4o8 Bulletin American Museum of Natural History. [Vol. XXII, 



inaccessible to the insects. The densely tufted moss is much like fur 

 and seriously impedes the insects in going to and from the nest, so that 

 in some nests invaded hy the Polytrichmn, a second zone of entrances 

 had been constructed nearer the summit and along the inner edge of 

 the moss zone. In one of the large twin nests (No. 6) shown in PI. 

 LXVI the ants constantly made a bridge of a fallen log running 

 from the ground to the umbilicus. It is certain, moreover, that 

 the moss, which absorbs and retains quantities of water and there- 

 fore reduces the temperature of its surroundings, must impair the 

 usefulness of the mound as an incubator. 



5. The Polytrichmn carpet continues to grow at its upper edge 

 until it envelops the summit of the mound and extinguishes the ant 

 communit}^. In one mound of this description, with an uncovered 

 area only 20 cm. in diameter, a few lethargic ants still lingered in the 

 earth of the summit; from other similar and older nests they had 

 completely disappeared. While the dense carpet continues to expand,, 

 herbaceous, and even woody plants of many species begin to settle 

 on the mounds and eventually in great part completely replace the 

 moss. The mound subsides somewhat, probably owing to the col- 

 lapsing of its galleries and chambers, and becomes flatter and less 

 conical. It may still be recognized, however, after the lapse of years 

 as was shown by one mound which supported bushes with stems 5-6 

 cm. in diameter. On these advanced nests at least three other mosses 

 {picranella heteromalla, Catherinea angustata, and Hypnum recurvans) 

 may be seen growing singly or in company. 



Do the ants emigrate and seek a new nesting site when the invasion 

 of the Polytrichum becomes intolerable, or do they gradually die off 

 without deserting the mound which they have reared with such 

 diligent solicitude? This is a difficult question to answer, because 

 a single exsectoides colony inay extend over several mounds and these 

 may be connected by subterranean galleries. The fact that the ants 

 in mounds invaded by Polytrichum have all the appearance of relicts 

 smitten with the weariness and dejection so characteristic of old and 

 depauperate ant communities, certainly favors the view that the 

 insects die o& in situ. 



It is equally difficult at pi'csent to answer the question as to the 

 time required to bring about the changes described in the above 

 stages. The age limit of the nest was considered by McCook (/. c, pp. 

 265-267), who concluded that he had "good reason to believe that 

 some of the hills are at least thirty years old." This estimate 

 is probably as close as any that can be obtained until some one 



