Page:Mind (New Series) Volume 8.djvu/548

534 CRITICAL NOTICES resemblance. Some such recurrence to an obsolete arrangement is imputed to Haeckel by the author in a passage of detailed criticism. 1 But the series is not linear. The organic form, which, in degree of evolution or organisation, might be the common ancestor of all living beings, coexists on the same genealogical level with the " higher " forms of life, as their remote cousin. If the classifica- tion is to be genealogical, such a form must appear in it at two extremes, as common ancestor, and as a collateral " uniplane " with existing forms. Admitting thus, then, the facts of collateral kinship in their fullest extent, can we make a genealogical arrangement coincide with a generic classification ? M. Durand does not deny the pos- sibility in toto, but points out a number of difficulties. i. The Ancestor's name, if it is to stand for a class, must imply an essential transmissible character; thus "Vegetable Monera " could not be a general name for all vegetable organisms. ii. It is conceivable that the " essential " quality may fail an Invertebrate species might spring from a Vertebrate one. The ancestor of the Eeptiles could hardly be called by a name which would serve as a general name to include Birds. Eegression, Convergence and the like, enhance this difficulty. It may be noted as to this that in any case the common ancestor in such a system must have his descendants' character read back into the meaning of his name, i.e., must have a sense put upon his name which his given qualities do not bear out. A true view of con- notation shows us this at once. To make the actual qualities of the animal Monera (calling it " Animal " par excellence) the uni versal of all the qualities of Animals, would be just like under- standing the connotation of " Animal "to be the remainder, after all attributes in which animal species differ, such, for example, as intelligence, had been omitted. iii. The author asks whether the generic name in a natural classification does always imply a single mother species ; adducing such cases as the common character of the Greek dialects or the Slav languages. From the mere common character, he urges, we cannot tell whether we may infer an actual " absolute " Greek, or an actual individual " Slav " language, as respective common ancestor. iv. By a sort of converse supposition toi., if the meaning of the actual common ancestor's name is confined to his given qualities, and nothing read back into them, then he will be very close to his 1 Haeckel, I gather from M. Durand, shows a genealogical tree with its trunk divided into four vertical segments, each named after a general division of the animal world. He also gives a list of kinds in linear order. A list must be linear, and should be interpreted by the genea- logical tree (see e.g. Lankester, art. Zoology in Enc. Brit.) ; but it is a question if the tree trunk can represent actual animal species (see below).