SEPTEMBER. 
195 
and amongst others Pear trees, are distinct invariable species, remaining always 
identical through all possible generations ; from whence it follows that these trees 
do not spring, as it is commonly believed, from a single or even from a small 
number of specific types which cultivation has caused to vary, but from as many 
primitive types as there are discernible varieties. Thus, to confine ourselves 
to the Pear tree, of which nurserymen already reckon more than five hundred 
varieties, we must admit at least five hundred primitive species, and as these 
exist nowhere in a wild state, logic induces M. Jordan to conclude that their 
domestication ascends to the antediluvian period of man, and that we only 
possess them now because they were preserved in the ark which saved Noah 
and his family. Strictly speaking, the fact may be conceived as possible ; but 
how many suppositions must we heap up, one on the other, to render it probable ! 
Is it not more simple to explain this always increasing multitude of congeneric 
varieties by the principle of the variability of species, if this variability can be 
demonstrated ? But I believe that this has been done. The Academy knows 
already the astonishing transformations which have been observed recently at 
the Museum in the group of Gourds and Melons, where the varieties are .counted 
by hundreds ; the facts which I have to describe in the Pear tree are of the same 
order, and lead to perfectly similar conclusions, which are, on one side the 
contemporary appearance of new races, on the other their instability by crossing, 
and particularly the specific unity of all the races and varieties of cultivated 
Pears. 
In 1853 I sowed numerous seeds of Pears, chosen the previous years from 
four varieties, reckoned as very distinct by all nurserymen,—viz., the old, uni¬ 
versally known Poire d’Angleterre ; the Poire Bose, whose form is that of an 
elongated calabash, and the skin uniformly cinnamon ; the Belle Alliance, short, 
and shaded with yellow and red ; and the Poire Sauger, a wild, or almost wild 
variety, so named because the leaves of the tree remind one, by their white 
down, of the common Sage. For this last sowing I employed all the crop of a 
tree which grows by itself on the road from Marcoussis to Gue. The seeds of 
these Pears sprouted the same year in which they were sown, with the exception 
of those of the Poire d’Angleterre, which did not do so till the following year; 
and this was the case in two different sowings in 1853 and 1854, without my 
being able to discover the cause. 
A small number only of these trees has begun to bear fruit, which I regret, 
because the results which they would have yielded, if all had borne fruit, 
would have been much more varied, and in consequence more conclusive 
than those which I have to submit to the Academy. We may see, never¬ 
theless, at the first glance, on an inspection of the coloured figures, how much 
the fruit, in each of these categories, has been already modified in the first 
generation. 
Thus, in the variety Sauger, four trees which have fructified have yielded 
four different forms of fruit; one ovoid and entirely green ; a second short and 
almost apple-shaped, coloured with red and green; a third still more depressed ; 
and, finally, a fourth, regularly pear-shaped, twice as large as the foregoing, 
and uniformly yellow. From the Belle Alliance nine new varieties arose, of 
which not one reproduced the mother variety, in form, size, colour, or time of 
ripening. There were two especially which I shall mention; one for its size, 
more than twice that of the Belle Alliance, the other for its short major axis, 
calling to mind the apple-shaped Pears or Bergamots. The Poire Bose produced 
three new fruits different from the type; one of the three so like one of those 
obtained from the Sauger, that one could scarcely distinguish it. The variations 
were not less in the Poire d’Angleterre, where six trees yielded six new forms, 
so different from each other and from the mother form, that there are amongst 
