614 BIOLOGY OF PNEUMOCOCCUS 



revealing potential or latent functions not active or even suspected 

 when the same cells are grown in ordinary media. Toxins as such 

 may not be elaborated by pneumococci themselves in an artificial 

 substrate but may possibly be a secondary product of the interac- 

 tion between the bacteria and the cells of the host. The question 

 is by no means closed. 



SEROLOGICAL TYPES 



The present division of pneumococci into thirty-two distinct 

 serological types may not be all-inclusive. Further study of strains 

 at present considered as unclassifiable and continued search of 

 fresh material will undoubtedly add other specific types to the 

 list, although the operation of the law of diminishing returns is in- 

 evitable. 



There is increasing need for a central source of type-specific 

 diagnostic serum to ensure the accuracy and uniformity of type- 

 determination throughout this and other countries. The problem 

 is largely an administrative one but the requirements come from 

 bacteriologists, immunologists, and epidemiologists. The move- 

 ment to establish such a center is to be strongly encouraged, and 

 might well be included in the deliberations of the Health Organiza- 

 tion of the League of Nations. 



DISSOCIATION 



A reconciliation of the many descriptions of pneumococcal vari- 

 ants is greatly to be desired. Modifications A to Zed, the bizarre 

 forms, the mucoid, smooth, and rough variants should be sorted 

 into a uniform, descriptive grouping, and the designations, if pos- 

 sible, should be made to conform to those of similar variant forms 

 of other bacterial species. 



The unknown principle in the virulent pneumococcal cell which 

 has the remarkable ability not only to restore a degraded form to 

 its original condition of type-specificity, but which also can actu- 



