34 SOME THOUGHTS ON CLASSIFICATION 



efforts before this could be done with any approximation to accuracy.. 

 In the mean time new wants arose : as men contemplated the variety 

 of organised beings passing continually under their view, they could 

 not help perceiving resemblances which had nothing to do with the 

 technical arrangements they employed. They felt, in spite of their 

 systems, that objects which they had placed together, were on the- 

 whole very unlike, whilst others, far removed by their classification, 

 impressed them with a sense of near relationship. Comparison of 

 systems founded on different characters, as seemed best to different 

 observers, showed the defect to be a general one, and thus arose the 

 perception of the difference between natural and artificial methods, 

 and the advantage of the former, provided they can be practically 

 applied. It was in relation to the Vegetable Kingdom that this sub- 

 ject was first discussed, and the great Linnseus, the author of the 

 clearest, most precise, and most practically useful system founded on 

 resemblances in some one class of characters, formed the opinion that 

 really natural groups could only be marked out by the sagacity of the 

 most experienced observers, being a sort of guesses at truths incapable 

 in the nature of things of satisfactory proof, and that such groups 

 could not be definitely characterized, so that however interesting to 

 the enlightened lover of nature, they could afford no aid to the stu- 

 dent in tracing the history of the objects passing under his notice. It 

 was with these views that he laid before his pupils his own most saga- 

 cious, and notwithstanding all the difficulties in his way, frequently 

 successful attempts at collecting plants in natural orders, as they were 

 called, whilst he had no doubt that his artificial system, formed 

 chiefly on the number of the most essential parts in flowers, or some 

 equally artificial plan, must continue to be used for tracing plants to 

 their name and what is known of their history. His principal followers 

 entertained the sam.e views as to the impracticability of the natural 

 system even after Jussieu had succeeded in giving good distinguishing 

 characters of natural families. But what it concerns us now to 

 inquire is, what is the distinctive quality by which a natural is to be 

 known from an artificial system, or what we precisely mean by calling 

 an arrangement natural ? Let it be observed then that in an acknow- 

 ledged artificial system each group is set apart by some single charac- 

 ter common to all its members, and wich is chosen for the conve- 

 nience with which it can be tested and applied, whilst the character- 

 istic marks of a natural group are numerous and employed with some 



