Activation of a visual pigment molecule to initiate phototransduction requi
res a minimum energy, E-a, that need not be wholly derived from a photon, b
ut may be supplemented by heat(1), Theory(2,3) predicts that absorbance at
very long wavelengths declines with the fraction of molecules that have a s
ufficient complement of thermal energy, and that E-a is inversely related t
o the wavelength of maximum absorbance (lambda(max)) of the pigment, Consis
tent with the first of these predictions, warming increases relative visual
sensitivity to long wavelengths(4-8). Here we measure this effect in amphi
bian photoreceptors with different pigments to estimate E-a (refs 2, 5-7) a
nd test experimentally the predictions of an inverse relation between E-a a
nd lambda(max). For rods and 'red' cones in the adult frog retina, we find
no significant difference in E-a between the two pigments involved, althoug
h their lambda(max) values are very different. We also determined E-a for t
he rhodopsin in toad retinal rods--spectrally similar to frog rhodopsin hut
differing in amino-acid sequence-and found that it was significantly highe
r. In addition, we estimated E-a for two pigments whose lambda(max) differe
nce was due only to a chromophore difference (A1 and A2 pigment, in adult a
nd larval bag cones), Here E-a for A2 was lower than for Al. Our results re
fute the idea of a necessary relation between lambda(max) and E-a but show
that the A1 --> A2 chromophore substitution decreases E-a.