CLASSIFICATION. 203 



266. (2.) Every natural classification takes into consider- 

 ation not only the adult characters, but also those of the 

 embryonic life of its objects. It is not enough to know the 

 differences and resemblances between two plants in their 

 adult state ; we must also know whether they differed or not 

 in their modes of reaching that state. In other words, in 

 order to determine the degree of relationship existing be- 

 tween two or more plants, all the characters of each plant, 

 as presented in its whole life, must be taken into the ac- 

 count. By ignoring this important law great confusion has 

 arisen, especially in the lower groups of plants. 



267. (3.) There is still another factor which should 

 enter into classification. Every classification should show 

 real relationship, not similarity alone ; it should bring to- 

 gether not those which simply show present coincidences, 

 but those in which similarity of form indicates similarity 

 of origin ; in addition to structural relationship, it should 

 show genetic relationship. This can be accomplished only 

 by a study of the genealogy of plants, a subject surrounded 

 by many difficulties. In but few cases can we trace an 

 ancestral line, and yet it is desirable that we should use the 

 facts we have, as by so doing we shall be the more likely to 

 discover others. 



(a) It is a mistaken notion that living things can be grouped natu- 

 rally by taking into consideration only one, or even two or three char- 

 acters. Botany and zoology are full of the debris of attempts at classi- 

 fications upon single characters, and in every case such classifications 

 have proved a hindrance to knowledge. The division of the vegetable 

 kingdom into Flowering and Flowerless Plants, by Ray,* in 1703, is an 

 illustration of one based upon a single character. The influence of 

 this classification, which is even yet much followed, has been injurious. 

 It has kept alive the notion that the so-called Flowerless plants are 

 quite different as to their reproductive organs from the Flowering ones ;' 

 it fixed an imaginary gulf between groups of plants, some at least of 

 whiph are in nature placed side by side. Endlicher'sf two great 

 groups, Cormophyta and Thallophyta, are likewise based upon a single 

 character, and are, as a consequence, misleading. The Thallophytes are 



* John Ray : " Methodus Plantarum emendata et aucta." 

 f Stephen Endlicher : " Genera Plantaruni secundum Ordines Natu- 

 rales disposita." 1836-40. 



