its consistently unspotted gall. While the gall of variety *centricola* may also be unspotted when immature, it is typically well covered with bright purplish spots.

Insects of this variety were bred on November 2 and 13, December 1, 9, 15, 16, 18, and 20, and January 2 (all in 1928). Despite the early onset of winter in much of its range, emergence timing is nearly identical to that of *centricola* and *strians*, both of which occur in warmer regions.

## *Cynips centricola* variety *strians*, new variety  
**Agamic form**  
**Figures**: 48, 256–257, 286  

**Synonymy**  
*Diplolepis centricola* err. det. Weld (in large part), 1926, *Proc. U.S. Nat. Mus.* 68(10): 23.

### FEMALE  
Largely black but often with a rufous tinge on the thorax and even more rufous on the head. Median groove almost always well evident for half the length of the mesonotum, deeply cut posteriorly. The tip of the second abscissa of the radius is slightly triangulate but not broadly so as in variety *centricola*. The marks in the cubital cell are not abundant nor large (Fig. 286).

### GALL  
Almost always well spotted with brownish purple; occurs on leaves of *Quercus stellata* (Figs. 256–257).

### RANGE  
- **Illinois**: Bonnie (types, Kinsey coll.), Christopher, Norris City, Bloomfield (Johnson County), Eddyville (hybrid with *centricola*; O. Buchanan in Kinsey coll.)  
- **Missouri**: Annapolis (E. S. Anderson in Kinsey coll.), Arcadia (gall, Kinsey coll.), Ironton and Poplar Bluff (acc. Weld 1926)  
- **Arkansas**: Hoxie (gall, acc. Weld 1926)

Probably restricted to the Ozark areas in Arkansas and Missouri, and to the eastward extensions of that area in Illinois and possibly western Kentucky and southern Indiana (Fig. 48).

### TYPES  
11 females and many galls. Holotype and paratype females and galls in the Kinsey collection. Paratype females and galls also in the American Museum and the U.S. National Museum. Labelled Bonnie, Illinois; gall collected October 15, 1927; *Q. stellata*; Kinsey collector.

### NOTES  
This is the Ozark variety of the species, known from the true Ozarks of Arkansas and Missouri, but, like many other Ozark Cynipidae, it is also common in the hill country of Illinois and expected in western Kentucky and southwestern Indiana, which are geologically related to the Ozarks. The region is bisected by the Mississippi and lower Ohio Valleys, where the Coastal Plain variety *centricola* also occurs. Since both *centricola* and *strians* have similarly spotted galls, material from this region must be determined based on insect characters.  

*Strians* is the only variety of the species which shows a well-developed median groove. Weld found pupae in galls likely of this variety on October 10. Kinsey bred adults on December 10 and 20 (1927).

The Gall Wasp Genus Cynips