in many minute details of venation, thoracic sculpture, abdominal and antennal characters; in atrimentus the parapsides are more convergent at the scutellum, the scutellum is less elongate, less rugose, with the foveal groove smoother and less distinctly bifoveate than in palustris. In the male of atrimentus the abdomen is not as small as in palustris. Atrimentus occurs on a white oak in one distinct faunal area; palustris on black oaks in a very different faunal area. It is exceedingly significant that the galls of the two species show similarity of structure, for as has been repeatedly shown the gall is an expression of the physiological nature of the insect. Physiologically then the two species are related, just as they are morphologically, but physiologically they differ even more markedly than morphologically. For though the galls are similar hollow bladders in the leaf, the Pacific Coast species does not have the thick and succulent walls of palustris; while the attached larval cell of atrimentus is distinctly unlike the remarkable, loose cell of palustris. Both forms are bisexual generations, developing very quickly in early spring on the unfolding leaves. The other generation of palustris, namely Andricus palustris form compressus (Gillette), is a wingless, agamic species forming a small, hollow, egg-shaped, separable gall on the leaves in late summer, the gall dropping to the ground in autumn, the insect emerging very early in the spring. It is not unlikely that atrimentus has a similar history, but it will be important that some one determine exactly what differences exist in the life-histories of the two insects.
Andricus attractans, new species
Plate XXIV, Figure 2
