Globin evolution was apparently very rapid in early vertebrates: a reasonable case against the rate-constancy hypothesis.

Kimura mistook ambiguous maximum parsimony codons for wrong codons. The maximum parsimony method performed well as judged by the two classes of serine codons (which can not be connected by silent mutations) on comparing the parsimony codons for serines in human, rabbit, and mouse alpha hemoglobin chains to actual codons determined by nucleotide sequencing. In genealogical reconstructions involving 247 eucaryotic globins, the maximum parsimony distances separating the contemporary sequences show that Kimura's Poisson and Dayhoff's PAM estimates of rate of globin evolution miss most of the superimposed replacements and are therefore seriously in error. Nor is Kimura's constant rate assumption and his belief in a single origin of myoglobin supported. Lamprey myoglobin appears to be most like lamprey hemoglobin, while gnathostome myoglobin seems closest to gnathostome hemoglobin. It was found that the three types of gnathostome globins (Mb, alpha Hb, beta Hb) evolved between the shark-boney vertebrate and bird-mammal ancestors at a much faster rate than from the latter ancestor to the present. The data indicate that rates were exceedingly fast during the origin of these globin chains because a high proportion of substitutions were adaptive. It was concluded that wherever strong stabilizing selection acts on a protein, somewhere in the past positive Darwinian selection must have spread the amino acid substitutions now being preserved.