The first substance is precipitated from X-19 crude extracts by a relatively low percentage of alcohol and electrolytes (from one to two and a half volumes of alcohol). When purified as far as possible, it gives a negative biuret reaction, a positive Molish and has a nitrogen content of 4 per cent. This material, which we call X factor, has the immunological properties of the common antigenic factor in Proteus X-19 and typhus Rickettsiae, described elsewhere. It has the property of precipitating with typhus serum as well as anti-Proteus serum, even after treatment with hot alkali.

The second substance we call P factor, suggesting a material which is proper to Proteus X-19 and has nothing to do with the Weil-Felix reaction. It is obtained from the crude extracts of B. proteus X-19 by treating the fluids from which the X factor has been removed with an excess of alcohol (seven to ten volumes, according to electrolytes in solution). The purified material shows a nitrogen content of a little less than 1 per cent, gives a negative biuret and a positive Molish reaction. The P factor produces precipitates with anti-Proteus serum in considerable dilution, but has no effect on typhus serum. It is quickly destroyed on treating with alkali, a fact in accordance with the results already cited, which were obtained by White with whole extracts of B. proteus X-19.

The duality of the X-19 extracts seems to be explained by the isolation of two immunologically different factors; one which is alkali-labile and which is proper to B. proteus X-19; and the other which is alkali-stable and is the common antigenic factor in Proteus X-19 and typhus Rickettsiae.