194 Willis. — The Relative Age of Endemic 
estimated in range by 360 miles. 1 Even for Ceylon, to equalize the figures 
(6, p. 4) 687 endemics must go up a class, which means discovery in new 
localities, and 699 wides must go down a class, which means proof of wrong 
identification for localities already recorded. 
The New Zealand figures confirm those of Ceylon to the necessary 
degree of safety. The Ceylon figures served their purpose in drawing my 
attention to the law of age and area which I based upon them. Had it not 
been that the species were thus conveniently classified into groups, I doubt 
if I should have thought it worth while to undertake the enormous labour 
of determining the area occupied by each, though I was determined to find 
out all that could be found out about the endemics. 
To apply my law to individual cases is to invite mistake, but no amount 
of evidence of individual exceptions will shake it. Exceptions must be 
brought up in gro 7 ips of twenty allied species , behaving alike. Assuming 
that lalang were native (see 1, p. 572) and spread without alteration of 
the previous conditions, no argument can be based upon it unless it were 
accompanied by nineteen other exceptional Gramineae. Even the other 
species of Imperata is VR, which at once halves the commonness of the 
lalang. 
Exceptions occur by the dozen, but there is no evidence in the figures 
to show that similar exceptions form any large percentage. Probably there 
are a fair number of relic endemics, of cases where adaptation really occurs, 
of cases where Natural Selection has been beneficially in operation, and the 
like, but they do not show in the figures, which simply go to show the over- 
whelming effect of mere age. 
Relative Age of Endemics and Wides. 
Mr. Ridley further states that the endemics are nearly all the relics of 
an old flora rapidly disappearing, and thus brings up the line of opposition 
which so far I have most frequently encountered. With the view of replying 
to it in advance, I read at the Newcastle meeting of the British Association 
a paper with the title ‘ Are endemics the oldest or the youngest species in 
a country ? * and I shall now proceed to quote a considerable portion of this 
paper, as Mr. Ridley’s objection would obviously make them out to be the 
older species. 
‘ In several papers recently published, I have brought forward a law 
which I propose, indicating that the geographical distribution of species 
(taken in groups of twenty or so allied forms) depends chiefly upon their age 
within the country, Natural Selection, whatever results it may produce in 
1 Rarity of wides 3*5, of endemics 6*5, in figures from 1 to 10. Each unit represents 
120 miles, 
