extinction.")

Down, November 15th [1855?].

In Schoenherr's Catalogue of Curculionidae (41/2. "Genera et Species
Curculionidum." (C.J. Schoenherr: Paris, 1833-38.)), the 6,717 species
are on an average 10.17 to a genus. Waterhouse (who knows the group well,
and who has published on fewness of species in aberrant genera) has given
me a list of 62 aberrant genera, and these have on an average 7.6 species;
and if one single genus be removed (and which I cannot yet believe ought to
be considered aberrant), then the 61 aberrant genera would have only 4.91
species on an average. I tested these results in another way. I found in
Schoenherr 9 families, including only 11 genera, and these genera (9 of
which were in Waterhouse's list) I found included only 3.36 species on an
average.

This last result led me to Lindley's "Vegetable Kingdom," in which I found
(excluding thallogens and acrogens) that the genera include each 10.46
species (how near by chance to the Curculionidae), and I find 21 orders
including single genera, and these 21 genera have on average 7.95 species;
but if Lindley is right that Erythroxylon (with its 75 species) ought to be
amongst the Malpighiads, then the average would be only 4.6 per genus.

But here comes, as it appears to me, an odd thing (I hope I shall not quite
weary you out). There are 29 other orders, each with 2 genera, and these
58 genera have on an average 15.07 species: this great number being owing
to the 10 genera in the Smilaceae, Salicaceae (with 220 species),
Begoniaceae, Balsaminaceae, Grossulariaceae, without which the remaining 48
genera have on an average only 5.91 species.