Cynips fulvicollis
GALL OF AGAMIC FORM. — Spherical, thin-shelled, gray, usually pubescent, the central larval cell supported by a dense mass of radiating fibers. Monothalamous ; up to 20.0 mm., in most varieties averaging nearer 8.0 mm. in diameter. Strictly spherical when fresh, but becoming a bit shrivelled and misshapen when dry; white, rose-tinged when fresh, becoming flesh-colored, ashen, or brown when mature; the surface shagreened, crystalline, slightly uneven, the raised spots bearing stellate hairs, the entire surface more or less pubescent, sometimes dense with a considerable pubescence when young, becoming more naked when mature. The outer wall moderately thin, papery or harder, not translucent; the young galls more or less solid, succulent, the older galls dry, well packed with fine, dense, radiating fibers which support the more or less spherical larval cell centrally. Attached by a fine point, singly, on the main veins, usually underneath the leaves, on all the more eastern American white oaks except those of the Q. virginiana group.
The galls of the agamic forms of fulvicollis are common on most of the species of white oak that occur over the northeastern two-thirds of the United States, often covering the ground under certain trees as closely as pebbles on a gravel drive. The more southern varieties of the species seem much more rare, being poorly represented in our collections altho I have done extensive field work in that portion of the country. Young galls of fulvicollis appear in June or July, earlier further south (June 6, 1927, in southeastern Kansas). Fully grown galls occur among the June collections from Kansas and in an August 25 collection from Roselle, New Jersey. The galls are deciduous, falling to the ground in September or October, by which time they contain some pupae. Adults are to be found in the galls in October of the year in which emergence occurs, but the insects do not chew out before the middle of November. Most of the emergence occurs before the end of December, but thruout the rest of the winter and early spring the insects continue to come out of the galls on bright days, especially if these have been immediately preceded by very low temperatures. The subapterous adults have been taken on several occasions running over the snow.
The most southwestern variety, vorisi, completes most of its emergence in the first winter after the development of the gall, thus holding to the life-history typical for the rest of the genus Cynips. On the other hand, some of the individuals of vorisi, and most of the insects of canadensis, fulvicollis, major, and gig as remain until a second or even a third winter in the galls before transforming into adults. Weld and Brodie first noted this, and my own experience confirms it. While only the northeastern varieties of fulvicollis were known, this two-year emergence seemed so exceptional as to suggest the exclusion of the species from the genus Cynips. The discovery of the shorter life cycle in our variety vorisi indicates how a physiologic quality that is ordinarily of generic rank may be modified by environmental factors. The small amount of first-year emergence which our records show for the more northern varieties may come from second-year galls that were indiscriminately included with first-year specimens in our collections.
On circumstantial evidence, we are now considering the long-winged pallipes the bisexual form of one of the short- winged agamic forms of fulvicollis. The data for this conclusion are detailed under pallipes. This bisexual insect is close to the bisexual forms known in the subgenus Acraspis, occurring in a seed-like bud gall very much resembling a bisexual Acraspis gall. The most distinctive thing in the bisexual Philonix is its unusual wing-body ratio of 1.17 which, however, is the length of the wing of the Southwestern Cynips (Philoriix) plumbea.
All of the previously recognized members of this species have been short-winged agamic forms. Few groups have had more varied taxonomic treatment, due to the difficutly of interpreting the relations of such specialized insects, and to an unusual amount of individual variation in the species. The genus Philonix was originally established by Fitch for his species fulvicollis and nigricollis, and Ashmead later designated fulvicollis as the type. Beutenmuller (1909) and Dalla Torre and Kieffer (1910) revised the generic assignments of these insects, but differed radically in their interpretations — chiefly because they worked with little material and studied few types. They nevertheless agreed in keeping these forms in genera which included none but nearly wingless forms. Weld maintained this policy when he revised the genera Acraspis and Philonix in 1922, but rendered a service by pointing out the two distinct lines of relationships represented by these two names. Those who have followed my use of the categories of species and variety will readily appreciate that what have hitherto been considered species of Philonix should now be considered varieties of one species.
Beutenmuller introduced confusion into our understanding of this group by trying to interpret it from published descriptions without an examination of types. His later study of the types led to an unfortunate synonomy that was, nevertheless, uncritically accepted by several of us for a number of years. He considered his erinacei a synonym of fulvicollis. I agree with Weld that the two represent distinct genera or subgenera. Beutenmuller was correct in considering (1909) gillettei a synonym of niger, but incorrect (1918) in considering this the same as lanaeglobuli. I believe that niger, gilletti, and Fitch’s nigricollis are synonyms of fulvicollis. My reasons for these interpretations, which are in every case based upon my studies of the type material supplemented by series representing a wide range of localities, are detailed in the following pages.
As I understand this species, we have the familiar story of the host or geographic isolation of each variety of the insect. The white oak inhabitants include variety canadensis which occurs chiefly in the sub-Canadian area of the northeastern part of the country, the highly variable fulvicollis (of which nigricollis, niger, and gillettei are synonyms) ranging from Massachusetts to Iowa, south to the Ohio River and still further in the mountains, variety major of the Ozark area extending from Indiana to Missouri and Arkansas, and variety rubricosa of the Gulf area from Texas eastward and northward to Tennesssee. Quercus macrocarpa and the closely related Q. bicolor harbor vorisi in the Ozark area from Kansas to Illinois and apparently without material variation all the way into northern Indiana. Quercus lyrata and the chestnut oaks bear variety gigas in the Ozark area. Quercus bicolor in Florida has variety lanaeglobuli. Whether there is a distinct Coastal Plain variety has not been determinable from the material I have seen. There is some irregular extension of the hosts of these varieties in areas in which the normal hosts are rare or lacking. Beutenmuller (1909, Bull. Amer. Mus. Nat. Hist. 26 : 252) notes a Q. prinoides gall from New Jersey, and I have material (C. J. Long, Jr., coll.) from the same host and the same state; and this may (or may not) represent another variety.
The geographic isolation of varieties in fulvicollis is, unfortunately for our determinations, not as sharp as with other species of Cynips. Indeed, I have never worked with any species of the Cynipidae in which there is more even grada- tion from one extreme (canadensis) to the other (vorisi). Such intergradation does occur commonly among some other organisms, both plant and animal, and it is rather remarkable that it does not occur more often among the Cynipidae. There will, no doubt, be some who would object that such an even series as fulvicollis should not be broken up into varieties; but the extremes of the present series are as different in feature and build and physiologic reactions as a Spaniard is from a Swede. There is more than a political purpose served in applying distinct names to those two races or varieties of Europeans, even tho intermediate areas of Europe may exhibit gradations from one to the other. Similarly, it would confuse the biologic data to apply one name to this whole complex of fulvicollis. The variety canadensis, for instance, is a distinctly small insect with a largely naked mesonotum, a very small gall, a two-year life cycle, and a restriction to white oak in sub-Canadian areas. The variety vorisi is a large and robust insect, always with a very hairy mesonotum, a gall which averages twice the diameter (which means eight times the volume) of nigricollis galls, a one-year life cycle, and a restriction to Q. macrocarpa and Q. bicolor in Ozark areas. In the territory between northern Michigan and Kansas, one may find every gradation and a continuous variation from one to the other. Wherever the two insects happen to come in contact, they certainly interbreed and produce individuals not fit for the cabinet of any systematist who believes that all specimens represent one species or another. Among the more than five thousand insects which we have of this group, it has proved possible to make varietal determinations for every series, altho the paucity of available taxonomic characters in these aborted, short-winged, forms and the amount of hybridization makes it impossible to guarantee our determinations of each individual as we can for most other Cynips.
The galls of fulvicollis sometimes contain a high percentage of parasites and inquilines. Most of the parasites emerge in the first spring after the development of the gall, the emergence occurring from February to July. The inquilines emerge for the most part in the later part of that same spring, even until late in June. The inquilines appear to act as para- sites that have attacked the gall at an early stage of development, for no remains of the gall maker are to be found in inquiline-inhabited specimens.
Cosens (1912:343) gives a description of the histologic structure of a gall of this species. His account is as follows:
Outside the nutritive zone is a wide crystal layer, each cell of which is completely filled with a crystal mass. The sclerenchyma of the protective zone is formed in a very unusual manner. The sides of contiguous cells are thickened in such a way that there is an almost spherical deposit at the points where the cells are in contact.
Radiating out from the protective layer are long narrow cells which form the minor part of the parenchyma zone. The remainder of this zone consists of irregularly elliptical, thin-walled cells. The epidermis is covered with a dense growth of trichomes with thick laminated and sclerified walls.
Cosens failed to note a thin collenchyma layer which is to be found directly beneath the epidermis.
Cynips fulvicollis var rubricosa, new variety
agamic form
Philonix nigra err. det Weld
GALL. — As described for the species; quite pubescent; up to 9.0 mm. in diameter; on leaves of Quercus alba and Q. stellata.
RANGE. — Oklahoma: Tuskahoma (galls, acc. Weld 1926). Texas: Marshall ( Q . stellata, W. A. Lansford in Kinsey coll.). Alabama: Athens (gall, Kinsey coll.). Tennessee: Charleston (types, Kinsey coll.). Possibly extending thruout a widespread area in the southeastern United States, from eastern Texas to northern Oklahoma, Tennessee, and western Florida. Figure 38.
This variety and Ashmead’s lanaeglobuli are the most southern forms described for this species. Cynips fulvicollis seems not at all common in the more southeastern portion of the United States; and we have little material of the present variety altho we have engaged in held work over many thousands of miles in what would appear to be the range of this insect. The host of three of our collections is Q. alba, but a Texas insect from Q. stellata seems identical with the holotype of rubricosa.
The insects of rubricosa are closest to variety gigas, but rubricosa is much smaller with a much shorter wing and a bright rufous to rufo-piceous patch at the base of the abdomen.
Our Texas material of rubricosa emerged on January 7, during the hrst winter after collecting the galls. All of the insects of the type series, from Tennessee, delayed emergence until the second winter, coming out on December 19 and January 4.
Cynips fulvicollis var vorisi, new variety
agamic form
GALL. — As described for the species, very large, pubescent, up to 17.0 mm. in diameter; on the leaves of Quercus macrocarpa and Q. bicolor. (Rarely on Q. Michauxii ?) . Figure 230.
RANGE. — Kansas: Winfield ( Q . macrocarpa, types; R. Voris in Kinsey coll.). 10 miles southeast of Winfield and Cedarvale ( Q . mac- rocarpa R. Voris in Kinsey coll.). Dexter ( Q . Michauxii, R. Voris in Kinsey coll.). Illinois: Olney and West Union (Q. bicolor, Kinsey coll.). Indiana: Steubenville ( Q . bicolor, Kinsey coll.; determination open to question). Largely restricted to the Ozark area from Arkansas to Indiana, but occurring westward into Kansas and northward to northern Indiana ; chiefly confined to Q. macrocarpa and Q. bicolor. Figure 38.
This is the Ozark variety of fulvicollis on the burr oak, Q. macrocarpa, and its close relative, the swamp white oak, Q. bicolor; but the species extends well outside the Ozark area both to the west in Kansas and to the northeast into Indiana. Even the Q. bicolor material which I have from northern Indiana seems no different from dark specimens of vorisi, tho the northern specimens do thus average differently in respect to color. Vorisi is very close to gig as, but has less bright rufous on its face and mesonotum. From major , vorisi differs in being a larger and much more robust insect. Altho vorisi, gig as, and major occur in much the same territory, they are on distinct oaks {gig as on Q. lyrata and the chestnut oaks, major on Q. alba), but the host isolation appears to break down in some cases, for perhaps ten per cent of our material of these insects is distinctly intermediate between the several varieties.
Very young galls of vorisi were found in southeastern Kansas as early as June 8 (1927). Full-sized galls were obtained by June 15, and galls large enough to breed were found on July 4. This represents earlier development than we know for the northeastern varieties of the species. Most of the adult insects emerge during the first winter, but about ten per cent of our material has emerged in the second winter after collecting. The emergence during the first winter is practically continuous from early December to the end of February, the bulk of it occurring well before the end of December; the emergence in the second winter was confined to December. The recorded emergence dates are December 2, 8, 9, 10, 12, 15, 17, 18, 19, 20, 22, 23, 26, and 28; January 2, 4, 5, 10, and 25; and February 2, 6, and 20.
Almost all of our Kansas material of this insect was collected by Dr. Ralph Voris, of the Southwestern Missouri State Teachers College. For several years Dr. Voris has been a companion in my field work, and many thousands of specimens of Cynipidae and many hours of help on field routine are to be credited, to him. He has an appreciation of the importance of individual variation, and to this fact we are indebted for the unusually large series which we have of several of the Cynipidae from Kansas and western Missouri.
Cynips fulvicollis var major, new variety
agamic form
Philonix nigra err det Weld
Cynips fulvicollis var B Kinsey
GALL. — As described for the species, rather large, pubescent, up to 10.0 mm. in diameter; on the leaves of Quercus alba; rarely on Q. Muhlenbergii, Q. Michauxii, and Q. macrocarpa. Figure 227.
RANGE. — Indiana: Steubenville, Nashville, Bloomington, Clinton, and Linton ( Q . alba, Kinsey coll.). Spencer (Q. alba and Q. Michauxii, Kinsey coll.). Letts ( Q . Michauxii, E. B. Ruth in Kinsey coll.). Charles- town (H. Spieth in Kinsey coll.). Illinois: Urbana (A. E. Miller in Kinsey coll.). America (types, Kinsey coll.). Bonnie, Bloomfield in Johnson County, West Union, and Norris City (Kinsey coll.). Eddyville (O. Buchanan in Kinsey coll.). Kentucky: Pinehill, Dawson Springs, Wickliffe, and Paducah (Kin- sey coll.). Cleveland (hybrid, Q. Michauxii, Kinsey coll.). Missouri: Rankin ( Q . Muhlenbergii and Q. alba, Kinsey coll.). Arcadia and Poplar Bluff (Q. alba, Kinsey coll.). Arkansas: Winslow (R. W. Shreve in Kinsey coll.). Kansas: Winfield ( Q . macrocarpa, R. Voris in Kinsey coll.). Apparently centering in the Ozarks, extending somewhat beyond this area, from southern Indiana to southern Missouri and (scatteringly) in Kansas. Figure 39.
As one travels southward and westward from Indiana and Illinois, he finds variety fulvicollis gradually giving way to a larger and somewhat darker insect which has a more hairy mesonotum. In southern Indiana as far north as Blooming- ton, and in Illinois as far north as Urbana, one may find many insects which are distinctly intermediate between fulvicollis and major and not always (altho sometimes) distinctly one or the other variety. In southern Missouri and Arkansas major occurs in more nearly pure? form. Further west, in the eastern part of Kansas, major seems to hybridize with vorisi, altho vorisi is on Q. macrocarpa and major seems confined to Q. alba except where alba is rare or lacking. We have a few specimens of what would appear to be hybrids of major x gigas on Quercus Michauxii and Q. Muhlenbergii in Missouri and Kansas.
I have numerous insects of major which emerged during the first season, but most of the insects I have bred waited until the second winter for emergence. The recorded emergence dates are November 22; December 1, 4, 8, 9, 10, 12, 14, 15, 16, 17, 18, 19, 20, 22, 28, 24, 26, and 28; January 2, 3, 4, 5, 7, 8, and 20; February 2, 6, and 20, and even one case of emergence as late as April 25.
[Kinsey's descriptions of varieties gigas and lanaeglobuli are listed as distinct species for the time being]
Cynips fulvicollis var fulvicollis
Philonix fulvicollis
Philonix nigricollis
Cynips fulvicollis
Teras fulvicollis
Teras nigricollis
Cynips nigricollis
Biorhiza fulvicollis
Biorhiza nigricollis
Acraspis niger
Biorrhiza nigricollis
Acraspis nigra
Biorrhiza fulvicollis
Acraspis gillettei
Philonix gillettei
Philonix nigra (in large part)
Philonix niger (in part)
Cynips fulvicollis var A and C Kinsey
NOT Diplolepis niger
NOT Biorhiza niger
NOT Andricus fulvicollis forms bicolens and erinacei
GALL. — Up to 10.0 mm. in diameter; naked to pubescent; on leaves, usually of Quercus alba, occasionally on Q. Michauxii (and Q. macrocarpa) . Figure 228.
RANGE. — Ontario: Toronto (Brodie in U.S. Nat. Mus.; also acc. Cosens 1912). New Hampshire: Alton Bay (galls, Mrs. D. Tenney coll.). Massachusetts: Forest Hills and Framingham (galls, Kinsey coll.). Connecticut: probably Waterbury (Bassett coll.). New York: near Albany? (Fitch, types of fulvicollis and nigricollis) . New York City (Beutenmiiller acc. Weld 1926). Riverhead (Crosby acc. Weld 1926). Ithaca (acc. Weld 1926). New Jersey: Ft. Lee (Beutenmiiller in American Mus. Nat. Hist.). Bennett (Kinsey coll.) . D. C.: Rock Creek (F. E. Mather in U.S. Nat. Mus.). Washington (H. S. Barber in U.S. Nat. Mus.). Ohio: Columbus (galls, Kinsey coll.). Michigan: Ionia Co.? (Gillette, types of gillettei and of niger?). Tekonsha, Owosso, Bay City and Traverse City (Kinsey coll.). Three Rivers ( Q . alba and Q. macrocarpa, Kinsey coll.). Martin (gall, Kinsey coll.). Indiana: Porter (acc. Weld 1926). North Webster, Winona Lake, Delphi, Mitchell, and Ontario (galls, Kinsey coll.). Crawfordsville (E. C. Stout and Hugh Lee in Kinsey coll.). Steubenville, Huntington, Bloomington, 10 miles southeast of Bloomington, 7 miles east of Bloom- ington, Linton, Clinton, and Bedford (Kinsey coll.). Ft. Wayne (C. M. Kinsey coll.). Nashville and Spencer (Q. alba and Q. Michauxii, Kinsey coll.) . Letts (Q. Michauxii, E. B. Ruth in Kinsey coll.) . Charlestown (H. Spieth in Kinsey coll.). Illinois: Evanston, Glencoe, Glen Ellyn, and Fountaindale (acc. Weld 1926). Oakwood, Oakley, and Urbana (A. E. Miller in Kinsey coll.). Danville (galls, W. V. Balduf in Kinsey coll.). Eddyville (O. Buchanan in Kinsey coll.) . West Union, Norris City, Olney, Bonnie, Bloomfield in Johnson County, and America (Kinsey coll.). Pana (galls, Kinsey coll.) . Kentucky : Paducah, Wickliffe, Dawson Springs, and Pinehill (Kin- sey coll.) . Cleveland (gall, Kinsey coll.) . Virginia: Bluemont (acc. Weld 1926) . Blue Ridge Mts. near Natural Bridge Station, and Winchester (Kinsey coll.) . Iowa: Corinth (C. Barracks coll. acc. Weld 1926). Ames ? (Gillette coll., types of gillettei? Also material of niger in U.S. Nat. Mus.) . Nebraska: Nebraska City (acc. Weld 1926) . Missouri : Kimmswick (acc. Weld 1926) . St. Louis and Allenton (galls, E. S. Anderson in Kinsey coll.). Rankin (Kinsey coll.) . Kansas : Dexter ( Q . Michauxii, R. Voris in Kinsey coll.) . Leaven- worth (galls in Kinsey coll.) . Apparently thruout the northeastern United States and adjacent Canada, westward to Nebraska, southward to Kentucky and Kansas. The variety appears to be a hybrid of major x canadensis, and since there are many forms of intergrades between these and fulvicollis, the extreme locality records are open to question. Figure 41.
This is the common variety on the white oaks of the northeastern quarter of the United States and adjacent Canada. It is possible that fulvicollis is replaced on the Atlantic Coastal Plain by a distinct variety. The southern Indiana material that is available shows gradation toward the variety major of the more southern Middle West. True fulvicollis is rare south of the Ohio River, except in the eastern mountains of Kentucky and still further south in the Blue Ridge.
The host of fulvicollis is usually Q . alba. Nevertheless I have four insects that appear to be fulvicollis from Q. Michauxii from Letts and more insects from the same host from Nashville in southern Indiana and from Q. macrocarpa in southern Michigan. Gillette recorded both Q. alba and Q. macrocarpa as hosts in “Michigan and Iowa.” Ashmead’s record (1885:304) for this insect on the chestnut oak, Q. montana , is open to re-determination.
The galls of fulvicollis begin developing in mid-summer (July 4 at Bloomington, Indiana, in 1929), but galls at Winona Lake in northern Indiana on August 12 (1927), and further south at Spencer, Indiana, on September 11 (in 1926) were still small, succulent, and almost solid, without differentiation of the fibers which later support the larval cell. Galls from central Illinois (Urbana) were full size but succulent and with very small larvae on August 18 (1927). The gall makers begin to emerge at the beginning of either the first or the second winter, but mostly in the second season. Fitch found his type material running- about on the snow. He thought the insects had come from root galls, but in this he was, of course, mistaken. Other breeding records range from early November to the latter half of February. Brodie secured adults at Toronto on November 10, 18, 20, and 24 (1886 to 1899, acc. U.S. Nat. Mus. coll.) and December 5 (in 1908 and 1907). From material collected in October at Glencoe, Illinois, Weld secured two adults before November 23 and five before December 11 (in 1916). Weld also gives November 1 to 19 (in 1917), December 2 (in 1919), and February 19 as dates on which adults were secured. F. E. Mather secured active adults at Washington on December 25 (in 1908). My own breeding records, scattered over several years, are for November 16 and 22; December 1, 4, 8, 9, 10, 12, 13, 14, 15, 16, 17, 18, 19, 20, 22, 23, 26, 28, and 30; January 1, 2, 3, 4, 5, 6, and 8; and February 2. Very little of this emergence occurs during the first year, practically all of it occurring during the second winter after the development of the gall.
Some colonies of galls are very heavily parasitized; other colonies yield a remarkably high percentage of gall makers. Parasites commonly emerge from the galls from late August until November of the first season, and again in abundance during the following April, May, and June.
The eggs of the agamic form of fulvicollis are apparently layed in the unopened buds of the white oaks, for Bassett’s pallipes , which produces a seed-like gall that develops in these buds in the next spring, is apparently the alternating, bisexual generation of fulvicollis. Fitch first drew attention to a perceptible ant or bee odor which is given out by the agamic gall maker of this species. Cosens (1912) shows a section of a larva of this variety that indicates there is an external opening to the digestive tract posteriorly. His account of the histologic structure of this gall is quoted under the specific description of fulvicollis in the present publication.
Altho fulvicollis is one of the oldest names among American Cynipidae, Fitch’s meager description went uninterpreted for nearly sixty years. Beutenmuller then found the types at the U. S. National Museum, but unfortunately considered them the “same as” his erinacei. I uncritically followed this synonomy in my paper (1920) on the life histories of American Cynipidae. A few years ago Weld studied these types and concluded that fulvicollis is a Philonix while erinacei is an Acraspis. My recent studies in the U. S, National Museum lead me to agree with Weld’s interpretation. At the same time, direct comparison of the holotype and paratypes of fulvicollis and nigricollis , of paratypes of gillettei, and of the holotype of Gillette’s niger leads me to conclude that all these names are synonyms. Each holotype is different from any of the others, but every variation covered by the types, and many additional variations are included in every extensive series we have represented in our collections.
What we appear to have here is a highly variable complex that grades into variety canadensis to the north and variety major to the south. In the intermediate territory, which is much of the northeastern quarter of the United States, one finds canadensis and true major and every conceivable intergrade between the two. The intermediate types are everywhere the most abundant. As one goes north, true canadensis or hybrid individuals appearing to have a great deal of canadensis blood are predominant. Toward the south, major similarly asserts itself.
One examining several thousand insects of this remarkable complex cannot doubt its hybrid origin from canadensis x major. The chief question to be raised is whether we are warranted in calling such a heterogeneous population a species. It is much like Cynips pezomachoides erinacei which also seems hybrid in origin, and the same geologic history that explains (p. 398) the origin of erinacei should fit fulvicollis . The two occur in the same area. In both cases the northern variety probably met its opportunity to hybridize with the southern variety upon the advance of the Pleistocene glaciers, and the retreat of the glaciers offered an extensive territory for the isolation and development of the hybrid. Fulvicollis is not as thoroly fused as erinacei. Nevertheless, the wide- spread distribution of the complex may warrant its recognition as a distinct taxonomic unit, whatever we may choose to call it.
The individuals which are the types of nigricollis , niger, and gillettei are only larger or smaller, lighter or darker, more hairy or less hairy representatives of this fulvicollis complex, and not nearly as diverse forms as a larger series of true fulvicollis will show.
Cresson’s list of the Bassett types in the Philadelphia Academy states that gillettei came from Fort Collins, Colorado. This probably represents an attempt to interpret Bassett’s original record which read : “from the ground beneath a large white oak on his [C. P. Gillette’s] lawn.” Professor Gillette has recently written me that he is not now positive of the locality from which he obtained the type material of gillettei , but he believes the galls “were taken either at my old home in Ionia County, Michigan, where there was a white oak tree of considerable size in our dooryard, or on the campus of the Iowa Agricultural College. . . . I do not recall any white oak near the house in which we lived on the Ames campus.” If the present species is ever found as far west as Colorado, it should be represented by a variety distinct from any now known from the more eastern areas.
Cynips fulvicollis var canadensis, new variety
agamic form
GALL. — As described for the species; often smooth, shining, and naked, with very little pubescence; the typical galls averaging about 6.0 mm. in diameter; on leaves of Quercus alba. Rarely on Q. Michauxii, Q. Muhlenbergii and Q. macrocarpa in the more southern extension of the range.
RANGE. — Michigan: Traverse City (types, Kinsey coll.). Bay City, West Branch (galls, Kinsey coll.). Owosso (Kinsey coll.). Pent- water (gall, F. Payne in Kinsey coll.). Interlochen (galls, R. Voris in Kinsey coll.). Indiana: Fort Wayne, Huntington, Nashville, and Clinton (Kinsey coll.). Morocco ( Q . macrocarpa, Kinsey coll.). Spencer ( Q . Michauxii, Kinsey coll.). Illinois: Olney, Bonnie, Norris City, Bloomfield in Johnson County, and America (Kinsey coll.) . West Union (Q. alba and Q. bicolor, Kin- sey coll.). Kentucky: Paducah and Dawson Springs (Kinsey coll.). Kansas: Winfield (Q. Miihlenbergii, Q. macrocarpa, Voris in Kinsey coll.). Cedarvale (Q. Muhlenbergii, Voris in Kinsey coll.). Probably thruout the sub-Canadian area of the northeastern quarter of the United States, still surviving in fairly pure form and as hybrids with varieties fulvicollis or major as far south as southern Illinois and the Kentucky mountains. Beyond the area covered by our collections, it is to be expected in northern New York, New England, and southeastern Canada. Figure 42.
This is the most northern variety of the species, best repre- sented at present from the northern end of the southern penin- sula of Michigan, but to be expected everywhere where white oak occurs along the Canadian-United States boundary. There seems every reason to believe that this insect was pushed southward with the advance of the Pleistocene glaciation, and that variety fulvicollis ( q.v.) is a hybrid of canadensis x major. If this interpretation is correct, it explains why fairly typical canadensis still occurs in small numbers as far south as southern Illinois, southern Indiana, and the hill country of Kentucky. We even have material from southeastern Kansas, but this determination may be open to question. The influence of canadensis in the fulvicollis complex increases considerably in northern Indiana and southern Michigan, but no considera- ble percentage of pure individuals of canadensis is to be found until one reaches more northern Michigan. The insect that Fitch described (1859) as nigricollis would appear to be an individual of fulvicollis carrying considerable third winter after the galls started growth. Only stray indi- viduals emerged the first year after collecting our more northern material, and it is not certain that these did not come from old galls gathered with the new crop. On the other hand, all of our Kansas insects emerged the first year after collecting. If our determinations are correct, these data raise an interesting question as to the factors affecting emergence. Direct temperature affecting the developing larvae cannot be responsible for the date of emergence of the insects, for Michigan material kept in the mild climate of southern Indiana for thirteen months emerged in the middle of November, and Kansas material placed in the same breeding box outside our laboratory windows emerged from late December to late January — in both cases near the dates they would have emerged in the normal environments. Our emergence dates for this insect are November 16; December 1, 4, 8, 10, 16, 18, 20, 22, 28, and 30 ; and January 4, 5, 8, and 25. The emergence of the material from northern Indiana and Michigan ranged from November 16 to December 20; that of the more southern collections began on December 4, but was concentrated chiefly between December 12 and January 25.
