Homology and Homoplasy 439that ancestry” we would now take as reflecting the presence of similar genetic and
developmental mechanisms in different lineages and the response of those lineages
to internal or external (mutational or environmental) influences to produce similar
phenotypes. Simpson [1965] saw no fundamental difference between parallelism
and convergence, because, he thought, mutations would be more likely to produce
similar effects in closely than in more distantly related organisms.^20
An example that separates parallelism and convergence is the study by Parra-
Olea and Wake [2001] of body (trunk) elongation in salamanders within the tribe
Bolitoglossini. Elongation can arise through the elongation of existing vertebrae,
as in the genusLineatriton, or by an increase in vertebral number, as in the genera
BatrachosepsandOedipina. Knowledge of phylogenetic relationships, emphasized
as so important by Lankester, Willey and Simpson, enabled Parra-Olea and Wake
to determine that these phylogenetic patterns of trunk elongation representparal-
lelismwithin the genusLineatritonbutconvergencewithin the family; it is all in
the depth of the phylogenetic relationships.
10 REVERSALSTo be identified as a reversal, rudiment, vestige or atavism a feature must bear a
high degree of similarity to a character found in an ancestor,^21 a circumstance that
suggests a conserved feature and so raises suspicions concerning relationship among
reversals, rudiments, vestiges and atavisms, and between homology or homoplasy.
Atavisms speak to us of reversibility of the phenotype associated with retention
of developmental information. Rudiments, vestiges and atavisms also speak to
retention of developmental information. Indeed, atavisms and reversals — and
neomorphs, which are novel features found in a few individuals, but characteristic
of all individuals in more derived taxa — can co-occur in the same population,
as demonstrated elegantly for tarsal (ankle) bones in the salamander,Taricha
granulosa, by Shubinet al.[1995].
Areversalis a reversion to a previous evolutionary state, a feature found in all
theadultmembers of a taxon, in phylogenetically related lineages or in an ancestral
lineage but not in the most recent common ancestor of those lineages. The desig-
nation reversal implies a phenotype in a descendent that develops from a develop-
mental programme retained from an ancestor but not expressed in one or more in-
tervening taxa. Phylogenetic character reversal is a synonymous name/process.^22
An oft-cited example, cited here again, is the loss of the second molar tooth in
felids in the Miocene and its reappearance in the extant lynx,Felis lynx[Kurt ́en,
(^20) See Kellogg and Shaffer [1993], Laporte [2000] and Hall [2003a] for discussion and amplifica-
tion of Simpson’s views.
(^21) Of course, in many if not most cases, it may be difficult if not impossible to identify the
common ancestor. In practice, a sister group or outgroup is used as the surrogate that indicates
phylogenetic affinity.
(^22) See McShea [1996], Bejder and Hall [2002] and Hall [2003a] for reversals, and see Stiassny
[1992] for phylogenetic character reversals.