There are, however, some observations which lead us to hope that the time may not be far distant when we shall have so £Eur obtained a knowledge of the conditions on which volcanic action depends as to be able to form some judgment as to its manifestations in the future at any particular locality. But we must recollect that these conditions axe very numerous and complicated, and that some of them may lie almost entirely outside our sphere of observation; hence hasty attempts in this direction, such as have recently been made, are to be deprecated by every true lover of science. Concerning the eruptions that have taken place at those volcanic centres which have been known from a remote antiquity, we have records from which we can determine the intervals separating these outbursts and their relative violence. A critical examination of these records leads to the following conclusions:-- (1.) A long period of quiescence is generally followed by an eruption which is either of long duration or of great violence. (2.) A long-continued, or very violent eruption is usually followed by a prolonged period of repose. (3.) Feeble and short eruptions usually succeed one another at brief intervals. (4.) As a general rule, the violence of a great eruption is inversely proportional to its duration. It will be seen that these general conclusions are in perfect harmony with the theory that volcanic outbursts are due to the accumulation of steam at volcanic centres, and that the tension of this imprisoned gas eventually overcomes the repressing forces which tend to prevent its manifestation. Before astronomers had learnt to determine all the conditions on which the production of eclipses depends, they had found that these phenomena succeed one another at regular intervals. The discovery of such astronomical cycles was a great advance in our knowledge of the heavenly bodies, and in the same way the determination of these general relations between the intensity and duration of volcanic outbursts and the intervals of time which separate them may be regarded as the first step towards the discovery of the laws which govern volcanic activity. In the actual determination of the conditions upon which the occurrence of volcanic eruptions depends, it must be confessed, however, that very little has as yet been done. This is in part due to the fact that some at least of these conditions lie beyond the limits of direct observation. But it must also be admitted, on the other hand, that little has been as yet accomplished towards the careful and systematic observation of those phenomena which may, and probably do, exert an influence in bringing about volcanic outbursts. [Sidenote: INFLUENCE OF ATMOSPHERIC CONDITIONS.]