SINCE earlier evidence (refs. 1–3 and Medawar, P. B., personal communication) suggesting a serological reactivity for deoxyribonucleic acid (or the deoxyribonueleic acid moiety of a nucleoprotein) has been indirect or controversial, the problem has been re-investigated with the aid of a preparation rich in deoxyribonucleic acid with transforming activity4,5, isolated from B. abortus cells by extraction with 0.5 per cent phenol in citrate–saline. The extracted material was ‘deproteinized’ by chloroform and amyl alcohol treatments6, but the final material still contained, in addition to typical deoxyribonucleic acid5, approximately 20 per cent protein that could not be dissociated by conventional procedures. Such preparations of deoxyribonucleic acid were used for immunizing rabbits, and the resulting antisera were tested first for precipitating antibodies against homologous deoxyribonucleic acid by the gel diffusion method7.