SEPTEMBER. 
197 
the Pear tree, but all do not succeed on the Quince; for example, the Ranee,* 
Clairgeau, Bose, Duchesse de Mars, &c. When we wish to multiply these 
varieties, and for want of the wild Pear we are obliged to employ the Quince> 
the last is grafted on the Jaminette, the Sucree Verte, the Crassane, the Abbeville, 
very vigorous kinds, which are suited to this sort of stock; and when the 
grafts have taken, they receive in their turn varieties whose sap does not 
sympathise with those of the Quince. It is an operation known and practised 
by all nurserymen. 
The relative size of the flowers and appearance of the foliage offer no less 
striking variations. Certain varieties, as the Catillac, St. Gall, Epargne, de 
Vallee, &c., together with wide, rounded and undulated petals, have blossoms 
5 or 6 centimetres (from about 2 to 2^ inches) broad; and their trees, in the 
early stage of foliage, are as white and cottony as the Sauger. Others, like the 
Heric, Sylvange, Fortunee, &c., with oval or lanceolate petals, have flowers half 
the size, their diameter not exceeding 3 centimetres (one-fifth of an inch). 
Finally, we possess at the Museum a Pear tree wrongly named Chartreuse, 
whose linear-lanceolate petals are scarcely 3 millimetres (scarcely one^fifth of 
an inch) broad and 9 millimetres (about three-quarters of an inch) long. It is 
vain, therefore, to seek for specific characters in the proportions of the flower 
or the parts of which it is made up. 
Can characters, however, be found in the size and form of the fruit ? We 
have already seen these elements vary in the experiments detailed above, and 
these were confined to four varieties, of which a few trees only have borne 
fruit. The variations would have been far greater had I been able to try all 
the known varieties of Pear tree. We may judge of the enormous differences 
which occur in respect of size, when I call to mind that the wild Pears, which 
botanists have somewhat permaturely called Pyrus longipes and Pyrus azaro- 
lifera, do not exceed the size of a pea, while our enormous Pears called Poires 
d’Amour and de Livre equal in volume a middle-sized Melon—that is, twelve 
or fifteen hundred times as much. Analogous remarks may be made as to the 
colour of the flesh, which is green, yellow, salmon-coloured or red. 
But perhaps' it may be said these are precisely characters which show a 
specific distinction in these different kinds of Pear trees. Assuredly I should 
ask nothing better, for nothing is so pleasing to the mind of a botanist as definite 
characters, those gaps in the series of congeneric forms, which at the same time 
facilitate his labour and furnish a fulcrum to his nomenclature. He is satisfied 
when these specific, well-defined divisions agree with his ideal notions of nature : 
but unhappily it is not so in the group of Pear trees: from the microscopic 
Pyrus azarolifera and longipes we pass by an insensible transition to the Mille- 
au-godet, a Pear cultivated in the neighbourhood of St. Brienc, which is scarcely 
larger; from this we arrive at the Sept-en-gueule, or little Nutmeg, another 
variety, or rather assemblage of varieties, in which the fruit varies from the 
size of a wood Nut to that of a Walnut. At the same time a multitude of races 
and sub-races, varieties and variations of wild Pears of all sorts of forms and 
magnitudes, from that of the Mille-au-godet to that of our common cultivated 
Pears ; and in these we pass from the smallest to the most gigantic by an 
indefinite series of intermediates, in which every difference of form and 
colour, from the Musette and Cornemuse, which are so curiously elongated, to 
those depressed Pears which have been justly compared to Apples. 
How then, I say, can we lay hold of a specific character of any value in an 
assemblage in which all the most extreme forms are united by insensible and 
* At least, if they do succeed, though they may hear abundantly, the fruit is extremely small 
as, for example, in the Beurre Kance, and scarcely to be recognised when compared with well-grown 
samples. 
