Andricus lustrans (agamic)

Xanthoteras ornatum (sexgen)

Loxaulus pattersoni (agamic)

Disholcaspis simulata (agamic)

Diplolepis bassetti

Heteroecus malus (agamic)

Heteroecus melanoderma (agamic)

Heteroecus pacificus (agamic)

Heteroecus chrysolepidis (agamic)

Neuroterus quercicola (pacificus) (agamic)

Andricus quercuscalifornicus (agamic)

Disholandricus lasius (agamic)

Callirhytis quercusoperator (agamic)

Callirhytis quercusoperator (sexgen)

Callirhytis ovata (agamic)

Amphibolips quercuspomiformis (agamic)

Cynips quercusechinus (sexgen)

Protobalandricus spectabilis (agamic)

Loxaulus brunneus (agamic)

Loxaulus atrior (agamic)

Cynips mirabilis (agamic)

Diastrophus kincaidii

Diastrophus austrior

Diplolepis fusiformans

Diplolepis radicum

Diplolepis utahensis

Diplolepis variabilis

Disholcaspis pattersoni (agamic)

Heteroecus bakeri (agamic)

Heteroecus dasydactyli (agamic)

Heteroecus pygmaeus (agamic)

Heteroecus sanctae-clarae (agamic)

Neuroterus evanescens (sexgen)

Neuroterus floricola (sexgen)

Dryocosmus asymmetricus (agamic)

Callirhytis quercusbatatoides (agamic)

Andricus chrysolepidicola (agamic)

Kokkocynips coxii (agamic)

Callirhytis elongata (agamic)

Callirhytis frequens (agamic)

Callirhytis perdens (agamic)

Callirhytis quercussuttoni (agamic)

Andricus dimorphus variety verifactor, new variety [Based on Kinsey's description of the gall and Weld's synonymy of the wasp, the association of this name with vacciniiformis is likely mistaken and this gall is most likely Callirhytis vaccinii]

Cynips vacciniiformis (gall only!)

GALL. — Clustered, seed-like leaf galls. Each gall monothalamous, elongate, rather cylindrical, urn-shape, broadest at the middle, less broad apically, flattened at the end, taper-pointed basally, up to 4.0 mm. in diameter by 6.0 mm. in length; colored dark green when young, becoming a dark red or purplish red when old. Mostly solid and fleshy when young, becoming hard, thin-walled, and hollow when old, without a distinct larval cell lining. In compact clusters of up to 30 galls, attached to the midrib, on the under sides of leaves of Quercus stellata (and Q. breviloba). 

RANGE.— Texas: Buffalo, Hearne, Elgin, (Leander?), Austin. Probably thruout more eastern Texas and a part of Oklahoma and Louisiana. 

This gall is very common on the post oaks in Texas. Patterson states that the punctures from which the galls will arise may be detected about the first of May, that the galls do not develop from the scars until about the middle of July, that the galls are fully grown in size by the first of October, and in a couple of weeks most of them fall to the ground. I have collected them in late November and December, but the larvae are then still so small that they do not mature after becoming dry. Evidently they need to be kept moist, as they are when lying on the ground, to allow the insect to develop. Patterson secured over a hundred adults which emerged from February 12, 1922, to March 8. Inasmuch as the breeding of the insect is difficult unless carefully handled on the field, we are considerably indebted to Dr. Patterson for successfully rearing the adult. 

I collected the galls, but did not obtain the insects from the other Texas localities listed. It is possible but not probable that other varieties occur at some of those points. The gall occurs on Q. breviloba at Leander, and Patterson reports it as occasionally on breviloba at Austin. It is not unlikely that the breviloba insect is a distinct variety with a range centering about Burnett County, Texas. 

This is certainly a variety of Andricus dimorphus, described as a Cynips (Beutenmuller, 1913, Trans. Amer. Ent. Soc, XXXIX, p. 245) from galls taken by Weld at Evanston, Illinois, on Q. macrocarpa. Similar galls have been recorded and are common on white oaks of many species from the whole of eastern United States. Unfortunately I cannot examine the types of dimorphus, and cannot furnish a comparative description. Cynips vacciniiformis was described from Q. stellata galls which are undoubtedly those of verifactor, and from Austin. But the description of the single insect obtained is far from correct for the true producer of this gall, and apparently applies to a different species mistakenly supposed to have come from verifactor galls. Vacciniiformis was described in 1913. Of course insects bred from galls collected in 1917 and now labelled cotypes in at least a couple of collections, cannot have any standing as type material. 

Trigonaspis ornata, new species

GALL. — Spindle-shaped, covered with a golden brown mass of short filaments; 6. mm. wide by 11. mm. long, widest slightly above the middle; distinctly tipped apically. Covered with a densely compacted mass of filaments, each filament short, wholly flattened with a slender, narrow blade 2. mm. long, the tips purplish when fresh. The central stem is swollen to form a thin-walled, empty larval chamber, apparently monothalamous. A bud gall, or on leaves, attached to the end of the mid-rib; on Quercus breviloba. 

RANGE.— Texas: Austin (Patterson). 

The galls were collected March 3, and adults emerged April 15. In this country the genus has been known previously only from root galls, tho in Europe it is obtained from leaf and stem galls also. 

The galls superficially resemble those of Neuroterus evanescens Kinsey, described in this paper, but the anthers in that gall are not particularly modified. The adult of this species closely resembles Trigonaspis radicola (Ashmead); the female of ornata differs in being more brilliant rufous in color, slightly smaller, the antennae are distinctly more slender, the scutellum is less rugose with the basal depression smooth (rugose in radicola), and the wing veins are much heavier, even the cubitus being heavy (quite faint in radicola). A further study of material from more localities and hosts may show that this is a variety of radicola, which occurs on Q. alba, and was first described from Missouri. Ornata will prove a distinct variety at least. 

The connection of ornata and radicola prompts a suggestion concerning the life history of the species, which though resting on circumstantial evidence, may be hazarded if it is taken only as an hypothesis. Brodie (1896, Ann. Rpt. Clerk Board Forestry, Ont., pp. 114-116) records Biorhiza forticornis ovipositing on the rootlets of Q. alba. In a 1920 paper (Bull. Amer. Mus. Nat. Hist., XLII, p. 374), I recorded the observation of Biorhiza forticornis (which is wingless, agamic, coming from terminal twig galls on Q. alba) ovipositing in December at the roots of Q. alba. I have observed galls of Trigonaspis radicola on the roots of trees which bore old galls of forticornis, and recently Weld (Proc. U. S. Nat. Mus., LIX, pp. 203-204) notes thirty-six instances of coincidence of the two galls on trees of Q. alba. There is considerable circumstantial evidence, then, of the cyclic alternation of Biorhiza forticornis and Trigonaspis radicola. In the paper above mentioned I further described a March brood of forticornis, obtained from the same galls as the December brood, ovipositing in the terminal buds of the tree; no galls were obtained from these buds in the experiments. Is it possible that the ornata gall (or more probably a similar variety on Q. alba) is the bud gall from eggs of the March brood of forticornis Are there two interlocking cycles in the life of this species? What is the solution of the heredity questions presented? Such are the alluring problems to be solved only after the cycle has been experimentally investigated. 

Compsodryoxenus pattersoni, new species

Gall.--A slight but distinct swelling of small roots. Polythalamous. Galls gradual enlargements from the stems, elongate, averaging 20 mm long by 4 mm wide, often larger, covered with natural or slightly roughened bark. Internally woody, but the tissue irregular, with the larval cells closely clustered, the lining not separable nor hardly distinct, each cell spherical, about 1 mm in diameter. On roots of Quercus virginiana. 

Range.--TX

One of the galls is constricted into two parts, the parts bent at an angle; the other galls are straight, undivided; the two types of galls give the same kinds of adults. These galls contained live adults in December; insects emerged on February 9. It is not impossible that adults are mature in the galls over the winter, not emerging until early in the following spring. Similar but much smaller swellings at the bases of petioles had adults emerge on Dec. 27 which appear exactly identical with those from the root galls. Some sort of alternation of generations may be involved here, but it is not unlikely that the species will produce galls on various parts of the plant, not being confined to the roots only. 

Disholcaspis simulata, new species

Hosts: Quercus dumosa, garryana

Gall. Rounded bullet gall with a nipple tip. Monothalamous. Fairly globular, averaging 12.-15. mm. in diameter, some specimens larger or smaller, bearing at the summit a short, blunt point; colored light brick red or yellowish brown, weathering dark; the surface is very rough, but finely so. Internally densely but not compactly woody, likely solid when young, becoming hollow when mature, but with the walls still thick; containing a thin-walled, hard-shelled, entirely loose larval cell averaging 3. mm. in diameter by 4.2 mm. long. Galls attached by a tongue in the concave base; laterally on twigs of white oaks.

RANGE: Oregon, California. Probably wherever oaks occur on the Pacific Coast.

Disholcaspis simulata variety simulata, new variety

GALL. — Mostly colored light brick red to a darker purplish red, becoming lighter and browner on aging, or weathering dark; on twigs of Quercus dumosa. 

RANGE. — California: Fallbrook, Sorrento, San Jacinto Mountains, Upland, Pasadena, Santa Barbara, Paso Robles. Probably occurs thru- out the southern Sierras and their extensions, from El Portal south. 

Galls collected in February and March, 1920, contained live adults in November and December 1920 and in March 1921 ; several adults had emerged before March 1921. It is not unlikely that it is a year and a half or more after hatching before the adult emerges from the gall. Externally some specimens, duller-colored, of this gall will be confused with less distinctly colored specimens of Disholcaspis plumbella Kinsey; both species occur in enormous abundance on the same host, Quercus dumosa; the adults of the two are very distinct. 

Disholcaspis simulata variety vancouverensis, new variety 

GALL. — Mostly colored light buff to yellowish brown, in part tinged rose red (not brick red!) when younger, weathering darker; on twigs of Quercus garryana. 

RANGE. — Oregon: Roseburg, Grants Pass, Ashland. California: Yreka. North of the other varieties. 

Tho structurally the two varieties are very much alike, their distinct color, correlated with the distinct hosts and ranges makes it important to distinguish two separated tendencies in evolution. This variety is likely confined to the geographic area of the more northern Pacific coast called **Vancouveran" by Van Dyke (1919, Ann. Ent. Soc. Amer., XII, p. 4) . Another variety occurs on Q. douglasii in central California, and still another distinct variety in the San Bernardino Mountains, but I have not seen adults from galls of these. 

Diplolepis bassetti (Beutenmuller)

GALL. — Mossy, containing a compact or agglomerated cluster of woody cells. Filamentous covering 4.0-8.0 mm. thick, probably green and reddish when young, becoming brown and blackened with age; filaments attached to a more or less compacted cluster of larval cells, each cell sub-spherical, 2.0-4.0 mm. in diameter, with woody but not thick walls, entirely hollow within. On the twigs, usually terminally, of Rosa nutkana. 

RANGE. — Oregon, Idaho, and Utah. Probably elsewhere in the northwest. 

Morphologically the insects of this species are decidedly related to Diplolepis rosae, and less closely to D. bicolor. Significantly, the galls of variety bassetti show striking similarity to those of rosae. The galls of the other variety, tho superficially similar to rosae, are in all respects merely a compact cluster of galls of the D. bicolor type, differing mainly in possessing the filamentous covering. The evolutionary transition from the galls bicolor to rosae, or the reverse procedure, cannot have been more profound than the evolution of the two types of galls in this one species. 

There will be no question that this is a single species, for the characters separating the two varieties are very few. But if one has any doubt of the validity of the varieties he should examine a large series of individuals, when he should be impressed with the constancy of the distinctions within these few characters. Geographic isolation has probably contributed materially to the separation of two forms, for as far as the ranges of the two are known at present they are separated by the high altitudes of the barren lava-bed country of western Oregon. The type localities of the two are two hundred and fifty miles apart. Moreover, variety bassetti occurs in a region of excessive precipitation and tempered seasons, while the other variety occurs in the very arid deserts of the northern Rocky Mountains, a country of extreme climate. I should not consider this a variety of Diplolepis rosae, tho the two are very close. Rosae seems to be a native of Europe, but has more or less of a world-wide distribution, being readily imported into new regions; and apparently the species reached the eastern part of the United States as an importation from Europe. Rosae has not been recorded from west of the Rockies. No species native to America is very closely related to rose except this Pacific Coast species, bassetti. There may be considerable significance in this situation. 

Diplolepis bassetti variety bassetti (Beutenmuller) 

Rhodites bassetti Beutenmuller, 1918, Can. Ent., L, p. 307, pi. IX, figs. 13, 14. 

GALL. — Superficially resembles the gall of the other variety, but is distinct as follows: Filaments finer, more dense and curled; often a hundred or more larval cells in a cluster, the cells thoroly fused together, forming a perfectly solid, woody mass; galls averaging large, often up to. 45. mm. in diameter. 

RANGE. — Oregon: Corvallis (Beutenmuller) ; Ashland. 

Beutenmuller states that galls collected in December gave adults from late February to the middle of March, indoors. Most of the adults had not yet emerged from the galls collected at Ashland on April 6. 

Diplolepis bassetti variety lucida, new variety

GALL. — Superficially resembles the gall of the other variety, but is distinct as follows: Filaments heavier, broader, straighter; usually only 30 or fewer larval cells in a cluster, each cell more or less entirely separate, with a proximate origin from the stem, the cells compacted into an insecure mass more by the intertwining filaments than by a fusion of the wood of each cell; the whole mass averaging smaller than in the other variety, 25. mm. the maximum diameter noted. 

RANGE. — Oregon: La Grande. Idaho: Mountain Home. 

Few of the adults had emerged on April 12 at La Grande, and on April 20 at Mountain Home Of 267 adults, 100 are males. 

The collection at Mountain Home, Idaho, has peculiar interest because that locality is in the heart of the Snake River desert. This is an alkaline, dust desert, barren of most vegetation except sage brush, with an excessively low precipitation ; because of its elevation of about 3500 feet, and its great, level sweep, the region is exposed to severe storms and extreme temperatures in winter, and to a surprisingly hot summer. Repeated searchings at several localities across this desert failed to disclose any plant which might bear Cynipidae. The oaks, mainly eastern species, planted as wind breaks in the towns, failed to disclose galls. Some distance outside of the town of Mountain Home, a few roses were growing beside an irrigation ditch, and there were two galls of this species and variety. I do not know whether the roses were native species or escapes from cultivation ; they bore a native cynipid. How far it may have been to the next colony of roses it would be hard to determine; I am inclined to believe it might have been a considerable distance. Insects and galls of this mate- rial agree in minute detail with material from La Grande, two hundred miles away, mostly across the desert. The desert does possess a system of streams, but roses are few at most along these streams, for I failed to find them at any of a half dozen other points at which I tried to collect. 

I collected galls of this type at Brigham, Utah, but did not obtain insects, so I cannot say whether yet another variety will be found for this species. 

Heteroecus malus, new species

Hosts: Quercus chrysolepis

GALL. — Ovoid to almost absolutely spherical, 8.-16. mm., in diameter, only very slightly drawn into a point for attachment basally, and with a very fine, very short point apically 1.0 mm. or much less in length; entirely smooth and naked, or the surface slightly roughened, light to dark green when fresh, tinged with rose red, upon aging becoming brown. Internally more or less solid, with a single larval cell centrally, closely embedded, and between the cell and the base only a discontinuous cavity indicating the usually cylindrical cavity of the species. Singly or in compact clusters of a dozen, more or less, laterally or terminally on twigs of Quercus chrysolepis.

RANGE. — California: Pasadena, San Bernardino.

This variety comes near bakeri, but is distinct enough. The manner of oviposition, many eggs being laid in a restricted area, is distinct from bakeri. Material from the San Bernardino Mountains is remarkably uniform in respect to both the insect and the gall. Material from outside the San Bernardinos, that is, from another faunal area, shows some considerable variation, a part of the Pasadena material fully matching that from the San Bernardinos. I cannot decide whether there are other varieties of this species until I can examine more material. It is not impossible that this should be considered a variety of H. chrysolepidis. 

Heteroecus melanoderma, new species

Andricus dasydactyli (gall only!)

Elongate, scurfy, like a date seed in shape. Cylindrical, averaging 6. mm. wide by 20. mm. long, bluntly, short, and conically pointed apically, slightly tapered basally, broadest near the apex; covered with a dense, very short, scurfy pubescence, rich golden brown, weathering dull brown to black. Monothalamous. Internally with a single, cylindrical, central cavity 2. mm. wide extending from the base almost to the tip of the gall; a larval cell, 1.7x3.5 mm., lies in the cavity, usually nearer the apical end. 

Sessile, on twigs of Quercus chrysolepis.

RANGE. — California: Cupertino (Fullaway) ; Los Gatos (Mc- Cracken coll.) ; Redwood Park (in Stanford Univ. coll.) ; Boulder Creek, Placerville. 

Adults were almost mature in the galls at Placerville, March 30, and emerged some time later. 

Material of this species in several museums is labelled Andricus dasydactyli, and this gall largely agrees with the gall described for that species by Ashmead. But the insect dasydactyli comes from galls covered with long wool, and is very distinct from melanoderma adults ; the latter species has been heretofore undescribed. Weld sends me galls from the Santa Catalina Mountains of Arizona, occurring on Quercus ohlongifolia; the galls are similar to those of melanoderma, but I have not yet seen the insects. 

Heteroecus pacificus

Naked, elongate, spindle-shaped. Cylindrical, with a constricted base and a greatly elongate, slender tip; entirely straight or curved; entirely naked, smooth, bright green splotched with brown to purple-brown, becoming dark brown upon aging. Solid, except for the cylindrical cavity extending from the base two-thirds to the tip, the larval cell not always at the very end of the cavity. On twigs of Quercus chrysolepis.

 RANGE. — California: San Jacinto Mountains to Dunsmuir and Ukiah; probably wherever Q. chrysolepis occurs. 

This species morphologically is closely related to dasydactyli, but altho Ashmead stated that the insects of pacificus ''cannot be separated from A. dasydactyli", his types of the two are distinct, showing differences which agree with those in the material I have bred from the two types of galls. Pacificus never bears any of the woolly covering which is characteristic of dasydactyli. Adult insects emerge in the spring, a couple of months earlier in southern than in northern California. 

[Kinsey describes three varieties of this species; see paper for details]

Heteroecus chrysolepidis (Ashmead)

Andricus chrysolepidis

Hosts: Quercus chrysolepis

Gall: A small, irregularly globular or ovate bud gall. Monothalamous. A slight nipple at the apex; the surface covered with a fine, scurfy pubescence, light brown in color; 7 mm in diameter by 10 mm long.

Range: CA

The insect is very close to [dasydactyli](https://www.gallformers.org/gall/1903). The gall is the most distinctive thing of the species. Galls from Diablo appear to match the type galls, but I do not have insects to check the determination. Unlike the type galls, some of the Diablo specimens are clustered, suggesting [H. malus](https://www.gallformers.org/gall/1731), and inasmuch as the insects of the two are rather similar perhaps malus should be considered a variety of chrysolepidis. 

Neuroterus cupulae, new species

On Quercus lobata. 

GALL. — Only a larval cell buried in the wood of the acorn cup. The cell oval, averaging 1.5 mm. wide by 2.2 mm. long; the walls hard but not thick; entirely hollow. Buried wholly or in part in the wood of the acorn cup, oftenest at the base; part of the cell sometimes visible within the cup, only rarely producing a slight swelling on the outside of the cup.

Adults were emerging at Paso Robles on March 7, after most of the cupules had fallen to the ground. This is clearly an agamic generation, and very probably has an alternate, bisexual generation, probably not occurring on acorn cups. A similar gall occurs on Q. garryana on the north Pacific Coast; the insects will probably prove to belong to a distinct variety. 

Range: CA, Paso Robles, Gilroy

Andricus californicus

Hosts: Quercus lobata, douglasii, garryana, dumosa

GALL. — Large, globose to reniform, smooth, twig gall. Monothalamous, or polythalamous, containing from one to a dozen larval cells. Irregularly rounded, ovate, ellipsoid to reniform, all sizes, up to 65. mm. in diameter by 113. mm. long; the surface smooth, sparingly, shallowly roughened; reddish to yellow and brown. Internally more or less filled with compacted, soft, crystalline, brown, yellow, or white material, woody from the point of attachment to above the center of the gall; the larval cells oval, averaging 4.-5. mm. long by 3. mm. wide, more or less radiantly in the woody tissue, toward the center of the gall. Laterally, on twigs of several white oaks. 

RANGE. — From the Mexican border of California to Washington. 

[Kinsey goes on to describe several varieties with distinct host, range, and gall morphologies; see paper for details]

Andricus lasius

Hosts: Quercus chrysolepis

GALL. — A more or less globular, wool-covered, leaf gall. Polythalamous, with a half dozen or fewer cells. Roughly globular, flattened basally, 10. mm. in diameter, covered with short, dense wool about 2. mm. thick, the wool at first bright golden brown, weathering light gray or black. Internally hard, woody, solid except for the larval cells, which average 2.5 x 1.5 mm., and which are arranged rather radiantly about the mid-point of the base. Attached by a slight projection of the woody gall to the mid-vein or often to the petiole, occasionally two together, on either surface, usually on the under surfaces of leaves of Quercus chrysolepis. 

RANGE. — California: El Toro to Ukiah and Dunsmuir. I cannot fix characters which will distinguish the galls of the varieties, for they are very similar. Here is an instance of the physiologic measure of the insect varying less than does the morphologic. Morphologically the varieties are also very similar. What differences do exist are correlated with the occurrence of the faunal areas. 

[Kinsey goes on to describe several varieties that do not differ by host or gall morphology, only adult morphology and range; see paper for details]

Andricus operator

Hosts: Quercus marilandica, coccinea, ilicifolia, velutina, schneckii [shumardii]

GALL. — An oval larval cell, mostly inseparable from and occupying most of a thin-walled, seed-like capsule. These capsules woolly, clustered in the bisexual generation, naked in the agamic generation, producing the following forms of galls; on eastern American black oaks. 

Agamic Generation : The larval cell occupying most of a triangulate or obconical capsule, somewhat compressed, up to 5. by 4. mm. in dimensions, one to six, usually one, between the nut and the cup of the acorn; on all of the black oaks which bear the bisexual generation. 

RANGE. — Ottawa to Virginia, Texas, and Illinois. Probably thruout more eastern North America wherever black oaks occur. 

[Kinsey goes on to describe several varieties that differ by host, range, and gall morphology; see paper for details]

Andricus operator

Hosts: Quercus marilandica, coccinea, ilicifolia, velutina, schneckii [shumardii]

GALL. — An oval larval cell, mostly inseparable from and occupying most of a thin-walled, seed-like capsule. These capsules woolly, clustered in the bisexual generation, naked in the agamic generation, producing the following forms of galls; on eastern American black oaks. 

Bisexual Generation: Large, compact masses of wool, irregular in shape, often oval, up to 40. by 55. mm. in dimensions; the hairs at first crisp, succulent; greenish, white, or rose-tinged, sometimes deep red; becoming straw color or golden brown upon aging, finally weathering bluish gray or black, shrivelling considerably. Within the wool, scattered or in small clusters, mostly toward the center of the gall, are the seed-like capsules with the closely imbedded larval cells, each capsule hard, oval, about 1.5 by 3.0 mm., sometimes 150 or more cells in a gall. On or involving young stems, new clusters of leaves, and especially the flower clusters. 

RANGE. — Ottawa to Virginia, Texas, and Illinois. Probably thruout more eastern North America wherever black oaks occur. 

[Kinsey goes on to describe several varieties that differ by host, range, and gall morphology; see paper for details]

Andricus ovatus

Hosts: Quercus laevis, shumardii, texana, myrtifolia

GALL. — Elliptical, hollow, standing out from the bark of roots. Monothalamous, altho several galls may more or less solidly fuse. Each gall globoid to ovoid or egg-shaped, up to 4. mm. in diameter by 6. mm. long, usually smaller, covered with bark mostly of normal color and texture, somewhat smoother; a direct continuation of the root bark, altho the galls stand out distinctly from the bark; a large, circular exit hole terminally. Internally entirely hollow (in mature galls), the walls moderately thin, a distinct woody lining inside of the bark wall. On very large roots or on rootlets not 2. mm. in diameter; on black oaks. 

RANGE.— Florida to Texas. 

This is a very distinct root gall. The insects pupate late in the fall, soon becoming adults, but not emerging until early in the spring. The three varieties I have seen are distinct as to color, two of them remarkably so. It has been a mistake to completely ignore variation in color as tho occurring without order in any group of individuals. These color differences are correlated with very definite morphologic characters, and with host and geographic distribution. It is too much to expect one variety to become the next if transplanted from one oak to another or one locality to another. But host and geographic isolation have separated distinct forms in this species. The galls of all of the varieties are very similar. 

[Kinsey goes on to describe several varieties that differ by host, range, and gall morphology; see paper for details]

Andricus pomiformis

Hosts: Quercus agrifolia, wislizenii

GALL. — Spherical, smooth, compact, in section suggesting an apple with seeds about the core. Polythalamous, often with fifty or more larval cells. More or less perfectly spherical, rarely flattened, or elongate ovate; of all sizes up to 55. mm., averaging nearer 40. mm. in diameter; the surface sometimes entirely smooth, usually finely, irregularly pitted and cracked, sometimes set with low, raised lines or irregular rows of low tubercles radiating more or less from the slightly pointed tip of the gall ; brilliant, rich apple red when fresh, with some apple green and lighter green, becoming bright brownish yellow on aging, weathering darker. Internally like compacted sawdust, crystalline, spongy only in old and weathered specimens; with a hard, woody core originating from the point of attachment, the core flaring and somewhat branching in the center of the gall, with the larval cells arranged radiantly in the extremities of this core, mostly in a single layer centrally of the gall; each cell averaging 2. x 4. mm., with a distinct and easily separable, thin-walled lining. Laterally, less often terminally, attached by a small point, sessile on twigs of Quercus agrifolia and Q. Wislizenii. 

RANGE. — California: from the Mexican border to Dunsmuir and Ukiah. 

This is one of the most prominent of Californian galls, being surpassed in size and abundance only by Andricus californicus. Both species are known as oakapples. As with californicus, the possibility of more than one variety, each with its own intensely interesting biologic problems, has been heretofore entirely overlooked. The biologist who will learn what these two species should easily disclose in ten years or less work on the held, may contribute as important data as we yet possess on such problems as the nature of alternation of generations, distribution factors, and host effects, for instance.

Adults mature in early spring, earlier at southern localities. The field data indicate that northern varieties (see variety pomiformis) complete the life cycle in a year, with the possibility of an alternation of generations; while southern varieties (see maculipennis) take two years or more to mature, and may not have an alternate generation. This must be considered a suggestion, to be verified only by experimental work. If this is shown to be the true condition, it confirms my previously expressed opinion (1920, Bull. Amer. Mus. Nat. Hist., XLII, p. 372) that alternation of generations is an extreme development of seasonal dimorphism. 

If galls are collected too soon before maturity, the larvae will develop into adults altho few insects will emerge. One finds this the case with galls in the laboratory more often than with galls in the open. Under adverse conditions in the field, such as a season of drought, there may be a similar destruction of adults before emergence. An experience with this species in the laboratory alone, in connection with some other similar cases, misled me into an over-statement in a previous paper (1920, Bull. Amer. Mus. Nat. Hist., XLII, p. 383) as to the lack of vitality of Cynipidae in general. 

Some of the most illuminating data I yet have on the effectiveness of the host in isolating varieties is that presented under variety distinctus. 

Some of the varieties of this species show a median depression of the scutellum much as in the genus Amphiholips. Other varieties do not show this. An evolution of a generic character has occurred within this single species. 

Callirhytis rossi Kieffer (1903, Marcellia, II, p. 84, figs. 1, 2) is probably the same as one of the varieties described for this species in this paper. The publication was made without a more definite locality than "California", and with- out a host determination, but with a host description which best fits Quercus chrysolejns. I feel certain there must be a mistake here, just as I do not believe Kieffer's eriophorus came from Wislizenii as published. Until we can see types of rossi it will not be possible to determine the variety it represents. I am inclined to expect to find it a synonym of maculipennis. 

[Kinsey goes on to describe several varieties that differ by host, range, and gall morphology; see paper for details]

Andricus ribes, new species

GALL. — Spherical, berry-like, very succulent, green and more or less translucent when fresh, bright red when very young; the surface pebbled, with a few, small, projecting points. Shrivelling greatly upon drying, becoming blackened, 4.-6. mm. in diameter; when drying upon the twig, the galls become rather obconical in shape, remaining lighter in color. Monothalamous, the cavity filling the whole gall, the wall of tissue moderately thin, thinnest furthest from the point of attachment. Attached by only a slight point to twigs (bud galls) of Quercus douglasii.

RANGE.— California: Oroville, Three Rivers. Some of the galls at Oroville contained pupae on April 1, while an equal number showed adults or exit holes thru which adults had already emerged. The whole life of this generation must be a month or so, with an alternate, probably agamic generation in the rest of the year. Of 40 adults I have, 14 are males. The species is not entirely unlike Dryophanta pulchella Beutenmuller, of which the male and gall are unknown, but the two species are distinct enough. 

Andricus spectabilis Kinsey

Hosts: Quercus chrysolepis

GALL. — Elongate stem swelling. Polythalamous. Large, elongate, oval to spindle-shaped, averaging 25. mm. wide by 50. mm. long; large specimens scarcely greater in diameter will reach 110. mm. in length; covered with bark of natural color. Internally hard and woody, only the peripheral tissue being less compact than the normal stem wood; larval cells toward the center of the gall, oval, 3. by 5. mm., tissue almost not at all distinct from the rest of the wood; exit holes upon aging show a distinct, smoother area on the bark. On smaller stems of Quercus chrysolepis. 

RANGE. — California: San Jacinto Mountains to Auburn and Ukiah. Probably wherever Quercus chrysolepis occurs. 

[Kinsey goes on to describe several varieties that do not differ by host or gall morphology, only range and adult morphology; see paper for details]

Compsodryoxenus brunneus variety brunneus (Ashmead)

 GALL. — None, or a slight stem swelling; the larval cells embedded in the wood, quite as in the other variety. On Quercus chrysolepis. 

RANGE.— California: Mt. Diablo (Ashmead) ; Placerville. Probably confined to more northerly localities where Q. chrysolepis occurs. 

Compsodryoxenus brunneus Ashmead 

 GALL. — None, or only a very slight swelling of the stem. The larval cells are elongate, averaging 2.5 mm. long by hardly 1.0 mm. wide, embedded in the wood, with the lining hardly distinct, and not at all separable. Mostly on young twigs of white oaks. 

RANGE.— California. 

Compsodryoxenus brunneus variety atrior, new variety

 GALL. — None, or a slight swelling; the larval cells embedded in the wood; quite as in the other variety. On Quercus lobata. 

RANGE. — California: Byron. Probably occurs thruout the range of Quercus lobata (and not improbably Q. Douglasii) . 

Cynips mirabilis, new species

Holcaspis maculipennis
Cynips maculipennis
Amphibolips quercus-inanis (error)

GALL. — Globular leaf gall; monothalamous; 15.-35. mm. in diameter, thin-shelled, smooth, yellowish brown, rather closely marked with irregular, purplish brown spots. Internally with a hard-shelled larval cell, 3. X 5. mm., held centrally by a moderately dense mass of silky, sparingly branched fibers. On leaves of Quercus garryana. 

RANGE. — Washington: White Salmon. Oregon: Ashland, Junction City, Grants Pass, Roseburg, Canby; Portland (E. 0. Hovey) ; Albany (Trotter). California: Ukiah(?), Yreka; McConaughy, Siskiyou Co. (Fullaway) ; Elsies Creek, Amador Co. (G. Hansen, in Gray Herb.). Probably found thruout northern California, Oregon, Washing- ton, and British Columbia, wherever Q. garryana occurs. 

By a mistake in identification of this Hovey material, this species has gone thru the literature as Gillette's maculipenninis. The two are generically related, but are very distinct species. Maculipennis was described from New Mexico, and is restricted as far as I know to a southern Rocky Mountain area; mirabilis is confined to the Vancouveran zone of the Pacific Coast, at present recorded only from Quercus garryana, or some of its varieties, on which oak it is very abundant. Apparently adults emerge late in the fall or early in the spring, for by early March the galls are all vacated. This species belongs to Cynips Hartig; and not to Cynips of most American authors ; the species shows also some characteristics of Amphibolips, with which genus the physiology reflected by gall characters would connect the insect. 

Diastrophus kincaidii variety kincaidii (Gillette) 

[Kinsey distinguishes this species from "Diastrophus kincaidii variety austrior," which is now Diastrophus austrior, as follows] 

These two varieties are about as similar as any two forms which I should call distinct. Both insects and galls share the similarity. The best single distinction between the insects is the nature of the foveae of the scutellum ; other characters are distinct mostly in their averages. The two have distinct ranges. This species is not so very different from Diastrophus nebulosus and D. turgidus of the eastern parts of the United States, and I am not entirely certain that all of these should not be considered varieties of one species. 

GALL. [same description for both varieties] — Large, irregular stem swelling, Polythalamous, with a great many cells. Swelling rather abrupt, elongate, up to 60. mm. in length by 20. mm. in diameter; somewhat irregular, smoothed, covered with bark of normal color. Internally filled with mostly loose tissue in which the larval cells are densely crowded, the cells not separate, but almost separable, broadly oval, averaging 2. x 2.5 mm. On stems of Rubus parviflorus (acc. B. G. Thompson in Mus. Comp, Zool.), and R. nutkanus. 

RANGE.— Washington: Seattle (in Amer. Mus. Nat. Hist.) ; Olym- pia (Gillette); White Salmon. Oregon: Corvallis (B. G. Thompson in Mus. Comp. Zool.). Idaho: Cedar Mts. (Beutenmuller det.). 

[Kinsey distinguishes this species "Diastrophus kincaidii variety austrior new variety" from "Diastrophus kincaidii variety kincaidii" as follows] 
These two varieties are about as similar as any two forms which I should call distinct. Both insects and galls share the similarity. The best single distinction between the insects is the nature of the foveae of the scutellum ; other characters are distinct mostly in their averages. The two have distinct ranges. This species is not so very different from Diastrophus nebulosus and D. turgidus of the eastern parts of the United States, and I am not entirely certain that all of these should not be considered varieties of one species. 

GALL. [same description for both varieties] — Large, irregular stem swelling, Polythalamous, with a great many cells. Swelling rather abrupt, elongate, up to 60. mm. in length by 20. mm. in diameter; somewhat irregular, smoothed, covered with bark of normal color. Internally filled with mostly loose tissue in which the larval cells are densely crowded, the cells not separate, but almost separable, broadly oval, averaging 2. x 2.5 mm. On stems of Rubus parviflorus (acc. B. G. Thompson in Mus. Comp, Zool.), and R. nutkanus. 

RANGE. — California: Point Arena (Fullaway) ; Alameda Co. (thru Riley) ; Mt. Tamalpais. 

Diplolepis fusiformans

GALL. — Small, elongate stem swelling. Polythalamous, with rather few cells. Irregular, usually longer than wide, sometimes rather globose, more or less abrupt, usually but not always symmetrical about the stem, covered with brown bark somewhat different from the normal bark; up to 30. mm. long by 8. mm. in diameter, usually smaller, sometimes several more or less fused. Internally woody, rather solid, the normal stem tissue more evident centrally; the larval cells more or less elongate oval, the lining rather distinct but inseparable. On stems of Rosa species. 

RANGE.— Canada: Toronto (Cosens). Illinois (Weld). Colorado. California. Probably occurs thruout North America, wherever roses occur. 

This is one of the primitive species of the genus, as indicated by the closed radial cell, the small second abdominal segment, the equal numbers of the two sexes, and the slight development of the gall. It is not the same species as Diplolepis verna (Osten Sacken), altho both species are primitive. This species should be discovered wherever roses occur in North America, and many other varieties will need to be described. Of 236 insects I have bred, only 89, or 38 per cent, are males. But material of one variety gave over 50 per cent males, suggesting that the sexes are probably equal if the collections are adequate. The insects emerge in spring, from February further south to April (or later?) further north. It is unlikely that an alternation of generations occurs. 

[Kinsey goes on to describe several varieties of this species which vary in gall morphology and range; see paper for details]

Diplolepis radicum

GALL. — A large, massive, pithy root gall. Polythalamous, larger galls with a hundred or more larval cells. Very irregular in shape, but generally rounded, globular, depressed at the points of attachment, irregularly folded, gnarled, sometimes as an unopened bud; variable in size, up to 105. mm. in length by 70. mm. in diameter; reddish brown, darkening with age. Internally soft, rather pithy, with only a small amount of woody fiber; the larval cells large, nearly round, often 4. mm. in diameter, with a smooth but hardly distinct, inseparable lining. On stems and roots, under ground, or under debris close to the ground; sessile or nearly so, laterally or terminally, on roses. 

RANGE. — Canada and Maine to North Carolina and Washington, Probably everywhere in North America where roses occur. The insects emerge rather late in the spring, in April or May or later, depending on the development of the season. Immature galls may be found in July or August. The species is bisexual, with the sexes about equal in numbers, and probably takes only a single year to mature, without an alternation of generations. The number of parasites bred is not great, altho they do occur. W. M. Davis (1908) records finding galls broken into by mice. 

Diplolepis radicum variety radicum (Osten Sacken)

GALL. — Does not differ particularly from the galls of other varieties; averaging smaller. 

RANGE.— D.C.: Washington (Osten Sacken). North Carolina (Beutenmuller) . Pennsylvania? (Beutenmuller) . New Jersey: Ft. Lee (Beutenmuller) . New York: Staten Island (Beutenmuller); Nyack (Zabriskie in Amer. Mus.). 

Diplolepis radicum variety johnsoni, new variety 

GALL. — Does not differ particularly from the galls of other varieties, averaging smaller. 

RANGE. — Massachusetts : Gloucester. Not improbably occurs thruout more northern New England. 

Probably no two localities in the United States have been more thoroly collected for Cynipidae than eastern Massachusetts and the neighborhood of New York City. It is no credit to the taxonomy we have been doing to have ignored the evident differences between material from the two regions. No two adjacent varieties of this species are more distinct than radicum and johnsoni. I have not yet worked out the extent of the cynipid fauna of eastern Massachusetts; it is probably the same as the fauna of most of more northern New England; variety radicum is very probably confined to the remnants of the old Atlantic Coastal Plain. 

[Kinsey goes on to describe three more varieties, one of which is listed separately as [Diplolepis utahensis](https://www.gallformers.org/gall/1395)]

Diplolepis radicum variety utahensis (Bassett) 

Rhodites utahensis

GALL. — Does not differ particularly from the galls of other varieties; averages larger. 

RANGE.— Utah: Thistle (Weld collector); Provo, Price. 

Bassett described utahensis from several males cut from old galls sent by Mr. Siler ''from Utah". The original de- scription was a brief comparison of the males with "radicum'', but such a comparison as would cover several of the western varieties, especially since it cannot be entirely certain which of the eastern varieties Bassett had in mind. As with other Cynipidae, there are probably several varieties of this species to be found in Utah, and I have no other reason for taking my Provo and Price material to represent Bassett's name than that this is the common variety in a large part of the more accessible regions of that state. This material does not disagree with anything in the original description. Altho the types are lost it may be as well to retain the Bassett name, restricting it as I have. 

Of 4077 insects I have bred of this material, 2554, or about 63 per cent, are males. This high percentage is evidently due to the premature collecting of the galls which appears to pre- vent a larger number of the less mature females from emerging. Of the first 2948 Provo insects emerging 69.5 per cent were males; of the last 765 Provo insects emerging only 49 per cent were males. More mature galls collected at Price gave only 38 per cent males. These factors in obtaining ratios of the sexes of Cynipidae must be borne in mind if we are to avoid wrong conclusions. It is safe, however, to estimate that the males are about equal in number to the females in this species. The adults emerged after collecting on April 18 and April 20, 1920. As with other species in this genus this insect probably has no alternation of generations. 

A part of the material from Pullman, Washington, matches this Utah material very closely, but until I can examine further collections I should hesitate to state that the range extends into Washington. Beutenmuller's 1907 description of utahensis was made from this Washington material, which is at least in part another variety, diver gens. 

Diplolepis variabilis

GALL. — Irregularly ovate, smooth, solid, leaf gall. Polythalamous, each gall with usually three or four larval cells. Very irregular in shape, globular, to ovate, elliptical, or massive, all edges rounded; aver- aging 10. by 20. mm., observed to 25. mm. in diameter; smoothed, naked of bloom or scurf, light to rich brown (in mature galls). Internally solid, compact, soft, like compacted sawdust; the larval cells large, 2.0- 3.5 mm. in diameter, more round than usual, with a distinct but thin and inseparable lining. On roses, attached to leaves, or replacing leaf- lets, or replacing the whole leaf and attached directly to stems. 

RANGE. — From Texas and Wyoming to Washington. A very closely related species, Diplolepis ignota (O. S.), extends over the eastern half of the United States. I have not seen the insects of this species which have been recorded from Texas (according to Beutenmuller) , Wyoming, or Washington, altho I have seen the galls of the Washington material. Undoubtedly these regions as well as others in the western part of the United States have varieties distinct from any yet described. What data have been over-looked by the approximate taxonomy which fails to distinguish the several varieties concerned! Important information as to the age of the species and factors in the origin of species are mentioned in connection with variety lutescens. Diplolepis ignota (Osten Sacken) with its varieties occur- ring thruout the eastern part of the United States is so closely related to variabilis that there cannot be found differences in the insects which one could presume are of more than varietal rank. With a single change in regard to the sculpture of the mesopleuras, the general description of this species will apply to all varieties of ignota. On the other hand, the galls of ignota varieties are covered with a white scurf which the galls of all varieties of variabilis lack, and all of the white- galled varieties occur in the eastern half and all of the brown-galled varieties in the western half of the United States. This indicates closer affinities within each group than between the groups. One must choose between considering the groups as species and over-emphasizing their distinctness; and consi ering it a single species involved and attempting to express the groupings of the varieties by a system of quadrinomials, which is objectionable. If we will keep in mind the close relations of the two it may prove most convenient to use two specific names. Of 1622 insects which I have bred, 784, or over 48 per cent, are males. I have a relatively small amount of material of ignota, and I had previously believed that the high percentage of males obtained there was abnormal. But it would appear as if these species have a more nearly equal sex ratio than some others of the genus. As with other rose cynipids I have bred, the males appear to emerge earlier than the females, accounting for instance for the 57 per cent males collected at Provo, and the 43 per cent collected at Holly. The insects emerge early in the spring, in late April in the Rocky Mountain country, probably earlier or later in regions of earlier or later seasons. I have recounted the life history of ignota (Bull. Amer. Mus. Nat. Hist., XLII, p. 331), and that species in Massachusetts does not have an alternation of generations. The field data for variabilis do not disagree, and it is prob- able that an alternation does not occur for any variety of either group. 

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

Disholcaspis pattersoni, new species 

GALL. — Large, bluntly conical, bullet gall, in compacted clusters about the twig. Each gall primarily a true cone, but distorted by pres- sure of surrounding galls, a slight projection basally extending into the twig, and a slight overlapping of the primarily circular base about the twig; dark purple when fresh, becoming rich brown, darker at the tips; smooth; averaging 8. mm. in diameter by 12. mm. high. Internally solid, the larval cell oval, 2. x 4. mm., with a distinct, shell-like wall, not at all separable; cell located below the mid-point of the gall but not at the very base. On twigs of Quercus breviloba. 

RANGE. — Texas: Austin (Patterson); Round Rock, Leander. 

The gall of this cynipid has been known to me for several years, but previously I have not had the adult. I am glad to be able to name this for Dr. Patterson who successfully reared the insect. In 1921 Dr. Patterson found young galls on July 28, mature galls on August 29, pupae by October 21, mature adults in the galls by November 5, and emerging adults on December 12. I collected galls in abundance on December 6 and 8, 1919, at Round Rock and Leander, not far north of Austin, but all of the adults had emerged previously. Emergence dates must vary considerably with the development of the seasons in different years. 

This is probably a variety of Disholcaspis bassetti Gillette, but until I can see the types and more material of bassetti I cannot be certain of the relationships of pattersoni. 

Heteroecus bakeri (Kieffer)

Callirhytis bakeri 

GALL. — A naked, robust, ovate gall with a slender tip. Monothalamous. Body of gall spherical to more elongate-ovate, more abruptly flattened apically, 9.-13. mm. in diameter, more or less constricted basally; bearing a slender point apically, arising more or less abruptly from the body of the gall, averaging 1.5 mm. in diameter by 5. mm. long; entirely smooth or with a few tubercles, or with part of the surface roughly rugose, or the whole surface granulose; when fresh, bright to dark green, speckled red, or reddish or purplish brown with buff yellow or light green speckling; on aging becoming more uniformly yellowish brown. Internally similar to compacted sawdust, with a cylindrical central cavity extending from the base almost to the tip. Scattered, on twigs of Quercus chrysolepis. 

RANGE. — California: Yosemite (Trotter); El Portal, Pasadena, Upland, San Bernardino, San Jacinto Mountain. 

I have seen the Pomona College cotypes, and the Baker material in The American Museum of Natural History, and they agree with my Upland material. Indeed, my locality, Upland, and Baker's Claremont, are nearby towns neither of which have Quercus chrysolepis, but these are post office names for very probably the same region in the nearby mountains. 

Most of the adults were emerged in the San Bernardino mountains on January 31, 1920, and at Upland on February 3 ; most of the insects were not yet emerged at Pasadena on February 7, and adults were still alive in the galls at El Portal on March 21. As usual, emergence is later in more northerly localities. 

Both the galls and adults of this species show considerable variation. The above description applies to an average of the material from Upland, that is, near the type locality. The areolet varies from small to large, the foveae are smooth or sparingly rugose or completely, closely rugose; the parapsidal grooves extend further in some individuals than in others. The galls vary from smooth to very rough or tuberculate, the tip arises very abruptly or only gradually from the body of the gall, the greatest diameter comes at the middle or nearer the apex of the gall. All of these variations occur at each of the localities : San Bernardino mountains, San Jacinto mountains, Upland, Pasadena, El Portal. I cannot perceive any regularity in the concurrence of characters, and have not yet discovered any tendency for a single type to occur in the San Bernardino range, contrary to the usual situation. Galls from Placerville, and part of the material from El Portal, may belong to a northern variety of this species. Some adults from Placerville may belong to a northern variety of bakeri, but unfortunately these galls and adults were not definitely connected in the breeding, and I shall need to see more material before deciding the point. 

Heteroecus dasydactyli 

GALL. — Spindle shaped, covered with long wool. Monothalamous, occasionally bithalamous. Cylindrical, with a constricted base and an apex which is long in one variety, the tip often curved; smooth, some- times with a few, scattered, blunt tubercles; light green when young, turning light straw brown to darker; completely covered by a thick mass of long-threaded wool, light buff yellow tinged with pink when young, golden brown when older, attached only to the slightly projecting base of the gall; the wool deciduous from older galls. Internally almost solid; more porous tissue surrounding a cylindrical cavity extending from the base of the gall hardly more than half way to the tip, with the larval cell in the apical part of the cavity. Singly or in clusters, on twigs of Quercus chrysolepis. 

RANGE. — California: San Jacinto Mountains to Dunsmuir. 

Ashmead described the galls of this species as ''covered with long, brownish wool", which is correct. But the rest of his description of the gall, type galls which I have seen, and gall material labelled dasydactyli in several collections, match the galls from which I bred Hetercecus melanoderma. The extent of Ashmead's confusion is indicated by his choice of the name dasydactyli for this rather than for the true ''date- seed" gall, melanoderma. The gall and insect both are most nearly related to H. pacificus. One might take pacificus galls to be older galls of this variety with the wool worn off, but when the wool drops from dasydactyli it leaves the gall smooth (unlike melanoderma), and showing the projecting base to which the wool was attached (unlike pacificus) . 

Heteroecus dasydactyli variety dasydactyli (Ashmead) 

GALL. — Differs from the galls of other varieties in being more often long, elongate spindle-shaped, averaging 15.-30. mm. long by 7.- 12. mm. in diameter, often with a long, slender apex, less often curved than in pygmxus ; usually singly on the twigs. 

RANGE. — California: Yosemite Valley, Placerville, Dunsmuir. Probably occurs in the central Sierras north of El Portal, wherever Q. chrysolepis occurs. 

Ashmead's adults were bred from January 18 to February 11, probably indoors. I obtained adults at some date after collecting the galls out-of-doors : March 30 at Placerville, April 3 at Dunsmuir. Galls collected in the Yosemite Valley, at a high elevation, while snow still buried most of the small trees, were quite immature on March 26. These Yosemite Valley insects belong, without doubt, to this variety rather than to eriophorus, the variety of the southern Sierras. Eriophortis occurs at El Portal, not twelve miles from the Yosemite Valley, but at an elevation which is a thousand feet lower, and in a locality not nearly as exposed to the severe climate of the higher Sierras. The Yosemite Valley belongs to one faunal area. El Portal to another! 

Heteroecus dasydactyli variety eriophorus (Kieffer) 

Callirhytis eriophora 

GALL. — Very similar to that of variety pygmaeus. Each gall short, ovoid or less often spindle-shaped, with the tapering point short, not usually curved as in pygmaeus, more or less smooth. Usually singly, sometimes a few in a cluster, on twigs. 

RANGE.— California: El Portal; Clareniont (Baker); Upland, Pasadena, San Jacinto Mountains. Probably occurs thruout the southern Sierras and their extensions, south of El Portal, except in the San Bernardino and Cuyamaca mountains. 

This variety comes very close to variety pygmaeus, but the insects can be separated by the darker rufous brown general color, and by the rugose bottoms of the fovese of eriophorus. Pygmaeus comes from a neighboring but isolated mountain range. I have not seen types of this variety, but Dr. McCracken very kindly compared types of pygmaeus with Baker material of eriophorus at Stanford University, and independently concluded that the two are distinct. I have insects from Pasadena and the San Jacinto Mountains, but only galls from Upland, which is my locality for the same mountain range from which the Baker material came. All of my material is from Quercus chrysolepis. Kieffer records Q. Wislizenii as the host. I have never seen such a gall on Wislizenii ; the only material which I have seen labelled eriophorus bore leaves unmistakably those of chrysolepis; and I doubt very much whether Wislizenii is ever the host for this variety. 

Heteroecus dasydactyli variety pygmaeus, new variety 

GALL. — Very similar to that of eriophorus. Each gall short, ovoid, or more usually spindle-shaped, with the tapering point short and strongly curved; more or less smooth. Singly or up to ten in a compact cluster on the twigs. [no host is listed]

RANGE, — California: San Bernardino Mountains; Pasadena (?). Probably confined to the neighborhood of the San Bernardino Mountains. 

The gall of this variety closely resembles that of variety eriophorus, but the two can be separated very definitely. Dr. McCracken has compared one of my types with Baker material of eriophorus, and she points out distinct differences as noted under eriophorus. The San Bernardino Mountains constitute a ''mountain island" area with a very distinct geologic history and separated quite definitely from the neighboring San Gabriels, in which latter range Baker collected Kieffer's eriophorus. Six insects from Pasadena and their galls match the San Bernardino material closely, tho most of the Pasadena material belongs to eriophorus. 

Heteroecus sanctae-clarae

GALL. — A smooth, two-parted gall. The lower portion of the gall is rather cylindrical, narrower apically where it is inserted into the second part, concave basally where it is attached to the twig; the upper portion is low, cylindrical, as wide or more often nearer twice as wide as the lower portion, as high or little higher than the low^er portion, drawn more or less abruptly into a long, rather slender, blunt point; gall measuring up to 15. mm. in greatest diameter, and 20. mm. in total length; externally smooth or slightly roughened, covered with a bluish white bloom when fresh, becoming light buff to golden yellow, blackening on weathering. Internally compact, not woody, solid except for a small, irregular cavity in the upper portion ; the larval cell about central, divided between the two parts of the gall, or nearer the base. Sessil on twigs of Quercus chrysolepis. 

RANGE. — California: San Jacinto Mountains to Dunsmuir. 

This is not an acorn gall, as stated in the original description, but a bud gall. The insect shows extremely close relationships to other species of the genus. Indeed, the distinctive characters of the species are few. These characters are not as marked as for most varieties of species of the genus, and but for the character of the gall, sanctae-clarae would never have been described when it was. In regard to the gall, the character of the external surface, the compact internal tissues, and the fact that it is a monothalamous bud gall would indicate even a physiology related to that of the other species. It is in general form only that the galls are distinct. 

[Kinsey describes three varieties which do not vary in gall morphology; see paper for details]

Neuroterus evanescens, new species

 GALL. — Very slight swelling of the stem of the ament, causing an abortion of the whole ament into a short, ovoid mass about 5. mm. long, covered closely with clustered but otherwise normal anthers. The larval cells are tiny cavities in the swollen stem, without distinct linings, covered on the outside by only very thin tissue. On aments of Quercus breviloba. 

RANGE.— Texas: Austin (Patterson). 

Only two males were obtained among the fourteen adults, but it is not unlikely that the sexes are usually more nearly equal in number. Patterson reports the galls appearing early in March, and the adults beginning to emerge March 15. With such a short life history for this generation, another form must occur in the remainder of the year. A bisexual generation on such an evanescent part of the host usually produces the alternate, agamic generation on some other part of the same host, and Patterson has observed the females ovipositing on the under sides of the leaves. The gall material of the Patterson collection number 68 occurs on Q. stellata ; it shows only a slightly greater degree of swelling of the stems, but when the insects are obtained it may prove a distinct variety confined to the single species of host. 

Neuroterus floricola, new species

GALL. — A minute. egg--shaped capsule in the aments. Monothalamous. Each capsule ovate, somewhat elongate, more pointed apically; golden yellow brown, the surface finely pitted like leather; averaging 1.5 mm. long by 0.7 mm. wide; thin-walled, entirely hollow. Singly or two or three completely fused; showing clearly the origin from the anthers, sometimes incompletely furrowed, or with parts of anthers attached; on the aments of Quercus Douglasii. 

RANGE. — California: Three Rivers. Probably occurs thruout the range of Quercus Douglasii. 

A form on the flowers of the oak must have a short life history, with an alternate, probably agamic generation. 

This tiny species is of especial interest because the gall so closely resembles that of Andricus gigas Kinsey, gigas galls differing mainly in being larger, averaging 2.0 mm. in length by 1.0 mm. wide. Gigas was obtained from the same oak, in the same faunal area. I also have gigas from the same flowers which gave this Neuroterus. Of course the two insects are very distinct, belonging to different genera, even if the name Andricus does not define a genus. Some other species of Cynipidae produce similar galls on the aments. Here is one of the few cases I know among cynipids producing distinct galls where the species and genus of the gall maker is not the primary factor in determining the type of the gall. The anther appears to have only certain possibilities for abnormal development; practically identical results are effected when the stimulus is supplied by either Andricus gigas or Neuroterus floricola. Galls of much the same type, altho clustered differently and of several different shapes, are effected by Andricus dubiosus (Fullaway), Andricus serricornis Kinsey, Neuroterus pallidus Bassett, and some other species. In fact, I do not now recall any species where any other sort of modification of the anther is produced. The diversity of these insects is great; it would appear to be the nature of the anther which restricts the type of modification. 

Plagiotrichus asymmetricus, new species

GALL. — Stem swelling containing compact clusters of larval cells. Polythalamous. The swelling covered with bark normal in appearance; a gradual enlargement, often asymmetrical on the stem; up to 30. mm. long, by 20. mm. broad. Internally, the larval cells in compact masses, each cell with a distinct and sometimes separable lining, not very thin walled; the cells usually densely clustered. On young stems of Quercus chrysolepis. 

RANGE. — California: Pasadena and San Bernardino to Dunsmuir and Ukiah. Probably wherever Q. chryslepis occurs. 

The gall of this species does not differ materially in external appearance from the galls of two other Pacific Coast species found on the same host: Andricus spectabilis Kinsey and Plagiotrichus chrysolepidicola (Ashmead). The species presents the typical condition of distinct varieties wherever it occurs in a distinct faunal area. Between the adults of the first two varieties there are few differences, but their galls are very distinctive; between the galls of the second and third varieties there are not appreciable differences, but their adults are very distinct. In the first instance the species varies more physiologically than morphologically, in the second instance the morphology varies more than does the physiology. 

[Kinsey describes two morphologicallly identical varieties of this gall; see paper for details]

Plagiotrichus batatoides

GALL. — Quite similar to the galls of the other agamic forms of the genus; generally large, robust, up to 20. mm. in diameter and 28. mm. long; usually more irregular, composed of rather distinct, rounded parts, more or less fused; internally inclined to become partly hollow, the larval cells closely embedded, not very distinctly lined. On twigs of Quercus virginiana. 

RANGE. — Florida, Georgia, and possibly elsewhere. The typical variety confined to an area including Jacksonville, Florida. 

At this time I can add little to our knowledge of this species because I do not have insects enough to determine how many varieties it may have, and the range and hosts of each. The above descriptions, made from types, will separate this species from others of the genus, and give further details probably characteristic of only the typical variety. I have galls from a number of localities in Florida and Georgia, rep- resenting certainly more than one variety. Moreover some of the galls are from Quercus virginiana, some from Q. laurifolia, and some from Q. stellata. Probably each host has a distinct form for each faunal area. The Q. stellata galls from Georgia may represent a variety of Plagiotrichus elongatus, altho the galls are more like batatoides galls. Dozier states that galls collected early in December gave adults late in January; Weld bred adults on April 12, 1914. All of my galls were collected early in November, and I failed to secure adults probably because the larvae were still too young when the galls were dried. These galls are pecked into by birds and gnawed by mice, as Ashmead also recorded with the original description. Older galls persist on the trees until many of the live oaks are thickly covered with them. They are the favorite abode of ants which tunnel the soft tissue, further its early decomposition, and establish their colonies inside. Many of the old galls are almost hollow, covered by mostly complete bark, but internally a mass of thinly separated tunnels and decomposing chewings. The insect shows very distinct relationships to the other white oak species of the genus, rather than to the black oak species, but in many respects it is the most unique of the white oak species. The discovery of other varieties may link it more directly with some other species. Plagiotrichus elongatus rufopleurae, occurring in Texas on Quercus breviloba, is in some respects similar to batatoides, but more closely related to the other varieties of elongatus. 

Plagiotrichus chrysolepidicola

GALL. — Similar to those of most species of the genus. Swelling up to 20. mm. in diameter by 70. mm. long, usually much smaller; the larval cells distinct, more or less imbedded in the wood, usually not completely surrounded by wood, never separable. On white oaks. 

RANGE. — California and Oregon. Apparently confined to the Pacific Coast states. 

I have grouped the following varieties into one species because they are all so closely related that further subdivision is not logical (as discussed under diminuens) ; because they all occur on white oaks in the Pacific Coast region, and are abundantly different from the black oak varieties of the same region. The most closely related varieties I have yet seen are the white oak varieties from the Rocky Mountain region, here treated as species frequens. In no case are these varieties as generally dark in color, or the abdomen as naked as in frequens. But inasmuch as in most other respects varieties of both groups are so similar, it must be borne in mind that frequens and chrysolepidicola are closely related, if they should not be considered one species. 

The insects emerge very early in the spring, from February 7 in southern California to April 7 in Oregon. New galls were found in some instances at about the same dates, too soon to have been produced by the insects then emerging, suggesting the possibility of the species having an alternate generation which takes a full year for its growth. The galls are abundant wherever white oaks occur, each variety being confined as far as known to a single species of oak, and where that oak ranges over more than one faunal area there will be as many varieties of the insect. 

It is unfortunate that the first variety described was named chrysolepidicola, which wrongly defines the habits of most of the varieties. If P. congregatus (Ashmead) should prove to be the bisexual generation of any of these varieties, that name will take precedence for this species. 

[Kinsey describes nine varieties that differ in host and gall morphology; see paper for details]

Plagiotrichus coxii

GALL. — A globose to elongate, solid twig gall. Polythalamous, averaging twenty or more cells to a gall. Smooth, covered with nearly normal bark, somewhat reddened; up to 20. mm. in diameter and 65. mm. in length. Internally rather solid but not entirely so, the larval cells with a distinct lining-, closely embedded in the less solid part of the tissue. On black oaks. 

RANGE, — Arizona. Possibly also in New Mexico, western Texas, and Mexico, wherever Q. Emoryi, Q. hypoleuca, and related oaks occur. 

The insect of this species is in several respects, particu- larly the almost naked mesonotum and the 14-segmented, short antennse, extreme for this genus. But species like P. suttonii and P. perdens of California are intermediate between coxii and the white oak species of the genus. This empha- sizes the artificial nature of our sharp generic lines. I doubt whether I should consider the species in this genus except for the additional evidence furnished by the gall. It is typically a Plcigiotrichus gall, related more closely to the other black oak species, as is the insect also, than to any other white oak species. 

The insects emerge in midwinter, December and January. The young galls appear immediately, which suggests that they arise from an alternate generation, the whole life cycle taking more than one year. 

I have galls from New Mexico which may belong to this species, but rather extensive collecting in both western Texas and New Mexico failed to give me other galls of this species. If the species occurs in those states, it certainly does not there reach the abundance with which it infests the black oaks of southern Arizona. 

[Kinsey describes a distinct variety on each of this gall's two hosts; see paper for details]

Plagiotrichus elongatus, new species

 GALL. — Slight twig swelling, elongate in two varieties, globoid in one. Monothalamous or polythalamous, one to fifty or more cells in a cluster. The swelling slight, two or three times the normal stem diameter, up to 8. mm. in diameter, averaging 60. mm. long in two varieties, 8. mm. long in one; covered with normal bark. Internally solid except for an irregular, more or less continuous cavity completely filled by crowded, distorted, larval cells, each cell separate, with a distinct lining, but not separable. Near the tips of twigs of Quercus stellata and Q. breviloba. 

RANGE. — Texas: Austin, Probably thruout a part of Texas and Louisiana. 

[Kinsey describes three varieties of this gall; see paper for details]

Plagiotrichus frequens

GALL. — Similar to the galls of the other agamic forms of this genus. A stem swelling, up to 25. mm. in diameter and 110. mm. in length, usually smaller; internally rather woody, mostly solid, soft, spongy; the larval cells closely embedded, sometimes somewhat separable. On Quercus Gambelii, its varieties, or closely related oaks. 

RANGE. — Colorado, Utah, Arizona, and probably New Mexico. Probably thruout the southern Rocky Mountains and adjacent regions wherever Q. Gambelii varieties occur. 

The two known varieties occur in distinct faunal areas of the southern Rocky Mountains of Colorado. Inasmuch as that country involves so many other faunal areas, there are probably many other varieties of the species to be described. I have galls from Q. Gambelii varieties from several localities in Arizona and Utah that probably belong to this species. A, rugulosus Bassett and its varieties produce rather similar galls on evergreen white oaks of parts of the southwest, but those insects are generically distinct from frequens. As far as I know, frequens occurs only on the deciduous white oaks. 

The species is most closely related to P. chrysolepidicola of the Pacific Coast, and it is not impossible that we shall consider these one. species after other, intermediate varieties become known. P. elongatus of Texas is the next most closely related species, but frequens does not resemble elongatus nearly as closely as it does some of the varieties of chrysolepidicola. 

Gillette described this species as from Q. undulata. I have seen it only from the white oak with large, long, deciduous leaves; this would appear to be Q, Gambelii, or one of its varieties or closely related species, according to the more recent treatments of these difficult oaks. 

[Kinsey describes two varieties of this gall; see paper for details]

Plagiotrichus perdens

GALL. — Stem swelling, raggedly split open, containing flattened, seed-like cells. Polythalamous. Up to 13. mm. in diameter by 200. mm. long, averaging somewhat smaller. Within are large cavities, 4 or 5 arranged more or less radially about the axis of the stem, and a great many in series along the stem, each cavity sector shape, extending to the bark, 10. mm. or less in width at edge. Inserted on the wall of each cavity is a larval cell; each cell is monothalamous, flattened, a somewhat squared oval in outline, 4. mm. high by 3. mm. wide, broadest at the top; concave at the base with a projecting tongue by means of which it is inserted into the twig; cells smooth, shining, buff yellowish; finely streaked, more or less, with purplish brown. The cell walls are shell-like, moderately thick, entirely hollow within. At maturity the bark splits raggedly over each cavity, the larval cell drops to the ground, and the affected twigs die. On terminal twigs of Quercus Kelloggii, 

RANGE. — California and Oregon. Probably wherever Quercus Kelloggii and Q. Wislizenii occur. 

This species in several respects represents a considerable specialization for this genus. In no other species, as far as I know the genus, does the larval cell drop out of the gall be- fore the insect emerges. The species is directly related to Plagiotrichus suttonii, which occurs on Pacific Coast black oaks; the galls of the two are distinct. These galls are very common on Querctis Kelloggii and Q. Wislizenii, but I am not certain that they occur on the other black oak of the Pacific Coast, Q. agrifolia. A couple of galls I collected at Descanso, California, on February 23, 1920, are recorded from agrifolia, but I cannot be positive of this determination, especially since the two oaks intergrade regularly in their southern range. 

As I noted before, these galls kill the stems beyond the point of infestation, and thus kill the whole tree. Extensive stands of young black oaks are sometimes badly hurt. 

I do not know when the insect matures ; emergence occurs apparently after the larval cells drop to the ground. I found old galls, with no fresh galls evident, from January in southern California through April in Oregon. 

[Kinsey describes two varieties of this gall; see paper for details]

Plagiotrichus suttonii

GALL. — Rather large, solid stem swelling. Polythalamous, with a great many cells. The swelling elongate or globose to massive, distinct but only moderately abrupt, solid, covered with normal bark, a some- what distinct, smoother ring about the exit hole; internally very solid, entirely woody, the larval cell distinctly lined but closely embedded. On stems of all of the Pacific Coast black oaks. 

RANGE. — California, Oregon. Probably also Washington and British Columbia, wherever black oaks occur. 

The insects of this species emerge early in the spring, in February and March, earlier further south. The galls are hard and solid, and I have not found them eaten into by birds and mice as are the white oak species of Plagiotrichus. 

The three varieties described here are confined to black oaks, each to a single species, all of the Pacific Coast black oaks being affected. Probably several varieties occur on each oak as it occurs in different faunal areas, but this has been proved only for the agrifolia varieties. This species is not so radically different from Plagiotrichus chrysolepidicola, but shows characters typical of the other black oak species of the genus. Its intermediate nature may be due to the close relations, probably dating from more remote geologic ages, of the black live oaks and the white live oaks of the Pacific Coast. I have included coxii and asymmetricus in this genus, even tho they are not closely related to some other species, mainly because suttonii nicely connects these more extreme species. The species is very closely related to Plagiotrichus perdens which occurs on two of the same oaks over much the same territory. 

[Kinsey describes three varieties of this gall; see paper for details]

Studies of some new and described Cynipidae (Hymenoptera)