Ridley . — On Endemism and the Mutation Theory. 569 
There is nothing within thousands of miles of them now from which they 
could be evolved. 
In cases like this, and there are very many,' his mutation theory utterly 
collapses. 
But what really puzzles him is the case of endemic species apparently 
actually evolved on the spot, such as Castelnavia , the various locally 
evolved Eugenias, Sonerilas, Impatiens , and such genera where [one finds 
a number of peculiar species in one area. To assume that because we 
have not worked out the whole life-histories and conditions of environment 
of all these different species, their specific differences cannot be due to 
local differences in condition, and they must have been produced by special 
creation, for that is what his mutation theory amounts to, is hardly 
scientific. 
In most cases of mutation and especially where one organ is much 
altered, other organs are also modified; e. g. a dwarfed plant of Liparis 
longipes was imported into Kew Gardens many years ago, in which not 
only the pseudo-bulbs but the leaves and inflorescence and all the organs 
of the flower down to the lip were shortened and broadened. Another 
example is the case of the Vitex trifolia previously described, where, though 
almost every organ in the littoral form was altered in shape, the plant proved 
to be not even a fixed mutation. 
Species absolutely isolated, either by the extinction of all allied forms 
or by accidental position, do not produce a series of mutations. 
There are only two species of the order Dioscoreaceae in Europe at the 
present day, the plants of this order being chiefly tropical. These are 
Dioscorea pyrenaica of the Pyrenees and Tamils communis . Though these 
plants must have inhabited Europe for an immense series of years, being 
undoubtedly the relics of an early tropical era, they have not produced yet 
anything that could be called a mutation. The same remark applies to 
Aphyllantkes monspeliensis , and many other instances might be deduced. A 
great contrast to this is the case of such genera as Hieracium , Rubus , 
Rosa , and Saxifraga , in which genera the number of mutations is very large. 
The same phenomenon occurs in all parts of the world. I would call 
attention to the endemic plants in Christmas Island in this connexion. 
This island has never had any connexion with any other land, being 
a volcanic island dating, it is believed, from the Eocene period. The flora 
which had not been interfered with by man, when it was first examined, 
consisted of plants nearly all of Javanese affinity; many were sea-shore 
species of wide distribution. The endemic flowering plants are thirty-one 
in number, all belonging to separate genera, except two Grewias, two 
Phreatias, ancj two Pandani. Here, in spite of the long period during which 
this island has been above water and capable of bearing a flora, there has 
been no evolution of a large number of any one species comparable with 
