Many other facts indicate that throughout the enormous periods required for the development of the varied forms of life upon the earth, the great phenomena of nature were but little different from those that prevail in our own times. The slow and gentle processes by which the various vegetable and animal remains were preserved are shown by the perfect state in which many of the fossils exist. Often trunks of trees, cycads, and tree-ferns are found standing erect, with their roots still imbedded in the soil they grew in. Large leaves of poplars, maples, oaks, and other trees are often preserved in as perfect a state as if gathered by a botanist and dried between paper for his herbarium, and the same is especially the case with the beautiful ferns of the Permian and Carboniferous periods. Throughout these and most other formations well-preserved ripple-marks are found in the solidified mud or sand of old seashores, differing in no respect from similar marks to be found on almost every coast to-day. Equally interesting are the marks of rain-drops preserved in the rocks of almost all ages. Sir Charles Lyell has given illustrations of recent impressions of rain-drops on the extensive mud-flats of Nova Scotia, and also an illustration of rain-drops on a slab of shale from the carboniferous formation of the same country; and the two are as much alike as the prints of two different showers a few days apart. The general size and form of the drops are almost identical, and imply a great similarity in the general atmospheric conditions. We must not forget that this presence of rain throughout geological time implies, as we have seen in our last chapter, a constant and universal distribution of atmospheric dust. The two chief sources of this dust--the total quantity of which in the atmosphere must be enormous--are volcanoes and deserts, and we are therefore sure that these two great natural phenomena have always been present. Of volcanoes we have ample independent evidence in the presence of lavas and volcanic ashes, as well as actual stumps or cores of old volcanoes, through all geological formations; and we can have little doubt that deserts also were present, though perhaps not always so extensive as they are now. It is a very suggestive fact that these two phenomena, usually held to be blots on the fair face of nature, and even to be opposed to belief in a beneficent Creator, should now be proved to be really essential to the earth's habitability. Notwithstanding this prevalence of warm and uniform conditions, there is also evidence of considerable changes of climate; and at two periods--in the Eocene and in the remote Permian--there are even indications of ice-action, so that some geologists believe that there were then actual glacial epochs. But it seems more probable that they imply only local glaciation, owing to there having been high land and other suitable conditions for the production of glaciers in certain areas.