CLASSIFICATION. 207 



of language, we find it convenient to classify words into 

 what we call parts of speech, and to divide and sub-divide 

 these again, in order to draw finer dictinctions, so, in our 

 study of plants, it will be necessary to arrange them 

 first of all in comprehensive groups, on the ground of 

 some characteristic possessed by every member of each 

 group. Just as, in Latin, every noun whose genitive 

 case is found to end in ce is classed with nouns of the 

 first declension, so in Botany every plant presenting 

 certain peculiarities will be placed in a group along with 

 all the other plants presenting the same peculiarities. 



358. Some hints have already been given you as to 

 the kind of resemblances upon which classification is 

 based. For instance, an immense number of plants are 

 found to produce seeds with a dicotyledonous embryo, 

 while an immense number of others have monocotyledon- 

 ous embryos. This distinction, therefore, is so pro- 

 nounced, that it forms the basis of a division into two 

 very large groups. Again, a very large number of 

 dicotyledonous plants have their corollas in separate 

 petals ; many others have them united, whilst others 

 again have no petals at all. Here, then, is an opportunity 

 to sub-divide the Dicotyledons into polypetalous, gamo- 

 petalous, and apetalous groups. And so we go on, always 

 on the plan that the more widely spread a peculiarity is 

 found to be, the more comprehensive must be the group 

 based on that peculiarity ; and so it happens, that the 

 smallest groups of all come to depend upon distinctions 

 which are, in many cases, by no means evident, and upon 

 which botanists often find themselves unable to agree. 



359. As our divisions and sub-divisions will necessarily 

 be somewhat numerous, we shall have to devise a special 



