1905.] Wheeler, Ants of the Genus Dolichoderus. 3^5 



The habits of the palaearctic D. quadripunctatus have been studied 

 by Latreille and Forel. Latreille's observations were embodied in a 

 remark that the species is frequently found on old tree-trunks and 

 that its societies are very small ("tres-peu nombreuse").i 



Forel has given a much more satisfactory account of this ant. 2 

 He finds that it is one of the components of a walnut-tree ant -fauna, 

 which comprises also Colohopsis truncatd and Leptothorax affinis. It 

 nests in the wood or bark of dead branches in colonies which are 

 rather small though much larger than the colonies of Colohopsis. As 

 soon as the morning has grown sufficiently warm the workers descend 

 the trees in files and distribute themselves over the surrounding 

 plants where they lap up the sweet exudations from flowers, leaves, 

 and twigs. Forel did not see them attending aphides. When dis- 

 turbed they timidly crouch in the crevices of the bark. He is in- 

 clined to believe that the close superficial resemblance between the 

 minor workers of Colohopsis truncata and the workers of D. quadri- 

 punctatus is due to mimicry. Not only are these two ants, belonging 

 to very different genera and even subfamilies, the only European 

 species with spotted gasters, but they closely resemble each other also 

 in gait, stature, and behavior. Forel found as many as nine different 

 nests of D. quadripunctatus in the dead branches of a single walnut 

 tree. When workers from seven of these were placed in the same 

 box, there were no hostilities. He concludes, therefore, that all of 

 these nests belonged to the same colony. In other words, the colonies 

 of this species are polydomous, but each colony contains several 

 dealated females. 



While Forel's statements go to show that the European Dolicho- 

 derus, like many of the tropical species, is strictly arboreal, the ob- 

 servations I have been able to make on our American species reveal 

 some interesting and important differences. These observations 

 show very clearly that our Dolichoderi represent at least three separate 

 species, which are ethologically as well as taxonomically quite dis- 

 tinct from the palaearctic form. 



The workers of D. plagiatus and its subspecies and varieties are 

 occasionally found in small companies, running over the leaves of 

 bushes and young trees in the sunny clearings of our northern woods. 

 They lick the surfaces of the leaves wherever they are covered with 

 honey-dew, i. e., the excrement of aphides, and undoubtedly also 



' Histoire Naturelle des Fourmis, Paris, 1802, p. 181. 



* Fourmis de la Suisse, 1874, pp. 286-288, and Vari^t^s Myrm^cologiques, Ann. Soc. Ent. Belg., 

 Tome XLV, 1901, pp. 380-382. 



