The coast-line of Co. Donegal will illustrate the foregoing remarks. On the southern and western side of the Slieve League promontory there is no real Guillemot station; only on the northern side--the quartzite in the vicinity of Tormore--are the birds to be found in large numbers. Northwards from here, a wild and rugged coast is passed over before other stations are reached--at the eastern end of Tory Island and on Horn Head; and beyond this, to the east, there are none, not even on the old rocks that form the promontory of Inishowen. Why, we ask, do countless numbers crowd the ledges of Horn Head, whilst they are absent from the precipices of Slieve League; why, too, are they absent from the granite cliffs of Owey? The reason is not far to seek. Either the face of the cliff is made up of a series of broken precipices, or the face of the precipices is too smooth, or the otherwise suitable ledges are situated too near the water, or the water recedes from the base of the cliff at low tide. Many miles of rock-bound coast are thus useless for the purpose of reproduction. Now when we bear in mind how large an expanse of coast is formed of blown sand or of rocks of low altitude, and how many miles of cliff fail to supply the three essential conditions that we have been considering, we can see that suitable breeding stations must be limited both in number and extent. From a wide expanse of ocean hosts of individuals are therefore obliged to converge at certain definite points; and hence, each recurring season, there must arise a competition for positions at the station, just as there is competition between individual Buntings for positions in the marsh. And the ability to obtain a position upon a suitable ledge involves, in the first place, an impulse to search for it; in the second place, an impulse to dwell in it; and in the third place, an impulse to resist intrusion upon it. It would be useless for an individual to be pugnacious if it had no fixed abode; equally useless for it to establish itself on a particular ledge if it had no power to defend it--all of which implies an inherited nature similar to that of the Bunting. But the proximate end to which the competition is directed is not alike in the case of both species. In the case of the Guillemot it has reference solely to the piece of rock whereon the egg is laid; in the case of the Bunting to a piece of ground capable of furnishing an adequate supply of food for the young; and the reason for the difference is this, that there is always an abundance of food in the water beneath the cliff, but breeding stations are scarce, whereas there is always an abundance of situations in the marsh in which the Bunting can place its nest, but the supply of food varies and at times can only be obtained with difficulty.