Williams - p. 7 
place of origin will, because of the factor of greater time available for 
differentiation and because of the severe competition of close relatives, 
show more derived states than those members of the lineage that have 
escaped to other areas. It is not in fact an unusual observation that the 
colonizing derivatives of mainland stocks show primitive states that are no 
longer represented in their place of origin. Within Anolis Williams calls 
attention to the persistence of fracture planes in the caudal vertebrae of 
species of the roquet group while all their mainland relatives of the 
latifrons group have lost caudal autotomy. Similarly, Etheridge regards 
the Jamaican betas as the most primitive in some regards of that entire 
‘series, Other instances could be provided. 
Differentiation of the carolinensis group on Hispaniola and the 
bimaculatus group on Puerto Rico puts each of these in a geographic position 
suitable for the colonizations each has achieved: the carolinensis group 
has colonized Cuba and Puerto Rico from Hispaniola, the bimaculatus group 
Hispaniola and the northern Lesser Antilles. 
Williams had imagined that Puerto Rico, as the first landfall for 
anoles in the eastern Greater Antilles, was a "generator island," i.e. an 
area that, by the accident of time of colonization was a first center of 
radiation, supplying stocks as they arose to later colonized islands. He 
is now compelled to admit for Puerto Rico, with the invasion from Hispaniola 
of the ancestor of occultus, the classic pattern of double invasion and 
accumulation, 
With the demonstration that Puerto Rico has been colonized by anoles 
more than once, it appears that all Greater Antillean anole faunas present 
a pattern of both radiation and accumulation, Jamaica is now the island 
that appears to be most nearly a "generator island," since according to 
current belief all but one of its species originated within the island, 
According to this Jamaica has, despite its relatively small size, furnished 
the sagrei group to Cuba (and the grahami-related conspersus to Grand Cayman) 
and has received in its turn only the back invasion of sagrei itself - a 
late and perhaps even human-aided invasion. Now, in view of the sheplani- 
occultus case, Williams wonders whether the sagrei group did not differentiate 
on Cuba (from, of course, a stock invading from Jamaica) and whether there 
have not been two back invasions - an old one, valencienni, and a much later 
one, sagrei. 
Discussion of these examples of radiation-colonization sequences logic- 
ally leads on to the review planned by Williams of these phenomena for the 
West Indies as a whole. Antecedent to this, however, is a more precise 
assessment of the relationships - affinity and phyletic distance - of all West 
Indian anoles. Williams has prepared a new arrangement of West Indian anoles 
into species, basing his judgments on the accumulated evidence of morphology, 
karyology and electrophoresis and, where relevant, behavior and ecology. He 
will, almost of necessity in so wide a review, be in disagreement at some 
point or other with the private opinions of almost every Anolis worker. He 
promises only to discuss the points of controversy. 
