M. J. DEGAXKEFE ON Vi-EIABILITY IN THE PEAll-TKEE. 43 



mixture of characters, to make an arrangement in the least degree 

 natural, and which would embrace all the known varieties. At 

 the commencement of -my studies, like my predecessors, I thought 

 that I might undertake this work with some chance of success ; 

 now I am disabused of this hope, and I do not fear to declare that 

 every classification will be purely artificial. The only useful prin- 

 ciple which can be adopted here will be, I think, the time of the 

 ripening of the fruit, because in an economical point of view 

 this consideration predominates over all others ; and even here, 

 again, we must assign very wide limits to these seasons of 

 maturity. 



Neither the form of the fruit, nor their size, nor their colour, 

 nor their flavour, any more than the habit and appearance of the 

 trees, the colour of the wood, the size of the leaves and flowers, 

 &c, can afford any base for a classification, because all these 

 characters are purely individual, which they do not transmit 

 faithfully by way of generation, and which, as there are not want- 

 ing examples to prove, change soon in one and the same individual 

 in consequence of local circumstances which one cannot always 

 explain. 



The partisans of the plurality of species may object, in the group 

 of trees with which we are occupied, that if in this multitude of 

 intermediate forms we are unable to recognize distinct specific 

 types, this depends on the fact that the primitive species have in- 

 tercrossed thousands of times ; that their fertile hybrids have 

 increased in an enormous degree the number of crosses, and that 

 from thence have sprung these innumerable forms which are the 

 despair of classifiers. I am far from denying the fact of these 

 crosses or of their influence ; I say even that nothing appears to 

 me more probable ; at least it is not possible to doubt it, when we 

 see what takes place in a pear-orchard when in flower, where the 

 bees, attracted from a distance of a league, pilfer from morning till 

 evening, mingling the pollen of all the varieties, and disseminating 

 it on stigmas for which it was not destined by nature. But we 

 may remark that these impregnations, which are supposed to be 

 unnatural, are always fruitful, that all the flowers which receive 

 pollen from any kind of pear whatsoever, set their ovary, and 

 that the fruit when developed always contains fertile seed*. 



* I know no apparent exception to this fertility, except in the Poires sans 

 pepins and Comte de Flandre, whose fruit contains no seeds ; but this does not 

 prove a want of power in the pollen, which, besides, might as well be that of its 

 own parent tree as of a tree of any other variety. In fact, I have ascertained 



