But by far the most satisfactory way of studying the ants is in their native haunts on the barrens, where I had ten nests under observation. One of these was so situated that it received the direct rays of the sun all day, and was protected from north and east winds by dense, low shrubbery. On sunny days, even with a cool wind from the north, when taking my seat in this sheltered spot, I would soon become uncomfortably warm. This hill was always active whenever I visited it, while in other localities the ants would often be all housed. Around this active nest I stuck stems of millet eighteen inches high, surmounted by the close-packed heads. The ants climbed the stems, loosened and secured the seeds, and stored them within the nest. They worked vigorously, sometimes twenty or more on one head pulling away at the seeds. In my artificial formicary they did not mount the stems, even when the heads were not more than three or four inches from the ground, but seeds that I scattered in the jar were always taken below. I threw down a handful of apple-seeds near the entrance of the active hill on the barrens. This immediately attracted a large number of excited ants. They rushed to the seeds in a warlike attitude, and began carrying them off, depositing them two or three feet away. But as soon as the excitement caused by the sudden pouring down of the seeds had subsided, they seemed to comprehend that they had been throwing away good seeds; and now, changing their tactics, not only carried the remainder into the nest, but finally brought back and stored all those that had been thrown away. On excavating the nests we found granaries of seed scattered irregularly throughout to the depth of twenty-two inches below the surface of the ground: some were near the surface, and a few sprouted seeds were scattered about in the mound. The mound is usually not more than four to six inches above the level of the ground. The great majority of nests that I have found are in the low pine barrens--so low that on reaching the depth of two feet the water runs into the cavity like a spring, and stands above some of the granaries. Notwithstanding this wet locality, I found no sprouted seeds in the deeper store-rooms, but only in the warmer mound. On sunny days the larvæ are brought up into the mound and deposited in chambers near the surface, where they receive the benefit of the sun's rays. On cool, cloudy days and in the early morning I found no larvæ near the surface. If the ants are intelligent enough to treat the larvæ in this way, why should they not store seeds where they will not sprout? And when they need to sprout them in order to obtain the sugar they contain, it would take no more wisdom to treat the seed as they do the larvæ--bringing them near the surface to obtain the right degree of heat for the required result.