454 LETTERS TO DARWIN, 1843-1859 
a great deal on amount of variability in great and small 
genera, and find it exceedingly difficult to explain logically 
the practical reasons there are against Botanists making 
varieties of well marked species, i.e. of small genera. Many 
of the small genera still kept up would never have been made 
at all, had the whole of the Natural Order as now known been 
known when those genera were made. E.G., in Europe 
we have, say 3 very different members of a large unkjiown 
Asiatic group of plants, certainl}" 100 species : of these 3 
as many genera are made in Europe : but after getting all 
the 100 Asiatic species, though these show that the said 3 
genera are naught, we do not therefore cancel them, but in 
9 cases out of 10 we group the Asiatic species as best we can 
under the 3 European genera. A thousand unphilosophical 
reasons occm-, of considerable {^present practical) weight to 
keep up the said old genera. 
We must never forget that Systematists have two very 
different ends to meet : 1. To provide a ready nomenclature 
without which the science cannot advance and which we 
change as little as possible— and further use every means 
to avoid even a necessary change — so important is it for 
all to get up the nomenclature, and so bulky and complicated 
is this nomenclature. 2. To arrange the members of the 
Vegetable Kingdom, scientifically, which is only done for 
the sake of scientific followers. Now we repeatedly find 
that to express our views scientificalty we must break up 
the whole nomenclature, and rather than do this excessively, 
we confine ourselves to stating our views without acting 
upon them. In no respect do we sacrifice more to the 
utilitarian purpose of nomenclature, than in keeping up 
small had genera. 
Practically no one (except a few of us) hesitates to remove 
a very distinct species of an old genus, especially if its 
characters are constant and it is an invariable jpkmt, and to 
make of it a new genus, just because it is more unlike its 
20 neighbours than they are unlike one another. The 
probabilities in this case are that the 20 are varieties of 
8 or 10, and being variable have varieties made of them- 
selves, whilst the one constant plant goes to a new genus, 
and is a small genus with no varieties. IX 
Again, practically very few do up an old genus of one 
