To return to the huanaco. After tracing the dying instinct back to its hypothetical origin--namely, a habit acquired by the animal in some past period of seeking refuge from some kind of pain and danger at a certain spot, it is only natural to speculate a little further as to the nature of that danger and of the conditions the animal existed in. If the huanaco is as old on the earth as its antique generalized form have led naturalists to suppose, we can well believe that it has survived not only a great many lost mammalian types, but many changes in the conditions of its life. Let us then imagine that at some remote period a change took place in the climate of Patagonia, and that it became colder and colder, owing to some cause affecting only that portion of the antarctic region; such a cause, for instance, as a great accumulation of icebergs on the northern shores of the antarctic continent, extending century by century until a large portion of the now open sea became blocked up with solid ice. If the change was gradual and the snow became deeper each winter and lasted longer, an intelligent, gregarious, and exceedingly hardy and active animal like the huanaco, able to exist on the driest woody fibres, would stand the beat chance of maintaining its existence in such altered conditions, and would form new habits to meet the new danger. One would be that at the approach of a period of deep snow and deadly cold, all the herds frequenting one place would gather together at the most favourable spots in the river valleys, where the vegetation is dense and some food could be had while the surrounding country continued covered with deep snow. They would, in fact, make choice of exactly such localities as are now used for dying places. There they would be sheltered from the cutting-winds, the twigs and bark would supply them with food, the warmth from a great many individuals massed together would serve to keep the snow partially melted under foot, and would prevent their being smothered, while the stiff and closely interlaced branches would keep a roof of snow above them, and thus protected they would keep alive until the return of mild weather released them. In the course of many generations all weakly animals, and all in which the habit of seeking the refuge at the proper time was weak or uncertain in its action would perish, but their loss would be an advantage to the survivors. It is worthy of remark that it is only at the southern extremity of Patagonia that the huanacos have dying places. In Northern Patagonia, and on the Chilian and Peruvian Andes no such instinct has been observed. CHAPTER XXII. THE STRANGE INSTINCTS OF CATTLE.