H YMENOP TERA 889 



Clistogastra the larvas are maggot-like in form and have no legs. The 

 pupae are of the exarate type, that is, the legs and wings are free, as 

 in the Coleoptera. With many species the larva, before changing to 

 a pupa, spins a cocoon about its body. With some this cocoon is 

 composed of comparatively loose silk, and resembles somewhat the 

 cocoon of a moth. In others the cocoon is of a dense parchment-like 

 texture, and in still others it resembles a very delicate foil. 



Parthenogenesis. — The production of young by females that have not mated is 

 known to occur in members of several families of this order. In some species the 

 young thus produced are all males; in others they are all females; and in still 

 others both males and females are developed from unfertilized ej"gs. Among the 

 well-known examples of parthenogenetic reproduction are the following. Some- 

 times a queen honey-bee produces eggs before she has mated; from such eggs 

 only males are developed. The eggs produced by fertile worker bees and fertile 

 worker ants, neither of which mate, develop only into males. In certain gall- 

 flies there is an alternation of a generation consisting of males and females and a 

 generation consisting only of females, which reproduce parthenogenetically; 

 the young of the latter are males and females. In some species of the Tenthred- 

 inidce the reproduction is believed to be entirely parthenogenetic, males of these 

 species being unknown. 



Polyembryony. — In several genera of minute parasitic Hymenoptera the 

 number of young produced is not dependent upon the number of eggs laid, for 

 with these insects many embryos are developed from a single egg. This type of 

 development is termed polyembryony; and has been investigated by several 

 workers. A recent paper on this subject is that of Dr. R. W. Leiby {'22), in 

 which there is a list of the earlier papers. Dr. Leiby traced the development of 

 Copidosoma gelechice, a parasite of the solidago gall-moth, the insect that makes 

 the spindle-shaped gall on golden-rod (Fig. 769). The parasite oviposits in the 

 eggs of the moth which it finds on leaves or stems of goldenrod in early fall. 

 By the arrival of cold weather the developing egg or parasite body is found as a 

 polynuclear mass within the completely formed host embryo, which has developed 

 synchronously. When the host larva hatches in the spring the parasite body is 

 found lodged in the fat-body of the host. As the host larva grows the polygermal 

 mass becomes a mass of embryos, which are later set free into the body-cavity of 

 the host larva as parasitic larvae. These larvae feed upon the body content of the 

 host devouring the blood, muscles, fatty tissue, and in fact everything except the 

 chitinous parts. An average of 163 adult-parasites is developed from a single egg. 

 The details of this development are described at length by Dr. Leiby and are illus- 

 trated by many figures. 



In a later paper Dr. Leiby and C. C. Hill ('23) described the development of 

 Platygaster heimales, a parasite of the Hessian fly. From some of the eggs of this 

 parasite a single larva is developed; but from others two larvae are produced. 

 This species is of great interest as illustrating the beginning of polyembryony. 



Aquatic Hymenoptera. — It has long been known that the adults of certain 

 parasitic Hymenoptera descend beneath the surface of water in order to ovi- 

 posit. One of these is a parasite of caddice-worms, the others whose hosts are 

 known lay their eggs in the eggs of various aquatic insects. Most of the observa- 

 tions on these insects have been made in Europe but recently Professors Mathe- 

 son and Crosby ('12) have described the habits of three minute species which 

 have been reared at Ithaca, N. Y.; they also give a list of the known aquatic 

 Hymenoptera. 



The classification of the Hymenoptera. — The classification of the 

 Hymenoptera, i. e., the sequence of the families and the groupings of 

 these families into superfamilies, adopted in this chapter is that 

 recently worked out by Messrs. J. C. Bradley, S. A. Rohwer, and 

 J. Bequaert. 



Authorities are not in agreement as to the proper application of certain generic 

 family and other group names in the order Hymenoptera. The matter is a 



