Mortality results for experiments with constant pulse energy achieved by proportionally adjusting the power according to pulse duration (25, 50, and 100 ms). Dosing was performed with both the (a) 532 nm and (b) 1,064 nm lasers, both using a 2.5 mm spot size, and set to the approximate LD75 fluence value from the dose–response curves (solid curves with dashed line 95% confidence intervals of logistic regression fit) from the corresponding 25 ms experiments. Error bars on individual points represent 95% confidence intervals of exact binomial probabilities. Statistical differences among mortality at the various pulse durations summarized in Table 3. The slight offsets on the x axis in (a) stem from slight changes in the beam spot size as a function of power.