Homology can't be evidence of ancestry if it is defined thus (Talk.Origins)
Homology is defined as similarity due to common ancestry. The claim then that it is evidence for common ancestry is a circular argument.
(Talk.Origins quotes in blue)
1. Homology is not defined as similarity due to common ancestry and then used as evidence for common ancestry. Rather, the evidence for common ancestry comes from the patterns of similarity of many traits. These similarities show that organisms group naturally into a nested hierarchy. For example, that ladybugs and scarabs are both types of beetle is based on various common traits such as hardened front wings; beetles, flies, and grasshoppers are types of insect; insects, scorpions, and centipedes are types of arthropod. Such grouping does not depend on any assumptions about origins and in fact was first codified by Linnaeus, a creationist. A grouping suggested by many common traits is evidence of common ancestry. This is true no matter what you choose to call the traits. The homology label gets added after the evidence for common ancestry is already in.
If one starts reading on page 59 instead of page 63 as Talk Origins says, it is obvious that Wells is well aware of the fact that the term homology predates Darwin and that it was not originally defined in terms of common decent. As a side note, it is interesting how Talk Origins takes issue with its own statement about Carolus Linnaeus being a creationist.
Furthermore Talk Origins claim that "Homology is not defined as similarity due to common ancestry" is wrong. It is often defined as such in classes and text books even if not formally defined as such.
Homology - similarity of features based on common descent.
Even if similarity due to common ancestry is not the formal definition of homology, it is indeed the practical definition of homology. Before similarities between two organisms—regardless of how similar they are—are considered homologous by Evolutionists, they must first be considered to result from common ancestry.