220 6 



of sterile normal saline as would make the total amount of fluid in each tube 

 2'5 c.cm. after the due addition of the agglutinating fluid — whether serum or acid — 

 had been made. Each series contained 24, 35, or 57 test-tubes in the different 

 experiments; and the amount of agglutinating substance put into each successive 

 tube diminished regularly from one end of the series (tube No. 1) to the other. 

 Each test-tube therefore contained the same quantity of bacterial suspension, and the 

 strength of the emulsion was the same in every tube; but the amount of aggluti- 

 nating substance added to each was, roughly, *k of that added to the tube imme- 

 diately preceding it in the series, and proportionately greater than the amount 

 placed in the tube following next to it. Thus tube No. 1 contained 10 times as 

 much agglutinating substance as tube No. 12, and No. 12 again contained 10 times 

 as much as tube No. 23; in the long series of 57 test-tubes, No. 1 would contain 

 100,000 times as much agglutinating substance as No. 56. The last tube in every 

 series was kept as a control, containing 15 c.cm. emulsion -f 10 c.cm. of normal saline, 

 and none of the agglutinating substance, thus serving to detect the possible occur- 

 rence of spontaneous agglutination. In many of the experiments the normal saline 

 and the agglutinating substance were measured into the tubes lirst, the To c.cm. of 

 emulsion being added after them, so as to diminish the experimental error due to 

 the difl'erent lengths of time during which the bacteria and agglutinating substance, 

 whether serum or acid, would be acting upon one another in the difl'erent tubes. 

 The actual measurements of the amounts of agglutinating substance and normal 

 saline placed in the tubes were made with a 10 c.cm. pipette divided into 100 parts, 

 while the emulsion was added from a 100 c.cm. pipette divided into 100 parts. 

 The diluted sera and acids were made up with normal saline, except in cases where 

 special mention is made to the contrary effect. 



In certain of the experiments, use was made of agglutinating sera that had 

 been exposed for a longer or shorter time to the action of an acid or of an alkali; 

 and here we found that a new source of error crept in. Taking, for example, a 

 series of 24 tubes, the acidified serum added in quantities varying from 10 to 

 013 c.cm. to tubes Nos. 1 to 11 would, in the ordinary course of the experiment, 

 be further diluted by the addition of 9 volumes of normal saline to it, before it 

 was added, in quantities varying from 10 to 01 c.cm., to Nos. 12 to 23. But it was 

 found that a serum which had been acted upon by an acid or an alkali would 

 exhibit a sudden change in its properties on being thus further diluted ; so that the 

 continuity that might have been expected to appear in the results was liable to a 

 sudden interruption between tubes 11 and 12, at the point where the further dilu- 

 tion of the acidified serum had been made. This etTect was no doubt due to the 

 sudden change in the degree of dissociation of some unknown serum-acid compound 

 formed in the acidified serum ; it was completely obviated by the employment of 

 a 010 c.cm. pipette, divided into 20 parts, for adding the serum to tubes Nos. 12 — 23, 



