All the stones of a Chinese arch are commonly wedge-shaped, their sides forming radii which converge towards the centre of the curve. It is observable, that, according to the opinion of Captain Parish who surveyed and made plans of a portion of the Great Wall, no masonry could be superior to it. The arched and vaulted work was considered by him as exceedingly well turned. The Chinese, therefore, must have understood the construction and properties of the arch long before the Greeks and Romans, whose original and most ancient edifices consisted of columns, connected by straight architraves, of bulk sufficient to support the incumbent pressure of solid masonry.

CHAPTER XIX.

NATURAL HISTORY AND PRODUCTIONS.

Chinese Classification-Result of their peculiar Language-European Researches in China-Zoology- -Mammalia-Birds-Reptiles-FishesInsects-Botany-Tea-plant-Timber Trees-Uses of the BambooDwarf Trees-Fruits -Flowers-Geological features-Chalky strata nearly unknown-Abundance of Coal-Unstratified Rocks and older strata-No active Volcanoes-Minerals and Metals.

AFTER a curious analysis of the great Chinese work on materia medica, which, although its name Puntsaou might literally imply that it was merely a herbal or history of plants, is in fact a classification of the chief productions of nature in the animal, vegetable, and mineral kingdoms, M. Remusat comes to the following conclusion: "I think we may infer that natural history has engaged the attention of the Chinese from the remotest antiquity, and that it became in consequence an object of pursuit among neighbouring nations, *which caused it to make some progress. The mode of writing employed in those countries, leading the people who used it to establish genera and orders, furnished them with the elements of an excellent nomenclature, and put them in the way of classification. All that could be learned from mere superficial inspection they have observed and recorded: all that demanded reflection or delicate research they have remained ignorant of, or misapprehended. Superficial, however, as are the ideas they have collected, they constitute a scientific whole, Using the same written characters, as Japan and Cochin

China.

which derives some value from the method to which it has been subjected. We conclude with a remark which is not destitute of interest to science itself: it is, that the Chinese and Japanese descriptions, when accompanied by the figures they refer to, may, with all their imperfections, enable us to distinguish the species we do from those we do not possess, augment our knowledge of facts, diffuse some light upon the distributions of the natural objects of the ancient world, and consequently may be consulted with advantage even by naturalists, so long as circumstances shall continue to interdict European philosophers from countries so abundant in objects of natural history, and hitherto so little explored." In the xvth chapter * allusion was made to the advantageous hints which the constitution of their written character had, from the earliest ages, afforded to the Chinese for a systematic nomenclature, and a rational classification of natural objects into certain genera or families, according to the most striking and obvious analogies that existed among them. The two hundred and fourteen roots, under which the whole language is arranged in Chinese dictionaries, include about one hundred and sixty, which serve at once (with the aid of other characters) as component parts in the written designations of all known objects in the animal, vegetable, and mineral kingdoms, and as heads under which they have been classed. "From this simple arrangement," observes M. Rémusat, "the very ideas appear which regulated the formation of the compound signs; which ideas frequently coincide with such as intelligent naturalists might acknowledge and adopt as a basis for their arrangements. This may be observed on a glance at even their modern dictionaries, although the written language of China has undergone alterations of all kinds, and admitted many * Page 140.

irregularities, which have affected the nomenclature of natural objects as well as other parts of the language. In turning over the leaves of the commonest of these works, we easily recognise genuine natural families, imperfect, no doubt, and founded upon inaccurate views, imperfect observation, and an unphilosophical analysis, but discovering almost always a judicious design, with sound and sometimes ingenious conclusions."

Of the thirty roots, or radical characters, which constitute the genera or families above alluded to, fourteen include the animal kingdom. The mammalia are comprised under nine of these, viz. three families of carnivora, one of rodentia, and three of ruminantia, as oxen, sheep, deer; while the horse and swine are the types of two other families. In the details of the above arrangement there has been (as might be expected) much confusion and a want of discrimination, in classing together animals between whom there was no real analogy, as well as separating others that were nearly allied: the ape and monkey tribes, for instance, are classed with the dogs; and numerous other examples might be adduced of the same kind. Birds, one of the most numerous class of animals in China, are all comprised in one family. Then come the tortoise and frog tribes under two heads. Fishes constitute one family, and improperly include the cetaceous and saurian tribes, as well as lobsters, crabs, and some of the molluscæ. The fourteenth family of animals, in the Chinese dictionary system, consists of insects.

This may serve to convey some idea of the notions which the Chinese have of classification, and show at the same time in what they have failed. Their vegetable kingdom is divided into eleven principal families. The first comprises all herbaceous plants, which have a common type, and are very numerous:

VOL. II.

T

the second family has wood for its radical character, and includes all trees, as well as plants with a woody stem: the bamboo, on account of its importance in use, and the great number of its varieties, stands at the head of the fifth class, and includes under it all reedy plants. No less than four separate radical characters serve as the heads under which the corn plants and esculent grasses have been arranged, and it follows of course that many repetitions and superfluous distinctions have taken place. The four together should have formed one natural family. The eighth family consists of leguminous plants, and has the bean for its type: the ninth comprises the cucurbitaceous, or gourd tribes: and under the tenth are included only about a dozen species and varieties of the alliaceous plants, as garlic onions, and leeks. The importance attached to some of the smaller divisions no doubt arose from their having been principal articles of food from the first. The eleventh and last family consists of plants analogous to the hemp, which, from its consequence, has from the earliest times been designated by a simple and radical character.

The mineral kingdom has been classed by the Chinese lexicographers under five radical characters. The first family consists of gems, of which the famous yu, or jade, is the type: to these have been improperly added all factitious stones, with glass, amber, &c. The four remaining families are distinguished into stones, earths, salts, and metals. "It must be remembered," observes M. Rémusat very correctly of the system, "that this was not a methodical or systematic arrangement contrived by naturalists, in order to classify the objects they wished to describe; but a mere distribution of written signs, brought together according to their orthography, and classed by the makers of dictionaries, solely with a