The

the appearance of organic beings on our earth, and their disappearance from it, than any other class of facts." 1 In passing southward over the continent of South America, he recognized another of Nature's classifications: he noticed the frequent recurrence of the fact that a species occupying a given region was replaced to the southward by a closely related species, and this serial arrangement impressed itself strongly upon him. His visit to the Galapagos Islands gave him an almost perfect example of simplicity in the working of Nature's forces. conditions for one of Nature's classifications were perfect. When Lawson, the Vice Governor, had declared to him that the tortoises from the different islands differed from one another, Darwin did not see the significance of the fact. He mixed up his collections from the various islands, and did not dream that there lay before him one of the most remarkable facts that Nature ever revealed to a naturalist. By some happy accident he compared the many specimens of mocking thrushes shot on Charles Island with those from Albemarle Island, and was astonished to find that they belonged to different species. It was not the fact that there were two species of mocking thrush liv

1 Naturalist's Voyage around the World, p. 173.

ing together on the two islands, but that they lived apart, each on its own island, and that they were closely instead of distantly related, - that several islands were stocked, each with its own species, or perhaps variety, of the same kind of animal, that struck him with wonder.

He

These facts haunted him, and drove him to look for others like them. Upon his arrival at home his insect collection proved the same law. had fortunately kept the plants of the different islands separate, and Hooker, at Darwin's request to see whether the law held good for them, found it to be so.1 Darwin had unwittingly carried one of Nature's beautiful classifications home to England with him. Keen insight into the relation of facts to one another had enabled him to recognize three striking examples of Nature's arrangements.

The habit of grouping facts to extract the truth from them was indispensable to Darwin's work, for he constantly dealt with large bodies of facts that were manageable in no other way. It will be seen how difficult it is even for a powerful observer to see facts for which he is. not looking, even though they lie under his feet. An act of classification, to be worth much, must usually be an effort to answer

1 Naturalist's Voyage around the World, pp. 393-398.

a direct question. Science has derived very little or no benefit from the miscellaneous collecting and grouping of facts without any previous notion of what they are likely to reveal. An investigation is usually made for the purpose of answering a definite question, or of verifying an anticipation. With some such end in view, with some principle by which the classification is guided, the result usually reveals not only what was looked for, but frequently still more fundamental characteristics; for it is impossible to throw facts into any order which reveals one truth without dragging others into the light with it. The character of Darwin's work required constant recourse to lists and tables; he appreciated fully both their value and their treachery, and his great ability to recognize all the points brought out, no matter whether he was looking for them or whether they bore directly on the subject which he happened to be investigating or not, made them enormously useful to him.

Darwin's original purpose in measuring the heights of the gravel-capped plains of Patagonia was to ascertain the heights at which recent fossil shells occurred. These measurements gave him all he sought, a notion of the amount of elevation in the recent period. On compar

ing his measurements with those of the Beagle survey, he was struck with their uniformity. He tabulated all the measurements representing the summit edges of the plains; and the tabulation proved to him that the elevation of the land had gone on at a remarkably equable rate over a north and south distance of at least five hundred miles. There is comparatively little danger of throwing away effort in a well directed classification. The danger lies in not comprehending the vast significance of the process both in actual investigation and in the presentation of results, and in the lack of persistent determination to exhaust its resources. Darwin himself sometimes owed it to happy accident that he did not overlook this powerful instrument.

Guided by deduction to the probable relation of the distribution of volcanoes to that of coral islands, and to the distribution of fringing and barrier reefs and atolls in relation to each other, he spent months in mapping them from the descriptions of voyagers, surveying vessels, etc. From the classification of a vast chaotic mass of facts scattered throughout geographic and geological literature, he extracted some of the most important conclusions of his whole work 1 Geological Observations, etc., p. 211.

on coral islands.1 Time has shown that the conclusions reached from this mapping of facts is too general. The conclusions that volcanoes are invariably absent from the areas which have recently subsided or are still subsiding, and are commonly present in areas that are rising or have recently risen, that fringing reefs lie in the areas of elevation and atolls in the areas of subsidence, may not be accepted without important reservations. But whether time shall ultimately substantiate or correct Darwin's conclusions, or shall even destroy some of them, his classifications will always remain essential to the study of coral islands.

1 Structure and Distribution of Coral Islands, p. 189.