A step in the direction of the social wasps seems also to have been taken by a small group of solitary Vespoids, allied to the Eumeninæ, namely the Zethinæ. These insects, which have been studied recently in Brazil by Ducke, in British Guiana by Howes and in the Philippines by F. X. Williams, have abandoned the use of earth as nest material and employ instead small bits of leaves or moss (Fig. 27). With such vegetable material Zethus constructs a beautiful nest with one or several tubular cells, and therefore approaches the social wasps which make their nests of paper, a substance consisting of fine particles of wood agglutinated with an oral secretion. The egg is laid loosely in the bottom of the cell and, according to Williams' account of the Philippine Zethus cyanopterus, the larva is fed from day to day on small caterpillars, which have been in part eaten by the mother. She faithfully guards the larva and, while it is small and there is still ample room, sleeps in the cell. She closes the latter as soon as the larva is full grown and proceeds to build another.

Each of these cases of progressive provisioning may be regarded as a very primitive family, or society, reduced to its simplest terms, i. e., to a mother and her single offspring. The seasonal or local conditions of the environment, in so far as they affect the abundance or scarcity of prey, have led on the one hand to mass provisioning and therefore to an exclusion of the mother from contact with her growing offspring, and on the other to an establishment

(After

Zethus cyanopterus of the Philippines and its nest. A, Adult wasp x 2; B, the beginning of a cell. It is attached to a twig by a mass of well-masticated leaf-bits and the wall of the cell is made of shingled leaf bits. (Somewhat enlarged); C, the first cell of the nest completed x 4 ;D, a four-cell nest showing roof-like structure and one emergence hole x 4. (After F. X.

Williams).

of that very contact. This, again, has developed an immediate interest of the mother in her young comparable with what we observe in many birds. Probably this interest is aroused and sustained in the mother wasp by simple, pleasurable, chemical (odor) or tactile stimuli emanating from the egg and larva, but whatever be the nature of the stimuli involved, I believe that we shall have to admit that the egg and the larva have acquired a "meaning" for the mother wasp, and so far as the egg is concerned, this seems to be true even in the species that practice mass provisioning. We noticed that many solitary Vespoids (Eumenes, Odynerus), before they bring in their prey, carefully attach the egg by a string

86

THE SCIENTIFIC MONTHLY

to the ceiling of the cell. This singular performance has been variously interpreted. Fabre and others regard it as a device for preventing the delicate egg from being crushed by the closely packed and sometimes reviving prey, on the same principle that in a crowded room an electric light bulb attached by a cord to the ceiling would be less easily crushed than one rigidly fixed to the walls or the floor. Others regard the filament as a device for keeping the egg free from the occasionally very damp walls of the cell. Ferton has recently shown that Bembix mediterraneus glues its slender egg to the floor of the cell in an erect position and with the base carefully supported by a cluster of large sand-grains (Fig. 28 A), and that Stizus errans glues its egg in a similar position

A. Egg of Bembix mediterraneus with its base supported by a cluster of large sand-grains. B. Egg cf Stizus errans glued to the upper surface of a carefully selected pebble. (After C. Ferton).

to the top of a small, carefully selected pebble placed on the floor of the cell (Fig. 28 B). Parker's description of the egg of our Microbember monodonta seems to indicate a condition similar to that of B. mediterraneus. In all these cases we seem to have an arrangement for keeping the very easily injured egg as free as possible from contact with the rough, sandy walls of the burrow. But even the Sphecoids and Psammocharids, which practice mass provisioning, attach the egg to a particular part of the victim and in such a position that the hatching larva can attack it at its most vulnerable point. Ferton, especially, has made a very interesting study of this type of behavior.

The following facts also indicate very clearly that the mother wasp may be aware, not only of sexual differences among her own eggs but also of the differences in the amount of food required by the resulting larvæ. Bordage, while investigating the Sphecoids of the Island of Réunion, found that three of the species, Pison argentatum, Trypoxylon scutifrons and T. errans, could be readily induced to make their cells in glass tubes placed between the

Phylogenetic tree of the various genera and families of Vespoids. (After
Ducke, with modifications).

88

THE SCIENTIFIC MONTHLY

pamphlets of his library. The Pison, under natural conditions, builds elliptical clay cells and provisions them with spiders, whereas the species of Trypoxylon nest in hollow twigs and the interstices of wall but use the same kind of prey. All these species adapted themselves to the glass tubes in the same manner. Each of them plugged the end of the tube with clay and divided the lumen into successive cells by building simple clay partitions across it. After the cells had been provisioned, Bordage observed that the first of them were longer by half a centimeter and contained more prey than those provisioned later, and he was able to show that the larvæ in the larger, more abundantly provisioned cells produced female, the others male wasps. Similar observations have also been published by Roubaud on the Congolese Odynerus (Rhynchium) anceps, which makes clusters of straight, tubular galleries in clay walls and divides each gallery into several cells by means of clay partitions. In this case also the first cells are much longer than the later, though there is no difference in the quantity of small caterpillars allotted to the different eggs. But Roubaud was able to prove experimentally that even when the amount of food is so greatly decreased that the larvæ produce adult wasps of only half the normal size, their sex is nevertheless in no wise affected. It would seem therefore that the mother wasp must discriminate between the deposition of a fertilized, femaleproducing and that of an unfertilized, male-producing egg, and regulate the size of the cell and in some instances also the amount of provisions accordingly.

In the accompanying diagram (Fig. 29), taken from Ducke but somewhat modified, I have indicated the hypothetical family tree. of the solitary and social Vespoids. The genera below the heavy horizontal line are solitary, and among them Eumenes and Odynerus seem to be nearest to the original ancestors, because they are very similar to the social forms in having longitudinally folded wings and in other morphological characters. It will be seen that there are six independent lines of descent to the social forms above the heavy line and that the genera plotted at different levels represent various stages of specialization as indicated by the nature of the materials and types of structure of the nests. With the doubtful exception of a few Stenogastrinæ, all the social wasps make paper nests consisting wholly or in part of one or more combs of regular hexagonal cells, in which a number of young are reared simultaneously.

(To be continued)