his own subject, no one has yet approached. His essay on life may well be likened to those broken fragments of ancient art, which, imperfect as they are, still bear the impress of the inspiration which gave them birth, and present in each separate part that unity of conception which to us makes them a complete and living whole.

From the preceding summary of the progress of physical knowledge, the reader may form some idea of the ability of those eminent men who arose in France during the latter half of the eighteenth century. To complete the picture, it is only necessary to examine what was done in the two remaining branches of natural history, namely, botany and mineralogy, in both of which the first great steps towards raising each study to a science were taken by Frenchmen a few years before the Revolution.

In botany, although our knowledge of particular facts has, during the last hundred years, rapidly increased,169 we are only possessed of two generalizations wide enough to be called laws of nature. The first generalization concerns the structure of plants; the other concerns their physiology. That concerning their physiology is the beautiful morphological law, according to which the different appearance of the various organs arises from arrested development: the stamens, pistils, corolla, calyx, and bracts, being simply modifications or successive stages of the leaf. This is one of many valuable discoveries we owe to Germany; it being made by Göthe late in the eighteenth century. With its importance every botanist is familiar; while to the historian of the human mind it is peculiarly interesting, as strengthening that great doctrine of development, towards which the highest branches of knowledge are now hastening, and which, in

170

169 Dioscorides and Galen knew from 450 to 600 plants. Winckler, Geschichte der Botanik, 1854, pp. 34, 40; but, according to Cuvier (Eloges, vol. iii. p. 468), Linnæus, in 1778, en indiquait environ huit mille espèces ;" and Meyen (Geog, of Plants, p. 4) says, 66 'at the time of Linnæus's death, about 8000 species were known." (Dr. Whewell, in his Bridgewater Treatise, p. 247, says "about 10,000.") Since then the progress has been uninterrupted; and in Henslow's Botany, 1837, p. 136, we are told that "the number of species already known and classified in works of botany amounts to about 60,000." Ten years later, Dr. Lindley (Vegetable Kingdom, 1847, p. 800) states them at 92,930; and two years afterwards, Mr. Balfour says "about 100,000." Balfour's Botany, 1849, p. 560. Such is the rate at which our knowledge of nature is advancing. To complete this historical note, I ought to have mentioned, that in 1812, Dr. Thomson says "nearly 30,000 species of plants have been examined and described." Thomson's Hist. of the Royal Society, p. 21.

170 It was published in 1790. Winckler, Gesch. der Botanik, p. 389. But the historians of botany have overlooked a short passage in Göthe's works, which proves that he had glimpses of the discovery in or before 1786. See Italianische Reise, in Göthe's Werke, vol. ii. part ii. p. 286, Stuttgart, 1837, where he writes from Padua, in September, 1786, "Hier in dieser neu mir entgegen tretenden Mannigfaltigkeit wird jener Gedanke immer lebendiger: dass man sich alle Pflanzengestalten viel leicht aus Einer entwickeln könne." There are some interesting remarks on this brilliant generalization in Owen's Parthenogenesis, 1849, pp. 53 seq.

the present century, has been also carried into one of the most difficult departments of animal physiology.""

But the most comprehensive truth with which we are acquainted respecting plants, is that which includes the whole of their general structure; and this we learnt from those great Frenchmen who, in the latter half of the eighteenth century, began to study the external world. The first steps were taken directly after the middle of the century, by Adanson, Duhamel de Monceau, and, above all, Desfontaines; three eminent thinkers, who proved the practicability of a natural method hitherto unknown, and of which even Ray himself had only a faint perception.172 This, by weakening the influence of the artificial system of Linnæus,173 prepared the way for an innovation more complete than has been effected in any other branch of knowledge. In the very year in which the Revolution occurred, Jussieu put forward a series of botanical generalizations, of which the most important are all intimately connected, and still remain the highest this department of inquiry has reached.17 Among these,

"That is, into the study of animal monstrosities, which, however capricious they may appear, are now understood to be the necessary result of preceding events. Within the last thirty years several of the laws of these unnatural births, as they used to be called, have been discovered; and it has been proved that, so far from being unnatural, they are strictly natural. A fresh science has thus been created, under the name of Teratology, which is destroying the old lusus naturæ in one of its last and favourite strongholds.

172 Dr. Lindley (Third Report of Brit. Assoc. p. 33) says, that Desfontaines was the first who demonstrated the opposite modes of increase in dycotyledonous and monocotyledonous stems. See also Richard, Eléments de Botanique, p. 131; and Cuvier, Eloges, vol. i. p. 64. In regard to the steps taken by Adanson and De Monceau, see Winckler, Gesch. der Botanik, pp. 204, 205; Thomson's Chemistry of Veg etables, p. 951; Lindley's Introduc. to Botany, vol. ii. p. 132.

It is curious to observe how even good botanists clung to the Linnæan system long after the superiority of a natural system was proved. This is the more noticeable, because Linnæus, who was a man of undoubted genius, and who possessed extraordinary powers of combination, always allowed that his own system was merely provisional, and that the great object to be attained was a classification according to natural families. See Winckler, Geschichte der Botanik, p. 202; Richard, Eléments de Botanique, p. 570. Indeed, what could be thought of the permanent value of a scheme which put together the reed and the barberry, because they were both hexandria; and forced sorel to associate with saffron, because both were trigynia? Jussieu's Botany, 1849, p. 524.

174 The Genera Plantarum of Antoine Jussieu was printed at Paris in 1789; and, though it is known to have been the result of many years of continued labour, some writers have asserted that the ideas in it were borrowed from his uncle, Bernard Jussieu. But assertions of this kind rarely deserve attention; and as Bernard did not choose to publish any thing of his own, his reputation ought to suffer for his uncommunicativeness. Compare Winckler, Gesch. der Botanik, pp. 261-272, with Biog. Univ. vol. xxii. pp. 162-166. I will only add the following remarks from a work of authority, Richard, Eléments de Botanique, Paris, 1846, p. 572: "Mais ce ne fut qu'en 1789 que l'on eut véritablement un ouvrage complet sur la méthode des familles naturelles. Le Genera Plantarum d'A. L. de Jussieu présenta la science des végétaux sous un point de vue si nouveau, par la précision et l'élégance qui y règnent, par la profondeur et la justesse des principes généraux qui y sont exposés pour la première fois, que c'est depuis cette époque seulement que la méthode des

I need only mention the three vast propositions which are now admitted to form the basis of vegetable anatomy. The first is, that the vegetable kingdom, in its whole extent, is composed of plants either with one cotyledon, or with two cotyledons, or else with no cotyledon at all. The second proposition is, that this classification, so far from being artificial, is strictly natural; since it is a law of nature, that plants having one cotyledon are endogenous, and grow by additions made to the centre of their stems, while, on the other hand, plants having two cotyledons are exogenous, and are compelled to grow by additions made, not to the centre of their stems, but to the circumference.15 The third proposition is, that when plants grow at their centre, the arrangement of the fruit and leaves is threefold; when, however, they grow at the circumference, it is nearly always fivefold.176

This is what was effected by the Frenchmen of the eighteenth century for the vegetable kingdom:177 and if we now turn to the mineral kingdom, we shall find that our obligations to them are equally great. The study of minerals is the most imperfect of the three branches of natural history, because, notwithstanding its apparent simplicity, and the immense number of experiments which have been made, the true method of investigation has not yet been ascertained; it being doubtful whether mineralogy ought to be subordinated to the laws of chemistry, or to those of crystallography, or whether both sets of laws will have to be confamilles naturelles a été véritablement créée, et que date la nouvelle ère de la science des végétaux. . . . L'auteur du Genera Plantarum posa le premier les bases de la science, en faisant voir quelle était l'importance relative des différents organes entre eux, et par conséquent leur valeur dans la classification. . . . Il a fait, selon la remarque de Cuvier, la même révolution dans les sciences d'observation que la chimie de Lavoisier dans les sciences d'expérience. En effet, il a non seulement changé la face de la botanique; mais son influence s'est également exercée sur les autres branches de l'histoire naturelle, et y a introduit cet esprit de recherches, de com paraison, et cette méthode philosophique et naturelle, vers le perfectionnement de laquelle tendent désormais les efforts de tous les naturalistes."

Hence the removal of a great source of error; since it is now understood that in dicotyledons alone can age be known with certainty. Henslow's Botany, p. 243: compare Richard, Eléments de Botanique, p. 159, aphorisme xxiv. On the stems of endogenous plants. which, being mostly tropical, have been less studied than the exogenous, see Lindley's Botany, vol. i. pp. 221-236; where there is also an account, pp. 229 seq., of the views which Schleiden advanced on this subject in 1839.

176 On the arrangement of the leaves, now called phyllotaxis, see Balfour's Botany, p. 92; Burdach's Physiologie, vol. v. p. 518.

The classification by cotyledons has been so successful, that, "with very few exceptions, however, nearly all plants may be referred by any botanist, at a single glance, and with unerring certainty, to their proper class; and a mere fragment even of the stem, leaf, or some other part, is often quite sufficient to enable him to decide this question." Henslow's Botany, p. 30. In regard to some difficulties still remaining in the way of the threefold cotyledonous division of the whole vegetable world, see Lindley's Botany, vol. ii. p. 61 seq.

sidered. 17 At all events it is certain that, down to the present time, chemistry has shown itself unable to reduce mineralogical phenomena; nor has any chemist, possessing sufficient powers of generalization, attempted the task except Berzelius; and most of his conclusions were overthrown by the splendid discovery of isomorphism, for which, as is well known, we are indebted to Mitscherlich, one of the many great thinkers Germany has produced.179

Although the chemical department of mineralogy is in an unformed and indeed anarchical condition, its other department, namely, crystallography, has made great progress; and here again the earliest steps were taken by two Frenchmen, who lived in the latter half of the eighteenth century. About 1760, Romé De Lisle so set the first example of studying crystals, according to a scheme so large as to include all the varieties of their primary forms, and to account for their irregularities, and the apparent caprice with which they were arranged. In this investigation he was guided by the fundamental assumption, that what is called an irregularity, is in truth perfectly regular, and that the operations of nature are invariable. Scarcely had this great

178 Mr. Swainson (Study of Natural History, p. 356) says, "mineralogy, indeed, which forms but a part of chemistry." This is deciding the question very rapidly; but in the mean time, what becomes of the geometrical laws of minerals? and what are we to do with that relation between their structure and optical phenomena, which Sir David Brewster has worked out with signal ability?

17 The difficulties introduced into the study of minerals by the discovery of isomorphism and polymorphism, are no doubt considerable; but M. Beudant (Minéralogie, Paris, 1841, p. 37) seems to me to exaggerate their effect upon "l'importance des formes crystallines." They are much more damaging to the purely chemical arrangement, because our implements for measuring the minute angles of crystals are still very imperfect, and the goniometer may fail in detecting differences which really exist; and, therefore, many alleged cases of isomorphism, are probably not so in reality. Wollaston's reflecting goniometer has been long considered the best instrument possessed by crystallographers; but I learn from Liebig and Kopp's Reports, vol. i. pp. 19, 20, that Frankenheim has recently invented one for measuring the angles of "microscopic crystals." On the amount of error in the measurement of angles, see Phillips's Mineralogy, 1837, p. viii.

1 He says, "depuis plus de vingt ans que je m'occupe de cet objet." Romé de Lisle, Cristallographie, ou Description des Formes propres à tous les Corps du Règne Minéral, Paris, 1783, vol. i. p. 91.

181 See his Essai de Cristallographie, Paris, 1772, p. x.: "un de ceux qui m'a le plus frappé ce sont les formes régulières et constantes que prennent naturellement certains corps que nous désignons par le nom de cristaux." In the same work, p. 13, "il faut nécessairement supposer que les molécules intégrantes des corps ont chacune, suivant qui lui est propre, une figure constante et déterminée." In his later treatise (Cristallographie, 1783, vol. i. p. 70), after giving some instances of the extraordinary complications presented by minerals, he adds: "il n'est donc pas étonnant que d'habiles chimistes n'aient rien vu de constant ni de déterminé dans les formes cristallines, tandis qu'il n'en est aucune qu'on ne puisse, avec un pen d'attention rapporter à la figure élémentaire et primordiale dont elle dérive." Even Buffon, notwithstanding his fine perception of law, had just declared, "qu'en général la forme de cristallisation n'est pas un caractère constant, mais plus équivoque et plus variable qu'aucun autre des caractères par lesquels on doit distinguer les

idea been applied to the almost innumerable forms into which minerals crystallize, when it was followed up with still larger resources by Haüy, another eminent Frenchman. 182 This remarkable man achieved a complete union between mineralogy and geometry; and, bringing the laws of space to bear on the molecular arrangements of matter, he was able to penetrate into the intimate structure of crystals.183 By this means, he succeeded in proving that the secondary forms of all crystals are derived from their primary forms by a regular process of decrement;15 and that, when a substance is passing from a liquid to a solid state, its particles are compelled to cohere, according to a scheme which provides for every possible change, since it includes even those subsequent layers which alter the ordinary type of the crystal, by disturbing its natural symmetry.185 To ascertain minéraux." De Lisle, vol. i. p. xviii. Compare, on this great achievement of De Lisle's, Herschel's Nat. Philos. p. 239: "he first ascertained the important fact of the constancy of the angles at which their faces meet."

162 The first work of Haüy appeared in 1784 (Quérard, France Littéraire, vol. iv. p. 41); but he had read two special memoirs in 1781. Cuvier, Eloges, vol. iii. p. 138. The intellectual relation between his views and those of his predecessor must be obvious to every mineralogist; but Dr. Whewell, who has noticed this judiciously enough, adds (Hist. of the Induc. Sciences, vol. iii. pp. 229, 230): "Unfortunately Romé de Lisle and Hauy were not only rivals, but in some measure enemies. . . . Hauy revenged himself by rarely mentioning Romé in his works, though it was manifest that his obligations to him were immense; and by recording his errors while he corrected them." The truth, however, is, that so far from rarely mentioning De Lisle, he mentions him incessantly; and I have counted upwards of three hundred instances in Haüy's great work, in which he is named, and his writings are referred to. On one occasion he says of De Lisle, "En un mot, sa cristallographie est le fruit d'un travail immense par son étendue, presque entièrement neuf par son objet, et très précieux par son utilité. Hauy, Traité de Minéralogie, Paris, 1801, vol. i. p. 17. Elsewhere he calls him, "cet habile naturaliste; ce savant célèbre,” vol. ii. p. 323; "ce célèbre naturaliste," vol. iii. p. 442: see also vol. iv. p. 51, &e. In a work of so much merit as Dr. Whewell's, it is important that these errors should be indicated, because we have no other book of value on the general history of the sciences; and many authors have deceived themselves and their readers, by implicitly adopting the statements of this able and industrious writer. I would par ticularly caution the student in regard to the physiological part of Dr. Whewell's History, where, for instance, the antagonism between the methods of Cuvier and Bichat is entirely lost sight of, and while whole pages are devoted to Cuvier, Bichat is disposed of in four lines.

183"Hauy est donc le seul véritable auteur de la science mathématique des cristaux." Cuvier, Progrès des Sciences, vol. i. p. 8; see also p. 317. Dr. Clarke, whose celebrated lectures on mineralogy excited much attention among his hearers, was indebted for some of his principal views to his conversations with Haüy: see Otter's Life of Clarke, vol. ii. p. 192.

184 See an admirable statement of the three forms of decrement, in Hauy, Traits de Minéralogie, vol. i. pp. 285, 286. Compare Whewell's Hist. of the Induc. Sciences, vol. iii. pp. 224, 225; who, however, does not mention Hauy's classification of “décroissemens sur les bords," "décroissemens sur les angles," and "décroissemens intermédiaires."

185 And, as he clearly saw, the proper method was to study the laws of symmetry, and then apply them deductively to minerals, instead of rising inductively from the aberrations actually presented by minerals. This is interesting to observe, because it is analogous to the method of the best pathologists, who seek the philos