xxx INTRODUCTION. 



(hem, to render their distinctions nugatory, and by another, clear and prc- 

 ,. |s ,.. x,, w . if the snpposed groups are really as closely allied, as for this 

 argument we maj assume them to be, it can be of no possible importance 

 theoretically, whether a given Genus or Order is placed in the one or the 

 nth. r. The near consanguinity of the two does away with all importance 

 ;„ guch B caB e. In Physical Geographyit is of no consequence whether Lon- 

 don is Btationed in liiddlesez or Surreyj and in like manner, in Theoretical 

 Botany, the place of a given Order may be equally indifferent. But it may 



f great consequence practically, because a definition of limits may be 

 possible or not, according to the arrangement. For example, let us take 

 the Solanal and Bignonial Alliances. These touch at the Orders of Night- 

 Bhades and Figworts respectively. If Nightshades are placed in the Bignonial 

 Alliance because of their intimate relation to Figworts, no apparent means 

 remain of clearly denning what is meant by the Bignonial Alliance. If, on 

 the other band, Figworts arc stationed in the Solanal Alliance, then the 

 distinctive characters of that Alliance are also rendered obscure and diffi- 

 cult, or impossible <>f application. But place Nightshades in the Solanal, 

 and Figworts in the Bignonial Alliance, and the language of Botanists 

 affords as clear a discrimination as can be wished for. And so of other cases. 

 Indeed, I am bo persuaded of this, that in my opinion all instances of con- 

 tused and vague characters are only so many proofs of Botanists not having 

 clearly understood the plants that they have endeavoured to classify.* 



It will, perhaps, be alleged that the doctrine just inculcated is directly 

 opposed to the first principles of a Natural System: but such is not the case. 

 No absolute limits, in met, exist, by which groups of plants can be circum- 

 scribed. Tlev pass into each other by insensible gradations, and every 

 group has apparently some Bpecies which assumes in part the structure of 

 some other group. Two countries are separated by a river whose waters 



common to both hanks : in a geographical division of territory the river 

 ina\ be assigned to either the left hank or the right bank, but such an 

 arrangement is arbitrary : and yet the interior of the countries is unaffected 

 hv it. So with the groups of plants; it cannot be of any possible conse- 

 quence whether an intermediate or frontier plant be assigned to one group 

 or another, and convenience alone should be considered in such a matter. 

 This long Bince led me to offer the following observations, the justice of 

 which, much more experience entirely confirms : — "All the groups into 

 which plants are thrown are in one sense artificial, inasmuch as Nature 



igniBes no such groups. Nevertheless, consisting in all cases of species 

 \.rv closely allied in nature, they are in another sense natural. But 



the Classes, Sub-cla »es, Alliances, Natural Orders, and Genera of 



mists, have no real existence in nature, it follows that they have no 

 fixed limit-, and consequently that it is impossible to define them. They 

 are to he cm-id. .vd as nothing more than the expression of particular ten- 

 dencies (nixus), on the part of the plants they comprehend, to assume a 

 particular mode of development. Their characters are only a declaration of 

 their prevailing tendencies." 



We Diust not. however; deceive ourselves with the expectation that by 

 this or anj other expedient definitions in Botany will become possible. 

 Mathematical precision is unknown in Buch subjects, and exceptions occur 



' " lis :r - n " >>'T''<>siv<- remark. It is the misfortune, not the fault of men 



everythtag with ti, w own eyes end that thev are compelled, from 



to take much of all thej study upon trust. In Botany this is most especially 



i >le to examine one-tenth of all the plants he "speaks of, with minute 



