450 MORPHOLOGY. 



sure ground, accurate knowledge of the unimpregnated germen, was 

 professed by no botanist. The want of a subdivision into classes, 

 orders, genera, and species, deduced from safe inductions, was con- 

 tinually more felt. Linnaeus had placed his forms of fruit side by side 

 as homologous members : the enlarged circuit of knowlege of materials 

 rendered that enumeration of forms insufficient ; and as new peculiar 

 phenomena occurred, new forms with new names were added, without 

 further trouble about the groundwork. This reproach especially applies 

 to that profound observer of individual things, Gartner, the very super- 

 ficial Willdenow, and Link, who always judges from solitary, accidental 

 ideas. In this, as so frequently in his casual notions, Link, of course, 

 has a perfectly correct idea, but usually he wants the scientific earnest- 

 ness to work it thoroughly out. He says, a very mistaken path has 

 been taken in making so many new words for solitary distinctions of 

 fruit, since individual different organs may indeed require especial words, 

 but not their modifications. Nevertheless he receives the whole of 

 the old nomenclature, which, in reference to the number of actual dif- 

 ferences, is partly made up of terms for very inessential modifications, 

 and he adds another new word to it. In the second edition of his Phi- 

 losophia Botanica (vol. ii. p. 253.), he says, Linnaeus, Gartner, and 

 Richard had given so many good descriptions of fruits, with their ter- 

 minology, that he would refrain from all new technical terms, and only 

 add amphispermium as a collective term for achcenium and caryopsis. 

 Nevertheless he made a wholly new definition for achcenium, called the 

 old caryopsis seminium, formed again two species of the new caryopsis, 

 according to characters actually non-existent, and called one carpelletum. 

 Besides these, he speaks only of capsula, pomum, legumen : nothing 

 is said of all the rest of the kinds of fruit ; and it is not stated at all 

 how the introduced expressions are to be applied to siliqua, drupa, 

 bacca, hesperidium, &c. L. C. Richard first (Analyse du Fruit, Paris, 

 1808), and subsequently DeCandolle (Organographie Vegetale, Paris, 

 1827), making use of the knowledge of the structure of the germen by that 

 time collected, sought, with somewhat more philosophic spirit, to give a 

 new basis to the matter. But they remained in the bonds of conven- 

 tionality, and so allowed a number of subordinate conditions to stand as 

 homologous members beside main divisions. L. C. Richard first dis- 

 tinguished the four layers, mentioned above, in the pericarp, namely, 

 the epicarpium and endocarpium, as external and internal epidermis, 

 and the mesocarpium as parenchyma between the two : of this latter, he 

 added, one layer is often separated, which forms the stone in the drupes, 

 &c. He therefore distinguished this layer strictly from the endocarpium, 

 because his distinction depended on minute observation. DeCandolle, 

 however, confused the whole matter again by introducing an imaginary 

 theory, tracing those three layers to the layers of a leaf, to which, from 

 imperfect knowledge of its structure, he ascribed three, and only three, 

 layers. Thus he made the endocarpium the third layer, counting in- 

 wards, and mixed it up with Richard's endocarpium, wholly overlooking 

 the woody layer of Richard's mesocarpium. Thus a pretended theory, 

 without observation, turned an excellent observation into a mass of 

 confusion. DeCandolle did the same with Richard's terminology for 

 the direction of the embryo, which he wholly misunderstood, and con- 

 sequently ascribed a superior radicle to the embryo of Ceratophyllum, 

 nearly a quarter of an inch long, truly easy enough to observe. De- 

 Candolle, indeed, started from the perfectly correct axiom, that the 



