DOWNS: BACILLUS TYPHOSUS. 157 



to give complete absorption and frequently absorption was not 

 complete in twenty-four or forty-eight hours. After a standard of 

 four days was chosen no more trouble was experienced. It was 

 always necessary to add more organisms as the supernatant fluid 

 became clear; the greater the tendency to agglutinate, the larger 

 the number of organisms necessary for complete absorption. It 

 was necessary to keep the serum at ice-box temperature because of 

 the well-known tendency of diluted serum to deteriorate at room 

 or incubator temperatures. A control of diluted serum which had 

 been incubated under the same conditions as the test sera was neces- 

 sary to show that no drop in titre had occurred, and a control of 

 the serum to be tested saturated with the homologous organisms 

 indicated the completeness of the absorption. Table V gives a 

 summary of the absorption tests. 



From table IV it will be seen that the strains of typhoid differ 

 perceptibly in their agglutinating properties. On this basis we have 

 placed the organisms tentatively into three groups. Group I is 

 made up of eleven organisms; group II of twelve organisms; group 

 III of two organisms. Group I serum agglutinates all other or- 

 ganisms in this group in dilutions practically as high as that given 

 for the homologous organisms. Group I serum also agglutinates 

 group II organisms, but in lower dilutions; conversely, the group I 

 organisms are agglutinated by group II serum, but in lower dilu- 

 tions than are the group II organisms. These two groups are closely 

 related and interagglutinate to the degree indicated in the table. 

 Groups I and II serum give slight or no agglutination with group 

 III organisms. Group III, consisting of two strains, Nos. 2 and 3, 

 interagglutinate perfectly at 1-15000, but this high-titred serum 

 agglutinates members of groups I and ir in low dilutions or not 

 at all. 



The results of agglutination tests using horse serum indicated 

 that the same antigenic differences were present, but that they ap- 

 peared in higher dilutions because of the higher titre of the serum. 



To illustrate: No. 12, the Rawlings strain, was completely ag- 

 glutinated at 1-80000, and No. 1 at 1-5000. 



Many of these agglutination tests were checked by using the 

 microscopic method, care being used to rule out the personal equa- 

 tion. Where partial agglutination occurred, the macroscopic meth- 

 od seemed to give more definite results. 



It will be seen that the absorption tests show an even closer 

 relationship between groups I and II than do the agglutination tests, 

 No. 1 being somewhat more irregular than the others. The ab- 



