If he means that the only legitimate data of geologists are facts of observation, classified and recorded, well and good; but to deny that they deal largely in hypotheses, and use them constantly as the premisses for inferences which are equally hypothetical, is palpably absurd. First of all we are to "assume the principle of uniformity" which Lyell is said to have established on an unassailable basis and to have made the fundamental axiom of geological science. He "has shown conclusively that while causes identical with ... existing causes will, _if given sufficient time_, account for all the facts hitherto observed, there is not a single fact which _proves_ the occurrence of a totally different order of causes." [81] This, however, is (1) limited to the period of geology which gives record of organic life, and not to the earlier astronomical period; nor (2) does it exclude changes in temperature, climate, distribution of seas and lands; nor (3) does it "_affirm positively_ that there may not have been in past ages explosions more violent than that of Krakatoa; lava-streams more extensive than that of Skaptar-Jokul, and earthquakes more powerful than that which uplifted five or six hundred miles of the Pacific coast of South America six or seven feet." [82] Now, seeing that all these cataclysms have occurred within the brief limits of most recent time, compared with which the period of pretended uniformity is almost an eternity, what sort of presumption or probability is there that such occurrences should have been confined to historical times; and is not the presumption all the other way? Again, it is largely on the supposition of this antecedently unlikely uniformity, that Mr. Laing argues to the antiquity of life on earth; whereas Lyell's conclusion warrants nothing of the kind, being simply: that present causes, "_given sufficient time_," would produce the observed effects. [83]