NATURE 
17 
THURSDAY, JANUARY 3, 1907. 
KINDRED 
PLANT DISPERSAL AND 
PROBLEMS. 
Observations of a Naturalist in the Pacific between 
1896 and 1899. By H. B. Guppy. Vol. ii. Plant 
Dispersal. Pp. xxvit+627. (London: Macmillan 
and Co., Ltd., 1906.) Price 21s. net. 
EW of the problems that confront the naturalist are 
wider in their range of interest than those connected 
with the origin of the present inhabitants of an oceanic 
island. Such a population is almost always a very 
mixed one, though it can usually be roughly divided 
into two classes, the one embracing the aboriginal 
or endemic element, whilst the other is composed of 
colonists hailing, it may be, from widely-sundered 
centres of emigration. But closer investigation shows 
that such a distinction is, after all, not a very pro- 
found one. The forbears of the endemic groups were 
themselves at one time colonists; time and circum- 
stance have permitted and encouraged divergent varia- 
tion, and so new types have arisen. The causes re- 
sponsible for the variation itself for the most part 
elude recognition, and their study is the business of 
the physiologist rather than of the naturalist, but the 
effects may well serve to concentrate the attention 
of the latter on the larger problems bearing on the 
nature and significance of adaptation no less than 
on those more directly concerned with the sources and 
mode of dispersal of the individual species. 
Mr. Guppy, in his ‘‘ Observations of a Naturalist 
in the Pacific,’? has kept both sets of problems clearly 
before him, and has produced a book that will 
deservedly appeal to a wide circle of biologists, and, 
indeed, to all who are interested, not merely in the 
details, but also in the philosophical aspects of dis- 
tribution. 
The author will probably not expect his ideas to 
command universal assent. It is inevitable, and in- 
deed desirable, that divergence of opinion should 
exist as to the true explanation of phenomena which 
are still but imperfectly comprehended. But such dis- 
sentience in no way detracts from the value of his 
work. Perhaps the highest praise that can be earned 
by any contribution to science is to say of it that its 
facts are really facts, and its theories, whatever be 
their ultimate fate, are stimulants to further re- 
search. 
Although the book before us deals with matters 
affecting the distribution of plants in general, the 
subject is more especially considered in relation to 
the littoral floras of the Pacific islands. The author 
takes as types to illustrate the wider bearings of his 
own observations the floras of Hawaii, Tahiti, and 
Fiji. He discusses very fully the various causes which 
in the case of the three island groups chosen have 
produced results so dissimilar, notwithstanding the 
equality of conditions which at first sight appears 
to prevail between them. 
The keynote of the explanation advanced to account 
for the facts is to be found in the buoyancy of the 
NO; T940, VOL. 75] * 
seed or fruit. But whilst many of the author’s con- 
clusions are both suggestive and interesting, it may 
be doubted whether others will be prepared to accept 
his views as to the connection between buoyancy and 
habitat. 
Put briefly, his position is this. The capacity of 
floating determines the position of those species pos- 
sessing buoyant fruits or seeds by the river or on 
the coast. Which of the latter stations is actually 
occupied will depend on the degree of transpiratory 
activity on the part of the plant, that is to say, on 
whether it happens to be a xerophyte or not. If the 
former, then it will be chiefly restricted to the coast, 
but if not it will be precluded from occupying a posi- 
tion that is only suited to plants that can withstand 
physiological drought, and it will be driven to take 
up its position by the river or pond. Mr. Guppy ex- 
pressly and repeatedly states his conviction that it 
is not the station which is responsible for the develop- 
ment of buoyant seeds, but that the plants so char- 
acterised reach and colonise littoral or riparian 
stations because they can be distributed to them; in 
other words, position does not determine buoyancy, 
but buoyancy determines the station, subject to the 
further sorting-out process which is associated with 
xerophilous or hygrophilous habit. 
But the author himself shows that seeds or fruits 
of the same species may exhibit great variation in 
their power of floatation, some sinking at once when 
placed in sea water, whilst others from the same tree 
may float without injury for months. He goes 
farther than this, and emphasises the need, in making 
such tests, of taking seeds from plants grown under 
similar conditions. Thus in Afzelia bijuga, experiment 
showed that of the seeds taken from a littoral ex- 
ample 7o per cent. floated in sea water, whilst of 
those gathered from an inland individual of the same 
species only 23 per cent. were able to swim. 
Several other similar examples could be cited. Such 
evidence would seem, however, to indicate that the 
environment is not without its direct influence in 
determining the floatability or the reverse of the 
seeds, and clearly if this is so natural selection has 
material enough to work on. At any rate, one would 
hesitate before accepting the author’s conclusions as 
to the real relation between buoyancy and station. 
A very useful account is given of the several struc- 
tural features to which buoyancy may be due. As they 
are often remarkably simple, it seems not unreason- 
able to think that an experimental study of the direct 
reaction of the plant to the environment in such 
cases as these would almost certainly yield interesting 
and valuable results. 
The author has some excellent remarks on the 
nature and origin of adaptive characters, and they 
deserve to be carefully read by members of that large 
but somewhat careless class of naturalists who 
imagine that when a structure has been shown to 
be useful for a particular purpose, its occurrence 
is thereby ‘‘explained.’’ Nothing can be farther 
from the truth, and the more we know of ‘ adaptive ”’ 
structures the less. directly does their origin appear 
| ite be related with the function they ultimately dis- 
L 
