day, but that is no proof that it has maintained itself constantly distinct by hereditary descent during the whole period. The branched spike is a form which wheat has a tendency to assume under particular circumstances, and an approach to it may be often observed in some of our corn-fields during remarkably luxuriant harvests; but as it affords no advantage to farmers in our climate, no endeavours have been used to preserve it.

This leads us to the next objection, that the several species of Ægilops cultivated in botanic gardens maintain for years their distinctive characters without any tendency to change into wheat, and that cultivated grasses in general do not break out into varieties, especially in their seed. This is, however, no proof that they may not do so under other circumstances. It is stated that from 150 to 160 races of wheat have been cultivated in the garden of Paris for thirty years, and that Messieurs Vilmorin have for half a century kept up very extensive collections of cerealia, which have remained always true and distinct. Yet it is never pretended that these 150 or 160 sorts of wheat are all distinct botanical species. In botanic gardens the object is not to raise new varieties, but to keep all the sorts as true as possible, and if the gardener selects his seed it will always be from the most characteristic plants. De Candolle's whole chapter on the variations of species shows how readily hereditary races are formed, and grasses are not an exception to the general rule. To carry out the experiment of the Ægilops with any chance of success the usual process of gardeners in search of new varieties must be adopted,-to sow the plant on a large scale in different soils and in a climate known to be favourable, -to search carefully for any seedlings that may show the slightest tendency to vary in the required direction, and to sow for the next generation the seeds of these only. We have no doubt that, by patient perseverance in this course for a number of years, not only many so-called species of Egilops would be reduced to one type, but Mr. Fabre's experiments, resulting in the conversion of Egilops into wheat would again meet with the

same success.

These positive experiments are met by the supposition that the plant upon which M. Fabre worked was an accidental hybrid or cross between the common Ægilops and the wheat growing in the adjoining fields. Mr. Godron, a careful experimentalist, has, they say, produced the same results by artificially crossing the two; M. Fabre's plant grew under circumstances which do not exclude the possibility of its being a natural hybrid, and therefore it is argued that all proof of the specific identity of the two parents is gone.

But natural hybrids between two species are very rare, and amongst grasses hitherto unknown. The fecundation of ce'realia takes place in the bud before the stamina protrude and ' whilst the glumes are closely applied against each other, which 'circumstance excludes all hybrid fecundation.' 'Experience shows that different sorts of wheat grown together never 'cross.' Hybrids, moreover, are weak in constitution, and are seldom maintained beyond the second generation without artificial fecundation from one of the parent species, in which case a gradual assimilation to the latter has place. And supposing, as we contend, that wheat is but an extreme race of Egilops, the crossing these two races would have precisely the result obtained by Mr. Godron, the production of an intermediate race. Mr. Godron's supposition that the plants successively cultivated by M. Fabre were always fertilised by the pollen floating in the air from the neighbouring wheat-fields is perfectly gratuitous, unsupported by a single observed fact, and in direct contradiction to much that we know of the physiology of grasses.

One argument remains to be noticed. It is absurd to suppose, we are told, that men in their early rude state would have been tempted to cultivate for their food any grains which were not tolerably weighty and nutritious, such as wheat, 'barley, rye, or oats, pretty nearly as they are at the present 'day. Do we see barbarous nations attempt the cultivation of Egilops and so many other grasses moderately farinaceous ? 'Have men but little civilised ever the idea that long cultivation 'could ameliorate a species? And if they arrived at this idea 'was not their state of society an obstacle to the effective appli'cation of the principle? The more ancient we suppose agri'culture to have been, and the more distant its origin in times of 'ignorance, the more probable it is that cultivators would have 'chosen species offering from the outset incontestable advantages.' We certainly admit that man in an early stage of civilisation would not theoretically set about agricultural experiments, but we contend that, from the moment he began to cultivate at all, the gradual improvement of the plants he sowed would be the natural consequence. And as to his selection of grain for that purpose, we believe that of the wild grasses of Palestine, Greece, and Sicily, none have a larger and more nutritious seed than those very Ægilops, Hordeums, Sorghums, Panicums, &c., whose presumed descendants are now known under the names of wheat, barley, millet, &c. Civilised nations do not now cultivate Ægilops, because they possess the long cultivated varieties which are much better, and barbarous nations do not

live where Egilops grows. A wild grass, glyceria fluitans, which we should never think of cultivating so small is its seed, is nevertheless extensively harvested in north-eastern Europe, and its grain imported even into this country under the name of Manna Croup.

If any of our cultivated races of wheat should be found growing really wild, as it is pretended, in the ravines of Mesopotamia or the valleys of Mount Sipylus in Asia Minor, we might be induced to modify our opinions; but the facts adduced are far too vague and partial to be entitled to any weight; and from all that is now known we are led to conclude that our common wheats are the offspring of Egilops ovata taken into cultivation in the earliest stages of human civilisation. The arguments which have led to this result point out the course to be followed in the investigation of the wild types of other long-cultivated plants of obscure origin. The plant must be traced back to the region where it was first taken into cultivation, and we must there look out, amongst the wild plants of the district, for that species which is at once most nearly allied to the cultivated plant in question, and sufficiently abundant to have attracted the notice of the early inhabitants in their endeavour to procure food or raiment from the herb of the field.

The careful researches of De Candolle in regard to above 150 of the most generally cultivated species are an important contribution to this most interesting inquiry. And although he has not been able to connect above half that number with any certainty with their wild types, yet he has much restricted the limits within which the native country of each is to be sought for; and we cannot do better than close the subject in his words:

'We see that almost all cultivated species have a known origin, if not as to the exact country, at least as to the primary divisions of the globe to which the country belongs. It is therefore useless to suppose the disappearance of any region under the ocean since the invention of cultivation, still less to imagine a miraculous and special creation of plants for cultivation independently of ordinary species. The whole of the facts taken together prove that men merely cultivated those species which were within their reach, and which appeared to offer to them some advantage. In certain cases they carried them with them from one country to another, but frequently also they found them on their arrival in a new region. In particular, the absolute diversity in the cultivated species of the old and the new world shows how much these two continents were isolated, as well as their inhabitants, from a period which it is now impossible to fix.'

If the theory of the dispersion of vegetation over our globe, as aiding in our conjectures as to its primitive geography, is

mainly founded on the distribution of species, there is another consideration by which the conclusions arrived at may often be essentially confirmed or controlled. We allude to the fact that species which botanically resemble each other, have their origin very frequently in the same or nearly adjoining regions. Genera and families have, like species, their centres of creation and geographical distribution.

What is a genus? is a question which has led to as much controversy and difference of opinion as any other of the fundamental principles of botanical arrangement. Yet without a clear understanding of the rules to be followed in the delimitation of generic groups, the whole science is in danger of being again plunged into the chaos from whence it was extricated by Linnæus's happy invention of the substantive genus. Many naturalists of the present day consider it indeed to have no existence in nature, but to be a mere creature of the imagination, a kind of instrument to enable man to classify the infinity of forms exhibited by nature, so as to bring them regularly before him one after the other in masses reduced to the capacity of his comprehension. Others, with De Candolle, believe them to be real agglomerations of species more evident and more natural than species themselves. The truth probably lies between these two extremes. We think that all species may be arranged into groups indicated by nature, so as to divide and sub-divide the whole vegetable kingdom ad infinitum, but that the precise limits and extent of each group, call it class, order, family, tribe, genus, or section,-are purely arbitrary. But the considerations upon which these conclusions are founded would lead us far beyond our present limits. It is from want of space, not from any deficiency in the interest or importance of the subject, that we pass over the definition of generic groups, and refrain from any details connected with their geography. Climate and station, which restrict within certain limits the area of species, have little or no influence on that of genera. This depends much more on circumstances of original creation, and must be well studied in speculating on the early history of vegetation. When we see large American genera represented by a very few species in Eastern Asia, with one or two extending westwards even to Europe; when we find numerous Californian and Northwest Mexican genera represented by analogous species in Chili and Bolivia; when we trace great resemblance in generic forms, if not in species, between New Zealand, South Australia, the southern extremity of America, and the Antarctic regions generally; between the East Indian Peninsula and North tropical Africa; between Java, Ceylon, and South tropical

Africa, we cannot avoid speculations on the ancient continuity of lands similar to those we are led into by the identity of certain species. But this part of the science is in its infancy. The known facts upon which we must found our reasoning are as yet few and often vague and insufficiently authenticated. It is to be hoped that De Candolle's work may induce enlightened explorers like Hooker and Thomson to collect materials with the express object of assisting our researches; that future theorists in geographical botany will see the paramount importance of a correct investigation of their facts; and, above all, that a true knowledge of the area of a species must depend on an accurate appreciation of its botanical limits and variations. Systematic botany, which it has been the fashion of late years to hold in so much contempt, is nevertheless the groundwork upon which the correctness of the speculation of the physiologist and geographical botanist must mainly depend. But the botanist who devotes himself to it should always bear in mind that it consists not in the technical description of specimens, but in the due appreciation of species and affinities; that he who demonstrates a fact such as the specific identity of two plants hitherto believed to be distinct, or the affinities of an obscure vegetable, renders a far greater service to science than he who discovers, describes, or invents any number of supposed new species.

ART. VIII.-Perversion, or the Causes and Consequences of Infidelity: A Tale for the Times. 3 vols. London: 1856.

WE HEN a work of fiction depends in any degree for its interest on the adroit management and natural succession of supposed events, the imagination of the writer is checked and limited by the ordinary conditions of life, and his unfitness for the work he has undertaken is soon detected and exposed. It is thus that the greatest masters in the art have always shrunk from strange, though possible, incidents, and have given to their design the effect of a series of occurrences, each of which might pass before the observation of the reader without exciting either incredulity or amazement, but which, taken as a whole, produce the agreeable impression of interesting reality. Thus the example of the best authorities has established or confirmed the paradox that truth is stranger than fiction,' and it is left to those who are content with the approbation of the thoughtless and the uneducated to excite the fancy of the public by the invention or even the reproduction of astonishing or inconsistent circumstances.