213 
Relation to the Extinction of Species. 
the converse of its mechanism of development, and hence consists in 
a progressive reduction of the area occupied ? If we follow this assumption 
to its logical conclusion, we must suppose that each species in dying out 
steadily reduces its area, until it finally expires at its birthplace ! This 
seems to be a fallacy comparable with the prevalent idea that structural 
degeneration is c an actual retracing of steps until the point of departure is 
reached \ 1 There are, it is true, certain instances in which the area occupied 
by a dying species shrinks from the margin inwards ; the Oxlip in East 
Anglia is an example , 2 but such cases are probably wholly exceptional. 
This plant [Primula elafior , Jacq.) occupies a restricted region and is 
surrounded on all sides by Primroses ( Primula vulgaris , Huds.), which do not 
penetrate into its area. Year by year the Oxlip and Primrose hybridize on 
the margin of the Oxlip’s demesne ; it seems that the Oxlip will ultimately 
be hybridized out of existence, and its place taken by the Primrose, which 
is present in such hordes that the loss which it suffers by hybridization is 
negligible. It appears to me, however, that, apart from such rare cases as 
this, the tangible evidence of the dying out of a species would more pro- 
bably be a progressive decrease in ‘ commonness on the ground while the 
total area over which the species spreads might remain unchanged almost to 
the last. This gradual decrease in abundance would be a subtle matter to 
gauge and to define, but it ought to come within the scope of ecological 
observations, when the progress of that branch of Botany has enabled 
connected records to be kept over a long series of years. Willis’s methods, 
on the other hand — valuable as they are for their own purpose — are not 
adapted for giving any indication of the progress of extinction, if it takes 
place in such a way that the total area from which the species is known 
remains unaltered. 
For the reasons which I have attempted to outline, I consider that — 
although Willis’s. statistics undoubtedly substantiate the truth of his main 
hypothesis regarding Age and Area— they have no necessary bearing upon 
the question of the extinction of species. 
Balfour Laboratory, 
Cambridge. 
1 There is a useful discussion of this and related questions in Demoor, J., Massart, J7, and 
Vandervelde, E. (1899). 
2 Christy, M. (1897). 
