296 PROCEEDINGS OF THE ACADEMY OF [1883. 



mouth of the Cannon Ball River, have no such covering. Prof. 

 Todd is inclined to think that the gravel roofing is selected from 

 the nest vicinage and placed iipon the mound ; but I believe the 

 stones to have been excavated from the underground galleries, 

 granaries and rooms, and brought up therefrom. However, I 

 think the construction of a roof by selection to be quite within 

 the ability of the Occidents, as I have observed them carrying 

 pebbles up, down and around the mound in all directions after 

 issuing from the gate. 



3. Harvesting habit. — Among the pebbles sent to me are a 

 number of husks, etc., of various seeds which appear to have been 

 taken from the kitchen-middens or refuse-heaps of the formicary. 

 These indicate that the Dakota emmets, like the more southern 

 examples, are harvesting ants. Mr. Thomas Meehan, to whom 

 was referred a small quantity of the deBi'is collected from 

 the margin of a nest by Prof. Todd, reports that there are no 

 seeds among the pebbles, but that there are a number of calices 

 and undeveloped capsules of a leguminous plant. Dalea alopecur- 

 oides, which is common on the American plains. I was puzzled 

 to explain why such intelligent creatures should be detected in 

 harvesting immature seeds, until, upon inquiry, I found that 

 leguminous plants have a succession of flowers, so that there may 

 be mature seeds and flowei'S on a plant at the same time. Mr. 

 Meehan actually found upon a specimen of the above plant in the 

 Academy's Herbarium, both the flower and the fully developed 

 seed; indeed, the two appear to occur upon the same spike. It 

 is thus evident that the ants were not harvesting out of season, 

 but were occasionally deceived, and cast out to the refuse-heaps 

 the calices that contamed no edible seed. 



