1. The Proteus strain XL, isolated by Dr. LIMA in the endemic area at Sao Paulo, represents the serological type of Proteus which corresponds to the local variety of typhus virus. 2. The Proteus strain "Muar", isolated in the Federated Malay States and described by MARTIN, antigenically corresponds to the virus of the local variety of tropical typhus (type XK). 3. The agglutinins for the two types OX19 and OX2, present in the serum from cases of "fie`vre boutonneuse" of the Mediterranean, are of the order of group agglutinins and are due to group antigens present in the virus. 4. It is suggested that the type of the main antigen of the virus of "fie`vre boutonneuse" will remain unknown until the corresponding type of Proteus has been isolated. 5. The relationship between the agglutinogenic and the immunogenic properties of different varieties of typhus virus is discussed. The established facts indicate that cross-immunity between typhus viruses is accompanied by identity of the serum reactions to Proteus X whereas failure to obtain cross-immunity coincides with dissimilarity in agglutination reactions. 6. The present knowledge of serological varieties of typhus, based on the occurrence in the patients' serum of main and group agglutinins for various types of Proteus X is presented in tabular form. It is suggested that serological methods, using various types of Proteus, are indispensable in the experimental study of the typhus group of viruses. 7. The technique of the agglutination reaction with Proteus X strains is discussed. The use of O variants and of the new Oxford standard suspension are recommended. 8. The isolation of Proteus X strains and the hypothesis of the transformation of Rickettsia into Proteus X are discussed and some technical details of the methods of cultivation are described. c 1933.