by wasps from other than the bred material, either by distance from any other plants of the sort (the Cynipidæ almost never fly and most likely never travel more than a very short distance), or by a shelter or covering. It is not desirable to have the plants indoors, for they do not develop at normal rates under these abnormal conditions and the insects will not attack the plants unless they are at the proper state of development. Wild roses and blackberries are easily transplanted, but it is profitable to secure nursery-grown oaks of several species, four or five years old, for the work. These should be growing thriftily before experiments are started, or else galls will probably not be obtained. To wait a year or so after the trees are transplanted is one of the almost necessary problems presented by the work.
If the alternate generation of a species which is being studied is to lay its eggs in the flowers or acorns of the oak, recourse must be had to older trees growing in the open, such additional precautions then being taken as will insure as far as possible the immunity of the part of the tree used from attack by any other gall-wasps than those under observation. Some species will oviposit in the roots of trees. These will be detected instantly, for they are positively geotropic and will start immediately to climb down the tree. These forms, while in the breeding jars, should be placed at the bases of trees having living roots near the surface and covered by only a very little loose earth or, better, leaf-mold, and which have the covering nets extending entirely to the ground, into which the cloth is pegged.
When the insects have emerged within the breeding jars, the trees should be covered with net bags. I employ large cylinders made of a closely-woven cheese-cloth, the cylinder open at the bottom and kept in shape by two heavy wire rings of the same diameter as the cylinder and sewed into place, one in the closed end of the bag and the other about a foot from the open end. These cylinders should be large enough to cover a large part of the tree or even the whole tree, for then the insects are given greater freedom in the choice of the spot for oviposition. Tapes sewed to the closed end of the bag will hold it to some beam or other support above the tree. A strip of sheet celluloid may be fastened into the cloth by a waterproof solution (a few drops of castor oil added to sheet celluloid dissolved in acetone) and the window thus supplied is a great convenience for observing the insects ovipositing. The breeding jars may be placed on wooden blocks or other supports on the ground at the base of the tree, the net drawn down to include the jars, the open end of the net brought together and securely fastened around the main
