40 

NATURE 
[JANUARY 13, 1923 
the circle will be a function of the age of the culture ] remains of such creatures form conspicuous fossils. 
until the medium is covered. The species or genera 
in the course of their dispersal are held to throw off 
new species and new genera, each of which again 
spreads concentrically from the focus of its inception. 
The throwing off of these new forms of life is regarded 
by Dr. Willis as a “ casual” process, and regarding it 
some very definite inferences are drawn, of which we 
will speak later. 
Now every evolutionist agrees that, apart from 
disturbing elements, area is a measure of age. If the 
matter rested there nothing would be in dispute, but 
nothing fresh would have been contributed to the 
discussion. We are, however, asked to believe that 
in practice this mode of estimating the age of a species 
is, on the whole, trustworthy: that endemic species 
and rarities in general can and must be for the most 
part accepted as new starters in evolution, and not 
as survivors. That is, of course, a paradox, but it 
constitutes the main thesis of the work. Dr. Willis 
takes the floras of Ceylon and New Zealand into 
special consideration, besides those of other isolated 
places, mountain tops and remote islands, and in 
brave defiance of all that science has hitherto taught 
us regarding the peculiar plants and animals limited 
to such localities, he tells us that, on the whole, the 
reason why those creatures occupy such small areas 
is that they have not yet existed long enough to have 
spread far. If any one objects that in application 
to the special cases which immediately suggest them= 
selves, Sphenodon, the dodo, Leucodendron, etc., 
such a contention is preposterous, Dr. Willis would 
reply that he knew as much already, and that he is 
concerned not with special cases, but with averages 
and general propositions. He is within his right. 
The second proviso is that comparative estimates of 
age are only to be based on area when forms within 
the same “ circle of affinity ” are compared. 
Everything then turns on the computation of these 
averages and on the criteria by which “circles of 
affinity” are to be recognised. Unfortunately no 
means are suggested by which we are to tell whether 
a species or genus is a novelty or a relic, and obviously 
none can be forthcoming. We may make shrewd 
surmises, but if things like that could be declared 
with certainty the study of evolution would be on the 
way to becoming an exact science. Meanwhile 
estimates of age based on area “ occupied ” must be 
exceedingly hazy. Giant tortoises live in the Mas- 
carenes and in the Galapagos, and therefore must be 
reckoned ancient, as they doubtless are. When they 
become extinct, say in the Mascarenes, which they 
presumably will, they would start again as novelties 
at the bottom of the list, but for the accident that the 
NO. 2776, VOL. III] 


Of the New Zealand shrubby Veronicas one, V. — 
elliptica, occurs also in Fuegia; having the widest 
recorded range it must be deemed! by far the oldest of 
these species. Once extinct in either locality, whether 
Fuegia or New Zealand, it would be ranked with the 
rest of the New Zealand species as new mutations. 
Then again the surface of the terrestrial globe is, 
as we all know, a medium of complex heterogeneity. 
By no provisos, safeguarding clauses, or anticipatory 
exclusion can considerable areas be defined in which 
dispersal may be observed which has not been pro- 
moted or limited, diverted or arrested, by countless 
interferences. Very rarely, if ever, do we find that 
reasonable uniformity and constancy of conditions, 
even in space, let alone time, without which we are 
warned the theory must not be applied. In areas 
which may be judged most uniform at a given point 
of time, the operation of sharply limiting causes is 
manifest. If, as in prairies and steppes, for hundreds 
of miles the conditions appear geologically and meteoro- 
logically uniform, the mere presence of living things 
introduces heterogeneity. Dr. Willis is well aware of 
this. In one of his best chapters he discusses 
“barriers? in the widest sense, and he makes us 
realise how difficult it must be for a new-comer species 
to get a footing or to spread among plant-associations 
already established. On the Central Asian steppes, 
for example, one can distinguish on the remote horizon 
by their colour the spots where encampments have 
stood. These patches are mainly characterised by 
the presence of nettles, which grow in such places. 
Nettles, as Dr. Willis remarks, are very easily dis- 
persed by wind, yet nowhere else do they establish 
themselves in the Artemisia steppes—only in places 
which man and his animals have made fit for their _ 
growth. European weeds abound in the Eastern 
States where the soil has been cultivated, but few 
invade patches of unbroken territory. Quantitative 
estimates of the allowances to be made for hetero- 
geneities and barriers in general cannot be attempted. 
Therefore in the hope that the heterogeneities will be 
so many and so various as to cancel, a reservation is 
introduced to the effect that the groups of species to 
be compared should each be not less than ten in 
number. But the difficulty is a real one, and in 
dealing with any troublesome or unconformable 
phenomena these considerations provide endless loop- 
holes for escape. 
A still more formidable difficulty is encountered in 
the endeavour to declare which classes of forms may 
be compared legitimately with the object of determin- 
' ing their relative ages from the areas they occupy, 
and which are not comparable. To have some con- 
