'These assertions are purely gratuitous. At the present time such deviations are not producible even by means of art, and it is known by examination of the fossils of the later formations, as well as by the comparison of the most ancient examples with their living analogues, that the forms remain unalterable.'

So Agassiz, after that complete survey of Fossil Fishes, which has earned him the perpetual gratitude of geologists, says on the question of the diversity of species due to development: 'We must necessarily rise to a higher cause, and recognize more powerful influences, exercising on the whole of nature a more direct action, if one wishes not always to move in a vicious circle. For my part, I am convinced that species have been created successively at different epochs; and that the changes which they have suffered during a geological period are but of secondary importance, and depend only on their greater or less fecundity, and on migrations subordinated to the influences of the period.'

On the same side must be ranked the great authority of Cuvier,

clarum et venerabile nomen.

"There is no proof that all the differences which now distinguish organized beings are of such a nature as to have been produced by circumstances. All that has been advanced on this subject is hypothetical;

R. L.

0

experience appears to shew on the contrary, that, in the actual state of the globe, varieties are restrained within rather narrow limits, and as far back as we ascend into antiquity we see that these limits were the same then as now.

'We are therefore obliged to admit certain forms, which have been perpetuated since the origin of things without exceeding these limits; and all the beings which appertain to one of these forms constitute what is called a species. Varieties are the

accidental subdivisions of the species.

'Generation being the only means of knowing the limits to which varieties can extend, a species may be defined, as comprising the individuals, descended one from another or from common parents, and those which resemble them as much as they resemble each other.

"These forms do not produce or change themselves; life supposes their existence; it can only be kindled in organizations ready prepared for it; and the most profound meditations, as well as the most delicate observations, reach no further than to the mystery of the pre-existence of germs.'

PRIMITIVE TYPES.

The illustrious author of the Systema Naturæ permitted his accurate and richly stored mind some

times to wander into general contemplation of the origin of the large series of living beings which he had subjected to classification. Perhaps no man who ever lived was more entitled to pronounce an opinion on the affinities and sequences of the hosts of 'specific' forms which he had characterized. His Natural Orders of Plants are among the most real and consistent groups yet assembled; many of his divisions of plants and animals are the clearest and most satisfactory. What was the deliberate opinion of this gifted man? I take it from the 13th and best edition of his great work, 1767. It applies to plants only.

'Suppose that in the beginning the Almighty proceeded from the simple to the compound; from few to many; and thus from a primary vegetable principle, created so many different plants as there are natural orders. Then that HE mixed these orders of plants in generation, so that as many plants should arise as now are distinct genera. Next that nature mixed these generic plants, by ambigenous generations (which change not the structure of the flower), and multiplied them into species, as many as possible, excluding however hybrids, as sterile.'

Without too minutely criticising the distinction here drawn between the work of the Almighty, and the performance of nature, which apparently may

be held to mean independent permanency of genus, and dependent mutability of species-we remark in this passage the recognition of general 'primary types,' the 'struggle for existence' and the final 'limitation of species,' which have often been appealed to in later speculations.

DISTINCTION OF SPECIES.

Ever since the days of Linnæus the tendency of naturalists has been toward a nicer discrimination of species and a more thorough appreciation of the affinities which constitute natural assemblages. The former analytical process has been pushed very far in Botany, Zoology and Palæontology, so that what Linnæus and his immediate followers regarded as mere varieties have since been freely admitted to rank as 'good' species. This decision in regard to the majority of plants and animals has been quite independent of and without any reference to proof of the descent of the objects from a common parentage which is the original ground of specific identity, or from different parents which should constitute specific difference. In the case of a very large majority of the living invertebrata, no proof of this sort has been looked for; a definite peculiarity of structure, however slight, if often repeated,

so as to appear to be hereditary, has seldom been rejected as a specific character; even colour and magnitude are not despised in this analysis.

But in Palæontology the list of species is even less satisfactory. For here the question of descent is seldom to be insisted on at all; very often we have only the knowledge of one age of the fossils, sometimes only one specimen-occasionally only the surface of that. For the purpose of diminishing these sources of error, and of determining in regard to each species the two most important parts of its history, viz.

1. Its geographical province, the area which it

occupies, in one stratum or in several strata; 2. Its geological range in differents parts of this area;

nothing is so effectual as those monographs of particular well selected districts, which are founded on accurate observations of the localities of fossils, and of the conditions under which they are observed. But these observations lose all their value if the species of fossils be not precisely and certainly determined, and this amidst the conflict of classifications, and unequal estimates of specific agreements and differences, is a condition of success not always easy to be secured. If for example we wish to trace the geographical province or geological range of a