CLASSIFICATION. 377 



through which classifications in general pass. In early 

 attempts to arrange organisms in some systematic manner, 

 we see at first a guidance by conspicuous and simple cha- 

 racters, and a tendency towards arrangement in linear order. 

 In successively later attempts, we see more regard paid to 

 combinations of characters which are essential but often in- 

 conspicuous, and an abandonment of a linear arrangement for 

 an arrangement in divergent groups and re-divergent sub- 

 groups. 



In the popular mind, plants are still classed under the 

 heads of Trees, Shrubs, and Herbs; and this serial classing 

 according to the single attribute of magnitude, swayed the 

 earliest observers. They would have thought it absurd to 

 call a bamboo, thirty feet high, a kind of grass; and would 

 have been incredulous if told that the Hart's-tongue should 

 be placed in the same great division with the Tree-ferns. 

 The zoological classifications current before Natural History 

 became a science, had divisions similarly superficial and sim- 

 ple. Beasts, Birds, Fishes, and Creeping-things are names 

 of groups marked off from one another by conspicuous differ- 

 ences of appearance and modes of life creatures that walk 

 and run, creatures that fly, creatures that live in the water, 

 creatures that crawl. And these groups were thought of in 

 the order of their importance. 



The first arrangements made by naturalists were based 

 either on single characters or on very simple combinations of 

 characters; as that of Clusius, and afterwards the more 

 scientific system of Cesalpino, recognizing the importance of 

 inconspicuous structures. Describing plant-classifications, 

 Lindley says : " Eivinus invented, in 1690, a system depend- 

 ing upon the formation of the corolla; Kamel, in 1693, 

 upon the fruit alone; Magnol, in 1720, on the calyx and 

 corolla; and finally, Linnaeus, in 1731, on variations in the 

 stamens and pistil." In this last system, which has been 

 for so long current as a means of identification (regarded by 

 its author as transitional), simple external attributes are 



