The immigration may have been from a southerly direction, and it seems that some few African as well as coldish plants are common to the mountains to the south.
Believing in occasional transport, I cannot feel so much surprise at there being a good deal in common to Madeira and Canary, these being the nearest points of land to each other.

It is quite new and very interesting to me what you say about the endemic plants being in so large a proportion rare species.

From the greater size of the workshop (i.e., greater competition and greater number of individuals, etc.) I should expect that continental forms, as they are occasionally introduced, would always tend to beat the insular forms; and, as in every area, there will always be many forms more or less rare tending towards extinction, I should certainly have expected that in islands a large proportion of the rarer forms would have been insular in their origin.

The longer the time any form has existed in an island into which continental forms are occasionally introduced, by so much the chances will be in favour of its being peculiar or abnormal in nature, and at the same time scanty in numbers.

The duration of its existence will also have formerly given it the best chance, when it was not so rare, of being widely distributed to adjoining archipelagoes.