62 AGE AND AREA [pt. i 



therefore be looked upon as in general descended from the wides, 

 and as young species just commencing their careers. 



I called this hypothesis, that on the average the area occupied 

 by species in a country depended upon their age within that 

 country, by the convenient jingle of " Age and Area" (123, p. 337, 

 footnote), and from the very first I was careful to point out that 

 this result Avas only true when averages of about 15 allied species 

 were taken. People, however, have nearly always insisted upon 

 applying the rule to individual cases, and then complaining that 

 it does not fit the facts. In regard to the facts that have just 

 been discussed, for example, they say "there are many VC 

 endemics, and a lot of VR wides, so it must be wrong." A simple 

 illustration will perhaps make my position more clear. 



Suppose that five wides are approaching Ceylon (then attached 

 to the mainland), spreading at a uniform speed, and let the dis- 

 tance from the foremost of A to the foremost of B be represented 

 by 2, that from B to C by 2 also, and so on. Then A will reach 

 Ceylon first, and when B reaches the island A will occupy there 

 a space represented by 2. When C arrives A will occupy 4, and 

 B 2. Ultimately they will occupy spaces represented by 10, 8, 

 6, 4, and 2. Now let each give rise in South India to another 

 species a, /3, y, S, e, each always at a distance beliind its parent 

 represented by 2. Then a Avill arrive in Ceylon simultaneously 

 Avith B, ^ Avith C, and so on, and these Ceylon-Peninsular-Indian 

 species Avill ultimately occupy areas represented by 8, 6, 4, 2, 

 and 0. And if, lastly, each species, Avhen it has reached a dis- 

 tribution in Ceylon represented by 2. gives rise to a Ceylon 

 endemic, then if avc subtract 2 from the figures of distribution 

 of all the preceding species, Ave shall get the distribution of the 

 endemics. This Avill be, for the endemics derived from the Avides, 

 8, 6, 4, 2, -, and for those derived from the Ceylon-Peninsular- 

 Indian species 6, 4, 2, — , — . 



NoAV the most Avidely distributed endemic, derived from A, 

 the first Avide to arrive, Avill have a range of 8, Avhile three out 

 of five of the Avides, and three out of four of the Ceylon-Penin- 

 sular-Indian species Avill have ranges of 6, 4, or 2, considerably 

 less. If one attempt to apply the rule to individual' cases, it is 

 at once liable to break doAvn. Bvit if Ave add up the dispersal of 

 all the Avides, and divide by the total of species, Ave get 

 (10 + 8 + 6 + 4 + 2 =) 30 ^ 5 = 6 as the average range of a 

 AAdde, 20 ^ 4 = 5 as the average range of a Ceylon-Peninsular- 

 Indian species, and 32 -^ 7 = 4-5 as the average range of an 



