The first generation of rats received an average of over 160 shocks each before learning to avoid the illuminated gangway. Each successive generation learned quicker than the previous one until, after 30 generations of rats were making an average of only 20 errors each. McDougall concluded this to be evidence for the inheritance of acquired characteristics, but his conclusion was extremely controversial and flew in the face of the orthodox theory of inheritance based on Mendelian genetics because the control animals also learned the behavior without ever being exposed to it! Some of the leading biologists of the time subjected his experiments to critical scrutiny, but they were unable to find any significant procedural flaws in his experiment. When they suggested that McDougall must have been breeding from the more intelligent rats in each generation he designed a new experiment in which he selected only the most stupid rats in each generation as the parents of the next. Thus, by our conventional view of genetics, subsequent generations should have learned more and more slowly. However, the reverse occurred. After 22 generations, the rats were learning 10 times faster than the first generation of stupid ancestors.