(1) Meningococci from 60 cases of cerebro-spinal fever have been submitted to serological tests and compared with meningococcus-like micro-organisms from the naso-pharynx of 71 normal persons. In these serological tests 19 monovalent anti-meningococcus of known pathogenic origin and eight with strains isolated from the naso-pharynx. The tests comprised careful estimation of the agglutinability of these 131 strains with the 19 different sera and also of the extent to which these strains absorbed and removed the agglutinin from one or more sera.
(2) The simple agglutination reactions effected a rough sub-division of these 131 strains into two groups, Group I and Group II, and indicated that in Group I there was a main group comprising the majority and at least two smaller groups containing a few strains only.
(3) Tests for absorption of agglutinin confirmed this rough subdivision into two groups and distinguished with much greater precision the different main groups and smaller groups. In general, each of these groups was found to differ in that they did not absorb the agglutinin for members of the others, and members of each produced sera from which members of the others similarly failed to absorb agglutinin.
(4) Two of the smaller groups, however, represent two “scrap-heaps” of strains unidentified by absorption reactions but suggesting by their agglutination reactions that they belonged to Groups I and II respectively. In the former there were placed three strains of cerebro-spinal and one strain of naso-pharyngeal origin, and in the latter four strains of the former origin and thirteen strains of the latter.
(5) The other main groups and smaller groups contained representatives of both cerebro-spinal and non-contact naso-pharyngeal strains, though not in equal numbers: in the main group of Group I there were placed 24 spinal strains and only one non-contact naso-pharyngeal; in the main group of Group II there were placed 16 spinal strains and 40 naso-pharyngeal, of which 23 were from non-contacts and 13 from contacts.
(6) Variations in agglutinability and absorptive capacity were shown to be so great as to interfere seriously with the use of serological tests for identifying meningococci in practice.
(7) An additional difficulty affecting both the identification of meningococcin and their classification into types or groups and sub-groups is that, even with seven sera, each corresponding to a different group, strains were found, both spinal and naso-pharyngeal, which failed to react typically with any and therefore could neither be ascribed to a particular type or group nor be identified on serological grounds with the other pathogenic strains.
(8) Hence it was concluded (a) that it is impossible to regard these types or groups as representing distinct classes limited by hard and fast lines, and (b) that it is unsafe to exclude any strain from possible pathogenicity on the ground of its failure to agree serologically with any of such sets of sera as are likely to be available in practice, since even the large series I employed failed to include all pathogenic strains.
(9) The conclusion in my first report is therefore maintained that any strain possessing the admitted morphological and cultural characters of the meningococcus should be regarded as potentially pathogenic without considering its serological reactions.
(10) I have found, however, that the strains obtained from the majority of carriers show by absorption tests complete serological identity with known pathogenic strains. To this criterion 58 of my 71 naso-pharyngeal strains conform. These were distributed as follows: 16 out of 138 out-patients attending Lambeth Infirmary in June and July, 1915 (11·6 per cent.); 12 out of 57 non-contact soldiers in February, March and May, 1916 (21 per cent.); 13 out of 24 soldiers in March, 1916, who had had no direct connection with cerebro-spinal fever but in whose neighbourhood two cases had occurred (54 per cent.); 17 out of 62 soldiers who had been in direct contact with cases of the disease (27 per cent.).
(11) Inclusion of those persons who were found to be carrying strains not fully identified serologically did not raise the percentage of carriers to a significant extent.
