ments and of the observed results. In a question of the permanence or mutability of forms, we may safely reason from analogy after a few cases have been fully proved. A much greater effect would, therefore, have been produced by a few detailed numerical statements, than by hundreds of vaguely reported observations. Though we are assured in the introductory part of the work that it contains nothing hypothetical, but is based on facts that cannot be disputed, when we come to investigate details we do not, in any single case, get satisfactory information regarding the conditions under which the observations were made. The results are in many cases opposed to our ordinary experience, and to the observations of other careful observers. For this reason the conditions under which they were obtained should have been precisely stated, so that any one might have had it in his power to repeat them. The great number of plants simultaneously observed is very perplexing, and makes the results still more doubtful.

In criticising some results published by M. Caspary, on the forms of some species of Biscutella, M. Jordan lays great stress on the importance of taking all the precautions necessary to avoid the possibility of error. We may, therefore, infer that he is in the habit of taking every precaution which he thinks needful, but it would be much more satisfactory to have some details, so that we might judge for ourselves of their sufficiency. The results of M. Caspary's experiment are so little in accord with M. Jordan's results in similar cases that he altogether refuses to accept them. He quotes from the "Flora," but we find the original account of these experiments in the 4th volume of Walper's "Annales." M. Caspary collected the seeds of six different forms of Biscutella in the autum of 1853 with great care. Sown in the spring of 1854, in the Berlin Botanic Garden, each set of seeds produced a glabrous, or almost glabrous, and a hairy fruited form. The smooth and hairy sorts are therefore not distinct species, but are reduced by M. Caspary to three. M. Jordan, on the other hand, has continued year after year to raise from seed the smooth and rough-fruited forms, and has found them to retain their character unaltered, as well as all the other distinctive characters peculiar to each form. Looking at the matter without bias, we may accept the results of both observers, the experiments of both, so far as we can see, being equally trustworthy; but instead of saying, with M. Jordan, that Caspary's result is insufficient to serve as a basis for conclusions, it seems to us that it goes far to negative any conclusions that can be founded on M. Jordan's.

The descriptive part of M. Jordan's work consists of 330 octavo pages, and contains diagnoses of 352 species, all but two (poppies), belonging to the families Ranunculaceae and Cruciferae. These 352 species are dismemberments of 58 Linnean types, so that on an average there are as nearly as possible six Jordanian species to each Linnean species. The extreme numbers are widely apart. Draba verna is divided into 53 species, but in a great many cases the Linnean species is divided into two only. The flora of France contains at least four times as many species of Ranunculaceae and Cruciferae, equal in value to the 58 here dismembered. We are not, however, to understand that the others are monotypic species. On the contrary, M. Jordan gives us clearly to understand that when all the species are investigated by competent observers, every one of them will be found to be composite, their apparent unity being the consequence of the superficial manner in which they have been looked at by botanists.

We cannot select a better illustration of the principles which guide our author in his subdivisions than Draba (or Erophila) verna, L. The genus Erophila is said to contain about five species, but in Europe it has hitherto been considered monotypic. The Asiatic species are little known, and as they are similar in habit and aspect to our common species, they will perhaps be found to be connected by a gradual series of intermediates with E. verna. This common little plant M. Jordan divides into 53 species, all but one (Corsican) natives of the south of France. He recommends the careful study of these forms to all observers, especially to those who are unwilling to adopt his views in their entirety, with the most complete conviction that no one can pay proper attention to them without at once becoming a convert. A little plant abundant every where, easily cultivated from seed, taking up little space and growing readily on any waste place, certainly offers every facility for observation. It would be too much to expect it to vary in England exactly in the same manner and to the same extent as in France; but we need not doubt that analogous variations will be met with, so that every one has an opportunity of judging for himself of the value of the species thus instituted. According to M. Jordan, they should be observed in autumn, their distinctive characters being most evident in the young leaves before they are injured by the winter frosts. The leaves which accompany the flowering scapes are not characteristic either in shape or colour. The points which it is most important to observe are the simple or forked hairs (on which character M. Jordan

divides the genus Erophila into two sections), the amount of pubescence, the shape of the pod, the shape, size and colour of the leaves, the size and colour of the flowers, and the general habit. The number of species of Erophila in France, he believes to be at least twice as many as those to which he has already given names, and if we consider the wide range of E. verna from temperate North America to the Himalayas, and its extreme abundance throughout the whole of its range, in the old world at least, we may fairly infer that with proper opportunities of observation the number of species would amount to at least two hundred. It is interesting to note the fact that these supposed species are all local forms. Not more than three or four of the fifty-three are usually found growing intermixed, and very frequently millions of individuals of a single form are found growing together without intermixture of a single specimen of any of the others. Each year too the same forms re-appear in the same places absolutely unchanged in character.

What then do we learn from M. Jordan's observations? We think that they may be considered to establish the fact that a practised observer can divide the assemblages of individuals commonly considered as species (using that word in the widest of the senses in which it is employed), into a greater or less number of forms to which definite characters can be given. These forms can, under favourable circumstances, be recognised with facility in a growing state, less certainly and often doubtfully in a dried state. In some cases, at least, they transmit their characters unaltered to their offspring during a certain number of years. The existence of these differences does not imply their permanence, nor does the transmission of slight characters by descent during a few generations prove that such characters will never vary. Whether, with M. Jordan, we call these distinct forms species, or with most other naturalists, races, depends on theoretical considerations which every naturalist must work out for himself, and on which, as is well known, there are at the present time the greatest possible differences of opinion.

Though M. Jordan is, as we have seen, fully convinced that these slightly distinct forms are permanent, and are therefore rightly regarded as species and admitted into our Floras, he does not, on that account, the less recognise the existence of the old species of which they are fractional parts, as distinct entities, to use a

N.H.R.-1865.

R

word introduced into our science by Bentham. Indeed, he not only sees no objection to their recognition as such, but considers it advantageous to unite all the allied forms into groups, corresponding to and bearing the name of the original type. In the book before us, he invariably notices these groups or types, and gives the number of species into which each is separable according to his observations. The Linnean type, from his point. of view, loses its rank as a species, but remains as an assemblage of true species all closely allied one to the other, and constituting a natural sub-division of the genus. No true species, in the sense in which M. Jordan understands that word, is distinguished by a single conspicuous character. There is, for him, no such thing as a well marked species (éspèce tranchée). Such characters distinguish groups of species. It is only in a certain number of cases indeed that the true species have hitherto been accurately distinguished from each other, but as in every case in which a Linnean type has been studied with due care it has been analysed into several, it is perfectly legitimate to conclude that it will be so in every instance.

We follow M. Jordan in his use of the expression Linnean type, more for the sake of convenience than as intending to express our belief that the Linnean species are in every case true to nature. Every working botanist is well aware that, in the great majority of cases, species as described in books are and must be empirical. Their value varies in the case of each describer, with the nature of his definition of a species, and perhaps in many cases according to his estimate of the importance, or possibility, of accuracy of diagnosis, and the consequent greater or less pains bestowed on them. In the case of exotic plants, the descriptive botanist has rarely an opportunity of examining more than a few specimens, so that their specific distinctness must be an inference from his knowledge of the relative constancy of characters derived from the various parts of the plants. This knowledge is acquired, in the first instance, by the teaching of others, and by the study of their works, and afterwards by the observation of the small number of species which it is possible to study in detail. When Linnæus had a sufficiency of materials, as in the case of the plants of his own country, and of other parts of Europe visited by him, he seems to have applied his great powers of generalization as successfully in this as in other things, and to have brought together, under a common name, or as

varieties of one species, forms which the less experienced authors who preceded him had kept asunder. Now and then, no doubt, he united very distinct species; but, on the whole, we may with M. Jordan regard the Linnean species as definite groups, and not passing by gradual transitions one into the other.

The progress of botanical science since it received its first great impulse from the labours of Linneus and Jussieu has been so rapid and continuous that it has, till quite lately, been impossible to stand still to review the question of species. In all branches of natural history there has been a gradually increasing tendency to their multiplication from the greater precision which has of late years been introduced into the modes of observation. This is most the case in local Floras, which, as a rule, are the work of botanists who confine their attention to the areas regarding which they publish, or at the most extend their range only to neighbouring countries, which possess a similar Flora. There are, of course, exceptions, and it will almost always be found that the more general the range of a botanist's studies the more wide his opinion of the specific limit. That the increase in the number of species in each Flora has been continuous, and at the same time extremely gradual, will be evident to any one who examines and compares the local Floras of the present day with those of older date. Each generation of botanists avails itself of the observations of its predecessors as a foundation from which to advance further in the same direction. Believers in the permanence of species will require no proof that the forms of Rubus, or Rosa, or Ranunculus aquatilis, which are now recognised, were much the same a century ago as they are now, and the advocates of the mutability of species, knowing the great lapse of time required to produce even a small modification, will readily admit the fact. Little or no change has taken place since that time in the external circumstances under which these plants grow, and no modifying causes can be suggested which are likely to have produced a sudden tendency to variation in these genera. If the possibility of this in a single case be contended for, it will, we suppose, be certainly conceded that it cannot have happened in all the variable genera. And yet while there has been during the period mentioned a gradual increase in the number of forms regarded as of specific value, there is nothing like a common consent among botanists of the same period as to the particular forms which are to be considered as species or varieties. No two authors, nay no two editions, agree with one another. Again,