1 6 Willis.— The Evolution of Species in Ceylon , 
logical structure than any other families. What is there in the differences 
between life on rocks in India and life on rocks in Brazil, in both cases 
without external competition, and submerged in running water, to cause 
the Indian species to develop flattened roots which adhere closely to the 
substratum, while the Brazilian species have flattened secondary shoots 
which do not adhere to it? The advocates of adaptation say that the 
water strain must be greater in India. This explanation was good until 
I had personally investigated some 120 waterfalls and rapids in which these 
plants grew in both countries, but there can be no doubt that the Brazilian 
forms, with far less holdfast, often grow in far more rapid water. Since 
writing that the fastest water in which I had found any was in the Rio 
Piabanha, four miles an hour, I have found a large leafy M our era , a 
Tristicha (probably kypnoides), and others growing in the cataract of Ronca 
Pao, near Cantagallo (state of Rio), where the water was flowing at least 
eight miles an hour. They were growing at the bottom of a great water-slide 
of perhaps 1 50 feet in height, with a slope of 30°. Where the water struck 
the ledge on which they were growing, it spouted up above the general 
surface. Never before have I seen plants in such an extraordinary 
situation. 
No explanation of such facts as we have been bringing forward in this 
paper can be produced by the advocates of Natural Selection, who are 
content to say that there must be differences in adaptation imperceptible 
to us, and which we do not and cannot perceive. After forty years no 
one has even been able to make a reasonable suggestion as to what these 
differences may be. But if it be admitted that dispersal goes with age, the 
whole is at once clear, and the way is open for innumerable new investigations. 
Another objection to my views admits the greater rarity of the 
endemic species, 1 and states that the mere fact of wide distribution makes 
a species commoner even among the endemics of a given country, which 
are (on the old theory) developed to suit that country. This position has 
always appeared to me an unsound one, and must have arisen from the 
necessity of explaining the fact that this is often the case, and from thinking 
chiefly of animals, which are able to range about over considerable areas. 
So far as the widely distributed species are concerned, Ceylon must have 
been stocked with them from the plants which were growing at or close 
to the point of junction with India. But why should the mere fact that 
a widely distributed species is growing near Tuticorin enable its offspring 
to spread in Ceylon more widely than a species evolved in the island? 
Why should the mere fact that the ancestors of A (a widely distributed 
species) have lived at all the various stages on the road from, say, Delhi 
to Tuticorin, make A more common in the island than B (a Ceylon- 
1 But does not explain why they are (family by family and genus by genus) graduated in 
rarity from VC down to VR, 
