AMERICAN IIYMENOPTERA. 209 



exercised an evident diversion, influencino; 18 per cent, of the lulls to 

 present the lnn<:ost slope to the north. In Section 3, with (he pre- 

 vailinsi slope of land toward the northwest, the north gains but a small 

 per cent, of the long slopes of the hill. On the whole, the above 

 exhibit strengthens the impression made upon my mind that the ants 

 have some regard to the points of the compuss in building their 

 mounds, although there is certainly nothing like the regularity that 

 Huber attributes to the Swiss yellow ant hills, which makes them a 

 safe com; ass to the mountaineers in foggy weather.* 



Galleries. — Much attention was given to the structure and extent 

 of the galleries. The mode of erecting them above ground has al- 

 ready been fully illustrated. 1 had half sections of a number of hills 

 sawed down and cut away in order to study the arrangement of the 

 galleries; and to obtain accurate figures for comparison and more 

 favorable study, I had a few of these section views photographed. 

 Some of the hills were cut east and west, some north and south, others 

 at raiHluai. I ftuiid that ([uite generally the greatest regularity in 

 the direction of the galleries was north and south, although one hill 

 showed equal regularity east and west. This fact may be obsi^ved by 

 examining carefully the two hills of PI. IV, the lower one of which 

 is cut north and south, the upper east and west. The double hill at 

 PI. V will illustrate the same feature. The lower hill is cut east and 

 west and a quarter section taken from the upper one thus exposing in 

 one view the result of both the north and south, and east and west 

 cutting. • The tendency of the galleries is to cross the hills at right 

 lines. They have the appearance of being laid regularly one above 

 the other. This feature may be noted most satisfactorily in PI. IV, 

 lower hill, in the shaded portion of the perpendicular section, at the 

 right and toward the base of the cone. The openings of the galleries 

 were carefully cleansed of soil, so that they might present as natural 

 an appearance as possible before they were photographed. Neverthe- 

 less, much to my regret, the plates fail to siiow perfectly the peculiar 

 structure of the interior ol' the hill. A tolerably accurate knowledge, 

 however, can be fornit'd from the plates. 



Underground Cialleries.— Thus far we have been dealing with 



*I had intencled to williliold the observations bearing upon this interesting 

 point, (and indeed other details herein recorded), until furtlier investigation 

 should lead to some definite conclusion. But, in the hope of stimulating some 

 one nearer the field to take up the inquiry, and influenced by the publicly ex- 

 pressed wishes of several eminent naturalists for all the information to be had 

 concerning the habits of ants, I have thought it better to submit my notes even 

 at the risk of burdening this paper with matter that may be valueless. 



TRANS. AMER. EJ(T. SOC. VI, (30) NOVEMBER 1877. 



