ten days or two weeks. The pupation period of the insect is very short, within a few days the adult emerges and within another few days the galls are mostly completely withered away or have decayed.
The adults emerge from mid-May to June 4, depending on the latitude and the progress of the season. The insects are produced in about equal numbers of the sexes. They copulate almost immediately upon emergence and run rapidly up the branches to the young leaves, on the under sides of which they settle to oviposit, often remaining there for several days before they die. In all, only about a month has passed from the time of the first appearance of the young gall to the time of oviposition. Over ten months will be required for the alternate generation to reach full maturity.
The experiments by which I was fortunate enough to discover the relations of the two generations of this species were made under the strictly controlled conditions which I have described in the introduction. The tree on which the sexual adults were put was covered with a net for almost a month and there was no chance of other insects having reached it during that time, nor is there much likelihood of the galls having been produced by insects from any other source. Of the thirty small oaks in the greenhouse, none produced any galls that season (1918) except the one on which the palustris had been isolated. Several scores of the adults of that form were observed to oviposit on the under surface of the leaves on which eight galls of Philonix compressa were found the first of September. The small size of the galls had prevented their detection earlier, for it is likely that they appear by the first of August.
P. compressa was a “species” known definitely from only two stations, Westchester Co., N. Y., and Ames, Iowa, while palustris had been known for sixty years to be a very abundant species throughout eastern America. It was rather surprising to find an apparently rare, local species to be the alternate form of one of the commonest of widely distributed galls. Fortunately, I recalled definitely the location of the very trees on which I had found the palustris galls from which the experimental material had been bred. I made a trip to these trees and found the leaves bearing an abundance of the compressa galls. Then I found the galls in the Millett Thompson Collection, taken in Massachusetts or Rhode Island; these were figured in the Thompson Catalogue but the collection was not recorded in the text. After that, I collected compressa galls at Melrose Highlands, Forest Hills, and Blue Hills, which are in the neighborhood of Boston, and always found it abundant at that date. The small size of the galls and the fact that they are very quickly decidu-
