45 s .-!\'TS. 



Dorylines in its frequent change of dwelling. 1 have already mentioned 

 the summer and winter nests of sun</itiiu'n, hut this ant is also fond of 

 changing its habitation in the >ame wood or field and of moving into 

 nests which it h;i> pillaged. Hence one often encounters sanyuinca 

 workers, in the act of moving their young or sexual forms to new 

 quarters. ( )n .Mich occasions they also carry their slaves in the same 

 manner as they carry small workers of their own species; the ant 

 carried heing held hy the mandibles while she coils herself up and 

 remains motionless under the head and thorax of her carrier. It is not 

 always easy at first sight to distinguish these changes of dwelling from 

 dulotic expeditions. 



2. The American Sanguinea. The typical sanyninca does not occur 

 in North America, but in its stead we have no less than six subspecies 

 and varieties: aseri'a, rubicund a, subnnda, subintegra (Fig. 268), pubc- 

 nila and obtusopilosa, and two species.: pcryandei and uttinda, which, 

 however, might be regarded merely as extreme subspecies. All of 

 these forms have the clypeal border notched, a character which serves 

 to distinguish them from our numerous other Formica: of the rufa, 

 c.rsccta, fusca and pallide-fulva groups. F. uinnda is confined to the 

 Rocky Mountains, where it lives in rather small colonies and never 

 makes slaves. F. peryandci is a more widely distributed species, but 

 seems to be very rare, as only a few of its colonies have been seen. 

 The one from which the types of the species were taken near Wash- 

 ington, D. C., contained also workers of F. pallide-fulva, and one which 

 I found near Colorado Springs contained several workers of subpolita. 

 The sanguined subspecies and varieties cited above present a maze 

 extremely difficult to disentangle taxonomically, and although I have 

 made many observations on dozens of colonies in different parts of the 

 country, I am quite unable to define their ethological peculiarities. I 

 have no doubt that such peculiarities exist, but their accurate definition 

 will require years of observation over a great area. 1 Some of the forms, 

 such as rubicunda, aseri'a and subnnda are preeminently boreal or 

 alpine, others, like subintegra and pubcrula prefer warmer latitudes a,nd 

 lower altitudes. In this general account I shall not endeavor to distin- 

 guish further between the habits of the various forms, but compare 

 them as a whole with the single European type. This comparison will 

 show that tiie American forms are peculiar in more than one particular. 



The colonies of our sanyuinca are quite as frequently slaveless as 



f 



1 1 have recently found ( igoS/) that the workers of one of our subspecies, 

 ascrva, are not slave-makers. The queens of this form of sanguined establish 

 their colonies by kidnapping the pup?e of F. glaclalis, but the workers do not 

 inherit this instinct. Hence the old colonies of aserra are pure, as Forel has 

 nl served dooor). 



