IDENTIFICATION OF BACTERIAL SPECIES 3 
Wiesbaden Congress {be) in April, and having more cases at his disposal, was able to publish a 
communication on serum diagnosis early in July* {14a). 
Pfeiffer's reaction is observed by injecting a mixture of cholera vibrios, together with a 
small quantity of serum, from an animal immunized against cholera into the peritoneal cavity of 
a normal animal. By withdrawing drops of fluid from the abdomen at different intervals it can 
be seen that the micro-organisms first lose their motility, then become transformed into spherules, 
and finally disappear. The rapidity and completeness of the reaction depends on various factors 
{e.g., potency of serum, extent of dilution, etc.). It is not so easily performed with typhoid bacilli. 
Pfeiffer thought vital action essential, but Bordet found that the same phenomena could 
be observed outside the body if the temperature were kept at 37" C. He generally performed 
the reaction with a mixture of bacilli, ' immune,'t and normal serum ; but he recognized that it 
could occur well with only ' immune ' serum, and also, altliough generally more feebly, with only 
normal serum from various sources. Bordet laid chief stress on the granular degeneration of 
the micro-organisms rather than on the agglomeration into groups. 
Durham and Gruber examined the phenomena much more thoroughly, and extended 
it to several kinds of micro-organisms. They found that the addition of a small quantity of 
' immune ' serum to an emulsion of the corresponding bacillus in a test tube caused a precipitation 
of the bacilli in the form of fine granules, the supernatant liquid becoming clear. In fact, the 
reaction resembled an ordinary chemical reaction occurring slowly. A control tube containing 
only bacilli would remain uniformly turbid. The rapidity and completeness of the reaction were 
found to depend on the same causes as in Pfeiffer's reaction. If the same experiment is per- 
formed with smaller quantities, so that the process can be observed microscopically, under suitable 
conditions the following stages are seen. The bacilli, at first isolated and moving rapidly about 
and across the field, gradually slow down and begin to Iiang on one to another ; groups of three 
or four are thus formed which then adhere to other groups, until, if the reaction be complete, all 
the micro-organisms are collected into large clumps, with total loss of movement. Should the 
serum be very potent and not sufficiently diluted, agglomeration and loss of movement may occur 
instantly, so that by the time the slide is brought under the microscope the reaction is already 
completed. On the other hand, with a weak or much diluted serum, the phenomena may 
remain permanently incomplete, so that several isolated micro-organisms are still seen, and even 
some of those in the clumps may retain their motility. So delicate, however, is the reaction 
that serum from a very highly immunized animal will produce an effect on its corresponding 
bacillus even when diluted 500,000 times. The action is to a considerable degree specific (or, 
as Durham prefers to term it, special), for a 'cholera' serum will not act on typhoid bacilli, and 
' typhoid ' serum will not act on coli bacilli. On the other hand, ' cholera ' serum will affect to 
some extent certain vibrios which are not cholera, and ' typhoid ' serum will agglomerate 
* Widal has persistently ignored the dates given in my articles, and consequently represents me as having merely confirmed 
his results. As a matter of fact, so far as I am concerned, the invention of serum diagnosis was made independently, possibly 
earlier, and arrived at from an opposite point of view. 
f For the sake of brevity, 'immune,' 'typhoid,' 'cholera,' etc., scrum will be used for the serum of an immunized 
animal, for the serum of an animal immunized against cholera, typhoid, etc., respectively. 
