stem of the tree, and then the covers worked off of the breeding jars. The insects are thus placed on the tree and confined immediately upon emergence from the galls. The nets may be left on the trees for three or four weeks, thus insuring quite completely that no other gall-wasps may attack that tree, but to keep the tree covered for a longer period will usually affect the vigor of the plant and prevent the growth of the galls. Adler, in working out the life histories of European cynipids, employed only smaller coverings for the trees, but by the use of nets large enough to cover the entire tree it is possible to give a much wider range of activity to the insect; and by placing the breeding jars directly into the nets the danger of injuring these delicate insects by handling is entirely avoided.
The adults of most species of cynipids have very little vitality, and to effect oviposition they must usually reach the part of the plant in which the egg will be laid within a few hours after emerging from the gall. Insects kept a couple of days after emergence rarely have vitality enough left to lay eggs. The occurrence of a rainy or even foggy day at the time of emergence of a species (most of the adults of any particular species will emerge within a very few days of each other) will prevent immediate oviposition, and by the time the weather has cleared the insects have lost all vitality and will not attempt to climb over the trees. In spite of all care, to obtain galls experimentally is a difficult process, often proving unsuccessful. From about two thousand gall-wasps placed upon plants the first year of my work, when the trees were newly transplanted and not growing very vigorously, I secured only twenty galls. The factors influencing the growth of these deformations are hardly at all understood, but it is very likely that the trees on which they are to grow must be in very vigorous condition. An abundance of insects of any one species will usually be necessary to secure any amount of results.
HISTORICAL
The alternation of two quite different forms in successive generations of an organism was discovered first in 1819 by Chamisso in a tunicate of the genus Salpa. “A Salpa-mother is not like its daughter or its own mother, but resembles its sister, its granddaughter, or its grandmother” was a remarkable statement to hear in a day when spontaneous generation theories had only recently been widely discarded and when the resemblance of all offspring to their parents seemed the foundation stone of biological science. Though plenty of other observations soon confirmed Chamisso's discovery. it was not until the time of the extensive observations of Lichtenstein with aphids and of Adler with Cynipi-
