Contribution of H. B. Barlow - Duplicity Theory of Vision - page 171

demonstrated that variation in the test parameters that produced
marked changes in threshold level during dark adaptation produced
the same changes in threshold level under light-adaptation conditions.
These results, then, passed the test of equivalence put forward by
Stiles and Crawford ( 1932 ).
Yet, Barlow found an even more striking confirmation of the
noise theory in the results obtained by Barlow and Sparrock ( 1964 ).
They had shown that when the brightness of a background light
under stabilized conditions matched the brightness of the positive
afterimage obtained at any particular moment during long-term
dark adaptation, the incremental threshold level of this background
brightness was the same as the threshold level measured during dark
adaptation at that moment. This result, of course, nicely fitted the
view that the positive afterimage represented 'dark light' equivalent
to real light with regard to adaptation.
22.4 Opposing evidence
Barlow ( 1972 ) also considered evidence that opposed his adapta-
tion theory. He found that although some of this evidence might
be surmountable, important objections remained. Firstly, Rushton
( 1965a , b ) in his so-called decisive experiment had demonstrated that
bleaching the retina with an array of small light dots and adapting the
retina with a similar array of luminous dots produced vastly different
adaptation effects, indicating that there was a marked difference
between the spatial spread of the desensitizing influence of real and
'dark' light.
This suggestion was further strengthened by Westheimer ( 1968 )
who found that real but not 'dark' light could produce the sensitizing
effect first demonstrated by Crawford ( 1940 ). Thus, he found no
evidence of sensitizing interaction between test spots and annuli
composed of real and 'dark' light, respectively.
Also, opposing electrophysiological evidence had been forth-
coming. Thus, Cone ( 1963 /1964) had found that the latency of the
b-wave of the rat electroretinogram was not affected equally by