Some day I will test it for a few families.
If you take for comparison with what I send you, the "British Flora," or Koch's "Flora Germanica," or Godron's "Flora of France," and mark the "close species" on the same principle, you will doubtless find a much greater number.

Of course you will not infer from this that the two floras differ in this respect; since the difference is probably owing to the facts that (1) there have not been so many observers here bent upon detecting differences; and (2) our species, thanks mostly to Dr.Torrey and myself, have been more thoroughly castigated.

What stands for one species in the "Manual" would figure in almost any European flora as two, three, or more, in a very considerable number of cases.
In boldly reducing nominal species J.Hooker is doing a good work; but his vocation--like that of any other reformer--exposes him to temptations and dangers.
Because you have shown that a and b are so connected by intermediate forms that we cannot do otherwise than regard them as variations of one species, we may not conclude that c and d, differing much in the same way and to the same degree, are of one species, before an equal amount of evidence is actually obtained.

That is, when two sets of individuals exhibit any grave differences, the burden of proof of their common origin lies with the person who takes that view; and each case must be decided on its own evidence, and not on analogy, if our conclusions in this way are to be of real value.

Of course we must often jump at conclusions from imperfect evidence.