7 221 



using, of course the same serum as had already been added (from a lOc.cm. pipette) 

 to tubes Nos. 1 — 11, without further dilution. 



The test-tubes in each experiment were all well shaken up, so as to mix their 

 contents thoroughly, as soon as the appropriate quantities of emulsion, normal 

 saline, and agglutinating solution had been measured into them. They were then 

 placed at once in an Ostwald's water-bath at 37° for from I'/a to 2 hours, when 

 they were removed and the measurements of the agglutination that had taken place 

 were made. When it was necessary, owing to slowing of the agglutination and pro- 

 traction of the time taken by the reaction, the measurements were made a second 

 time after the tubes had stood for a further 18 hours at the temperature of 

 the room. 



The actual measurement of the degree of agglutination attained, one of the 

 most important questions in researches such as these, was made by Madsen's 

 method, which gives admirable results in cases where it can be applied. The 

 method consists, as is well-known, in the arbitrary selection of a single moderately 

 agglutinated tube in one of the series as a standard, and then noting which tube 

 in each of the other series is agglutinated to just the same degree by direct com- 

 parison. This can be conveniently done as a rule; but in cases where the size or 

 the character of the flocculi of agglutinated bacteria vary in the different series of 

 tubes, it is impossible to take the measurements bj' Madsen's method. 



On such occasions the measurement was set down as standing at the tube 

 containing the minimum of agglutination uisible to the naked eye: it was necessary 

 to use this method, for example, in many of the experiments upon the influence 

 of heat on the agglutinability of bacteria. 



To conclude with a brief estimate of the value of these two methods of mea- 

 suring agglutinating power, it must be emphasised that in a considerable number 

 of cases they lead to discordant results, giving values that differ relatively as well 

 as absolutely; so that given a number of series of tubes, the amounts of agglutinat- 

 ing substance needed to produce the smallest amount of agglutination macroscop- 

 ically visible will bear no constant relation of the amounts present in the tubes 

 that shew the same degree of agglutination when the measurements are made by 

 Madsen's method. In some of the series the two tubes will be close together, in 

 others they will be at some distance from one another ; in the former case con- 

 taining somewhat similar amounts of agglutinating substances, in the latter amounts 

 that are widely different. The explanation of these discrepancies we are unable to 

 give; they appear to depend upon a variety of conditions that will not be further 

 particularised. Which of the two measurements is to be accepted as the correct 

 one is really a matter of opinion; but care must be exercised in accepting as final 

 the results that may be deduced from the application of either of them alone, with- 

 out reference to the other. 



