of the Dichotomous System. 139 



but I first pointed out its nature and its general application, 

 and called the attention of naturalists to the subject. 



Dr. Fleming says, that the distinction between these two 

 relations is only respected by me (for, as I have shown, La- 

 marck never made it) when they suit my views. This is to a 

 certain degree true; but let us examine the full value of the 

 remark, and we shall find that it is a proof of my respect for 

 nature. Let us take his definition of relations of affinity 

 and analogy which he fathers off upon me, but to which he 

 nevertheless appears to give his full consent. Relations of 

 affinity, says he, are relations of resemblance between different 

 objects compared with each other; and relations of analogy 

 are the relations of particular parts of different objects. Why, 

 if this be all the distinction, there is in reality none ; the last 

 kind being clearly involved in the first — a resemblance between 

 parts being only a partial resemblance between the wholes. 

 How then is this confusion between the two relations to be 

 prevented? By applying another and most necessary con- 

 dition to relations of analogy, namely, their parallelism. And 

 now Dr. Fleming will understand the reason why I only 

 respect the distinction he makes between relations of affinity 

 and analogy, when the latter suit my views of their necessary 

 parallelism. I will repeat here for him, what I long ago said 

 on the subject in the Linnaean Transactions: " The theoretical 

 difference between affinity and analogy may be thus explained. 

 Suppose the existence of two parallel series of animals, the 

 corresponding points of which agree in some one or two re- 

 markable particulars of structure. Suppose also that the ge- 

 neral conformation of the animals in each series passes so 

 gradually from one species to the other, as to render any in- 

 terruption of this transition almost imperceptible. We shall 

 thus have two very different relations, which must have re- 

 quired an infinite degree of design before they could have 

 been made exactly to harmonize with each other. 



" When therefore two such parallel series can be shown in 

 nature to have each their general change of form gradual, or, 

 in other words, their relations of affinity uninterrupted by any 

 thing known; when moreover the corresponding points in 

 these two series agree in some one or two remarkable circum- 

 stances, they afford relations of analogy, and there is every 

 probability of our arrangement being correct. It is quite in- 

 conceivable that the utmost human ingenuity could make these 

 two kinds of relation tally with each other, had they not been 

 so designed at the Creation."— See also Linn. Trans., vol. xiv. 

 note, p. 52. 



If naturalists did but study the works of MM. Fries and 



T 2 Agardh 



