BEHAVIOR OF ANTS 427 



a form of parasitism in which the adaptations are practically 

 all on the part of the parasites. Wasmann concludes that "the 

 symphilic instincts of ants and termites are phylogenetically 

 acquired differentiations and specializations of the general brood 

 — and adoptive instincts of these social insects. Through their 

 hereditary relation to the adoption and nurture of certain true 

 guests they are to be regarded as specialized instincts. Assump- 

 tion of these instincts is necessary to an understanding of the 

 facts." Wasmann reviews the various adaptive characters of 

 the symphiles (trichomes, exudate organs, antennal structure, 

 etc.), and concludes that these could be developed only by a 

 kind of selection, which he calls " amical selection," exercised 

 by the host ants, and that this kind of selection is merely a func- 

 tion of the symphilic instinct. Amical selection is supposed to 

 be built on natural selection and to heighten and perfect the 

 adaptation of the symphile to its host. From the standpoint 

 of the host, however, amical selection is supposed to be inde- 

 pendent of, and to be even inimical to, natural selection. 



Wasmann "^ reviews Holmgren's recent studies on termites 

 with special reference to his theory of the role of the exudates 

 in the behavior of these insect's. The exudate organs and tissues 

 were first described by Wasmann in 1903 in the guests of ants 

 and termites. The tissues consist of fatty or adipoid elements, 

 which seem to represent merely a modified portion of the corpus 

 adiposum, and give oft" through the cuticle to the surface of 

 the body a thin liquid which resembles and in some cases may 

 actually be blood plasma. Exudate tissue has now been found 

 by Holmgren also in the termites themselves, and is most highly 

 developed in the queen. He regards the exudate as a nutritive 

 substance which is eaten by the termites, and owing to its different 

 constitution in the different castes, sexes, and developmental 

 stages of the individuals in the colony acts as a varying stimulus 

 which causes the insects to vary their behavior adaptively. Thus 

 he maintains that the exudate tissue is at the basis of caste 

 development. The existence of a similar tissue in myrmecophiles 

 and probably also in ants themselves, as indicated by certain 

 guests {Myrmecophila, Oxysoma, Attaphila, Leptothorax emer- 

 soni) which lick the surfaces of their host ants, suggests a wide 

 and interesting application of Holmgren's theory. 



