Cynips (fulvicollis) insulensis, new species
agamic form

GALL.--Typical of the fulvicollis complex; pubescent, up to 9 mm. or more in diameter. Figure 121.

HOSTS.–In the Rockies: Q. Gambelii. In the Black Hills: A member of the Q. macrocarpa complex. East of the Rockies: Q. macrocarpa.

RANGE.-Illinois: Stockton, 3 W (gall). Wisconsin: Portage, 14 W. Portage, 7 W (gall). Stevens Point, 5 S (gall). Stevens Point, 14 S (gall). Waupun, 8 N. Lake Geneva. Sparta, 7 E (gall). Wausau, 5 S (gall). Minnesota: Badger (galls, R. Voris in Kinsey coll.). Harmony, 6 N. Iowa: Postville, 3 E (galls). Nebraska: Seward (galls). Nebraska City (galls). South Dakota: Black Hills State Park. Deadwood (types). Wyoming: Sundance (galls). Colorado: Happy Canyon (20 SE Denver). Manitou (galls). Palmer Lake (galls). Occurring in the Rocky Mountains east of the Divide, and throughout the northern Middle West north of Missouri and west of Lake Michigan. Figure 120.

LIFE HISTORY.—Galls: immature mid-August to September 1. Adults: November 4, 13, 20, 25. December 1, 10, 24, 26, 28. Emergence in part in the first year, in part in the second year after formation of the gall.

The present species closely resembles the eastern canadensis and the Utah species latigenae; but insulensis has parapsidal grooves which are more sharply defined than in any other species of the fulvicollis complex.

It is interesting to find the range of this species discontinuous. It extends into the Rocky Mountains east of the Divide in Colorado (and Northern New Mexico?); into the essentially Rocky Mountain island which is the Black Hills; and into the eastern deciduous forests extending from the eastern part of the Dakotas through Minnesota and Wisconsin to Lake Michigan. While these three areas are well separated by arid plains, there is no good barrier to limit the range toward the south and east, and through Iowa and Northern Missouri we may expect to find the species hybridizing with C. vorisi. We have numer ous insects of the species, but select the types from our limited material from Deadwood, on the north side of the Black Hills. There the complete isolation of the area eliminates all hybridization of this with other species of the complex.

Most interesting of the data presented by this species is the evidence that Quercus Gambelii, representing the only complex of broad-leaved white oaks in the Rocky Mountain area, has its closest relative in Q. macrocarpa in the eastern part of the United States. Extending east from its original home in the Rockies, the fulvicollis complex changed from Gambelii to macrocarpa, and only subsequently to Q. alba and the chestnut oaks. Indeed, the change from Gambelii to macrocarpa was so simple that the one species, insulensis, bridged the gap without having to develop mutant forms which would adopt the new host.