In 1928 Frederick Griffith, British microbiologist, made a series of unexpected observations while performing an experiment with the disease-causing bacteria pneumococcus and laboratory mice. Griffith's experiment dealt with two strains of the bacteria pneumococcus. One was a virulent strain with a smooth polysaccharide coat necessary for infection and colonies of this strain appear smooth. The other was a non-virulent strain with a rough coat that could not cause infection and colonies of the strain appear rough.

Griffith injected one group of mice with the smooth virulent strain and these mice died after a few days. He then injected another group with the rough non-virulent strain and these mice continued to be healthy. Griffith took a heat-killed strain of the virulent bacteria and injected it into mice and observed that they did not die. Griffith's fourth experiment was to inject heat treated, killed, smooth virulent strain mixed with the non virulent rough strain. He injected this mixture and found that after a few days the mice died. The blood of the dead mice showed high levels of virulent pneumococcus. Griffith theorized that some type of transformation takes place from the virulent to the non-virulent strain for it to synthesize a new polysaccharide coat.