KATIN TEST 171 
minimum amount of salt soluti(m l)e used in ])re])arin^ the antiji;en 
suspension. 
Choice of an Antigen. — In attempting- to prepare an antigen of suffi- 
cient concentration of lipoids, experiments indicated that powdered 
heart muscle instead of wet heart muscle should be employed in the 
alcoholic extraction. It was further observed, after comparative 
studies with different alcoholic extracts, that a Wassermann antigen 
proposed by Neumann and Geiger, gave promising precipitation results 
when used with syphilitic serum. The method of preparation of this 
antigen consisted of first removing the ether-soluble substances from 
dried beef heart by extraction with ether and then making an alcoholic 
extract of the same beef heart. After devising a quantitative method 
for its preparation, this alcoholic-extract antigen formed the basis for 
developing a practical precipitation test. 
With an antigen prepared from a given lot of heart muscle, it was 
found that the most sensitive reactions are obtained when the lipoid 
concentration in the antigen is adjusted to a certain, optimum, deter- 
minable value. Increasing or decreasing the concentration from this 
optimum reduces the sensitiveness of the precipitation reactions. As 
will be seen later, this observation made possible the continued use of 
an antigen of a uniformly standard degree of sensitiveness in the 
Kahn test. Briefly, if an antigen is more sensitive than standard 
requirements, it may be corrected by either one of two methods: 
dilution with alcohol or the addition of a sufficient excess of antigenic 
lipoid material. In practice the former method is adhered to. Under- 
sensitiveness in an antigen, when associated with over-concentration 
of lipoids, is corrected by dilution with alcohol; when associated with 
deficient concentration, it is corrected by the addition of lipoids. 
Instability of the Antigen Suspension.— Under proper conditions it is 
possible to mix an antigen with physiological salt solution so as to 
produce a suspension of lipoid particles of such instability that they 
will become immediately dispersed on contact with additional salt 
solution or serum. This is the type of antigen suspension used in the 
Kahn reaction. When this suspension is mixed with syphilitic serum, 
the lipoid particles become immediately dispersed and a stable pre- 
cipitate appears within a minute or two. When mixed with non- 
syphilitic serum, no precipitate appears. 
Quantitative Relation Between Antigen Suspension and Serum.— 
Optimum precipitation results are obtained only when the number of 
reacting units in the antigen suspension approximates the number of 
reacting units in the syphilitic serum. If the number of antigen units 
in relation to serum units is either very small or very large, no precipi- 
tation will take place. In a less marked degree the same holds true 
if the number of serum units in relation to antigen units is either 
excessive or deficient. Based upon clinical and serological studies, the 
following three proportions of serum and antigen suspension appeared 
desirable in a practical test: one part of serum to one-third part of sus- 
