brata, convince himself that these manuals, though small in size, are yet comprehensive, and that though they be running through successive editions at home, they still keep abreast of the tide of scientific progress, as it speeds along its thousand channels abroad.

LIV. DE CANDOLLE'S PRODROMUS.

PRODROMUS SYSTEMATIS NATURALIS REGNI VEGETABILIS, SIVE ENUMERATIO CONTRACTA ORDINUM GENERUM SPECIERUMQUE PLANTARUM HUCUSQUE COGNITARUM, JUXTA METHODI NATURALIS NORMAS DIGESTA, editore et pro parte auctore Alphonso De Candolle. Pars XV. Sectio 1. Parisiis, Maio, 1864.

THIS great work is now drawing towards a close, or rather, approaching that termination which circumstances have prescribed—far short unfortunately of the object which its original author hoped to attain. It has been nearly half a century in progress. Forty-six years have elapsed since the publication of the first volume of the "Regni Vegetabilis Systema Naturale," forty years since it was recommenced on a reduced plan. During this period we have witnessed the gradual progress and final triumph of those principles of method which it was originally intended to illustrate and promote. It may therefore be not uninteresting to pass in review the state of systematic botany, which first induced the elder De Candolle to undertake so formidable a task, and the various causes which have brought on a gradual modification of the original plan of his work, through twenty successive volumes or parts.

After the pacification of Europe in 1814 and 1815, scientific men, as much perhaps as any others, began to take advantage of that free intercommunication of ideas and observations, of which they had been almost deprived during the isolation of the previous quarter of a century. Botanists amongst others, in England, France, and Germany, now perceived that, in each country, materials had been separately accumulated towards the illustration of the natural history of the globe, and that although the publication of many of them was being separately carried on in each country, a new general systematic enumeration of all the plants known, in which all these separate collections should be brought together and compared was much wanted. The most recent works of the kind, Willdenow's elaborate "Species Plantarum," purporting to be a fourth or fifth edition

of the Species of Linnæus, and the admirable compact Synopsis published under the name of Persoon, had both become out of date; Vahl's excellent Enumeratio had been cut short by the author's death; and the question arose, who should undertake new editions of either of these, or a new work to replace them. In England there seemed little chance of carrying out any such project, without great pecuniary sacrifice. Although we could already boast of having produced the first botanist of the age, scientific botany was not popular. Very few of our cultivators or amateurs of the science went beyond the local flora or ornamental plants, and there was nothing to induce a publisher to risk a scientific work of which the cost would be so high and the probable sale so small.

On the Continent it was different. Previous publications bad given to France and Germany a footing in the market, the cost of printing and publishing was far below English prices, and, besides a much more generally diffused taste for general botany, there were a number of small university towns and other centres of science which insured a more general sale for works of the kind. Enterprising publishers in both countries were therefore ready to undertake a new systematic enumeration of all known plants, either on the model of Willdenow, or on that of Persoon, for which they found there was a demand. But in both countries it was debated how far the systems adopted in those works could be improved upon. The immense facilities in the pursuit of their science which botanists had derived from Linnæus's systematic nomenclature, from the admirable rules he had laid down and practically carried out, and from the fixity he had given to natural genera and species, had inspired an almost universal veneration for his system in every detail. This was more especially the case in Germany, where the feeling was aided by the general reactionary spirit of conservatism which then prevailed, and a keen sense of the miseries they had undergone, vaguely attributed to French innovation. Even in England the Linnean classes were almost universally adhered to. Robert Brown had indeed shown the impossibility of dealing satisfactorily with the mass of known plants, then already five or six times as large as in the days of Linnæus, without following Jussieu in extending to higher groups the principles upon which Linnæus had established genera. But the principal work in which he had taken up and carried forward the Jussieuan arrangement related to the plants of Australia, a far distant country, then scarcely known; it was moreover never completed, and the por

tion published was withdrawn from circulation; its influence, therefore, on the progress of science was for the time arrested. At the same time, the head of British botanists was possessor of the library and collections of Linnæus, whose infallibility he was therefore unwilling to see attacked, and no one had as yet ventured to introduce the natural orders to the notice of beginners. In Sweden, Russia, Italy, and Spain, they had scarcely been heard of.

In Germany, accordingly, botanists and publishers were agreed, not only that the Linnean arrangement should be followed in the new work, but that, like those of Willdenow and others, it should be given out as a new edition of Linnæus' Species Plantarum, and on this plan Roemer and Schultes began their bulky Systema. In the preface to their first volume, published early in 1817, they give three principal reasons for the method adopted: 1. Respect for the authority of Linnæus,-as if, when a man of extraordinary genius and ability has produced some great invention, and worked it out successfully, his method should ever after be strictly followed, without further improvement, even though foreshadowed by himself. 2. Because the Linnean system is the easiest, an argument still so frequently brought forward that we may have occasion to recur to it; we will now merely hint that the alphabetical arrangement of many earlier botanists is still easier. 3. Because the Linnean classes are more in conformity to nature, for like the laws of nature they never change, whatever be the additions made by new discoveries, whilst the socalled natural system is always changing; the arrangements of Jussieu, De Candolle, and Brown (the principal ones then known), differing from each other in the number, order, and limits of classes and orders-a syllogism which, from any other but a scholastic German, would be treated as a joke or conundrum. Roemer and Schultes were, however, generally accurate in detail, and bestowed much care on their work, which, notwithstanding the drawbacks as to method, would have been useful had it soon been completed. But the seven or eight volumes to which it ultimately extended can be treated only as fragmentary. No part, on such a system, can be complete in itself. A monograph of Pentandria, for instance, has but little more of unity and completeness than would have been a monograph of genera whose names begin with the letter C. The work is, therefore, now only referred to for such species as happen to be there first described, although as to several genera it was nearly contemporaneous with the corresponding portions of the Prodromus.

With the French, the case was very different. The reaction against useless innovation had not restored any blind respect for ancient institutions. They fully appreciated the Linnean nomenclature, genera and species, and even his classes were used in many local Floras; but a countryman of their own had successfully applied the generic principles to higher groups,-the Jussieuan method had become national. His families or Orders were taught in all their schools. De Candolle's admirable Flore Française had rendered them familiar to local botanists, beginners and amateurs, and the Linnean classes were being passed over as obsolete, or taught only as Keys to those foreign or older works in which they were still made use of.

In exotic botany, Kunth was exemplifying the advantages of the natural method in the great work he had commenced on the tropical American plants of Humboldt and Bonpland, and it was universally admitted that any new Species Plantarum, which must now be the work of many years, to be useful must be arranged in natural groups, so that each volume as it appeared should be complete in itself. It is true that the then most recent synopsis, that of Persoon, although French, was under the Linnean arrangement, but that was owing to a bookseller's stipulation, founded on a consideration of foreign sale, and said to have been so much against the wish of the principal author, L. C. Richard, that he refused to affix his name to it. All French botanists of note, therefore, called upon De Candolle to carry out the plan he was known to entertain of a comprehensive Species Plantarum arranged under natural Orders. He was eminently qualified for the purpose. His enlarged general views, his methodical mind, capable of embracing at once very complicated affinities, his talent for description and indefatigable industry, had been exemplified in his above-mentioned Flora, in his already numerous monographs and detached memoirs, and the principles upon which botanical works should be carried out had been admirably laid down in his "Théorie Élémentaire." He was still further stimulated to exertion by Roemer and Schultes' attempt to perpetuate the old system, and he spared no pains to render his work as complete as possible. Quietly established in his native town of Geneva, with an extensive herbarium and nearly complete botanical library, in constant and active correspondence with the principal Parisian botanists, and more especially assisted from the herbarium and library of his friend M. Benjamin Delessert, he also paid a visit to London to

inspect the collections there, a journey which he hoped to repeat occasionally as his work advanced. At last, in the summer of 1818, the appearance of the first volume of his "Regni Vegetabilis Systema Naturale" was hailed with great satisfaction by the botanical world, and even announced with some kind of public ceremony by his friends and admirers in Geneva.

In point of execution, this work answered all the expectations which had been entertained. The ordinal, tribual, generic and sectional groups were distributed and characterized with the author's wonted method and clearness; the relative requirements of contracted diagnoses and detailed descriptions well kept in view, and above all no labour was spared to draw up his characters, as far as possible, from actual observation, copying only where all his efforts to procure a sight of specimens had failed. But the plan adopted was far too extensive. In the synonymy especially, the attempt to refer to all works of any importance where a plant had been described and figured, to all general works in which it had been entered, including the old herbalists and ante-Linnean compilations, entailed an amount of labour and research, by no means compensated by any special utility in the result. Three years, therefore, elapsed before the second volume was published. In these two volumes were comprised complete monographs of ten natural Orders, including the complicated European ones of Ranunculaceæ and Cruciferæ, but the number of species they contained was but little above 1900, whereas the total number to be described was already estimated at above fifty thousand. Thus it was clear that, if the same plan were pursued, sixty volumes and three generations of authors would not suffice to bring the work to a conclusion. Unwilling, at first, to give up entirely the project he had so fondly entertained, but which he now saw indefinitely deferred, the author determined on preparing in the mean time a succinct enumeration of all plants known on a reduced scale, partly compiled from published works, but verified as far as possible from actual observation of specimens, and incorporating all the novelties he could collect, re-arranging the whole according to the method adopted. To this work he gave the name of a Prodromus, for it was intended as a preliminary sketch of the groups and species to be fully developed in the series of monographs which he still hoped might in some measure be carried out by a combination of authorship. He published a first volume at the close of the year 1823, or very early in 1824 (the date of the title page), comprising the whole of Thalamiflora,