( 784 ) 
but f(V) or f(a), as instead of all possibilities only the most probable 
distribution is considered. When, however, e, and «, are f(a), the 
. ; de d(e, He, +8,) 
above reduction is not allowed, since there 7 Ln 5 "has 
a a 
been put. 
So it seems to me that two errors have been made in this reduction, 
which have yet made it possible for us to arrive finally at the 
correct result. First of all ¢, and ¢, have been taken constant over 
the ensemble, which would have the result that A would be the 
: deren 
same as if ¢, and «, did not exist, because Aen is independent 
a 
of «, and «, for a certain phase. It is just in consequence of the 
change in phase-distribution which they cause, that «, and e, have 
Eg 23 
influence. In the second place e 7% and e 7 should now have been 
ARO f 
put as constant factors before the sign rae in consequence of which 
„ga 
we should, in fact, have again obtained the same value as without 
es, and «,. At all events I think I may be justified in calling in 
question the strictness with which the method has been applied up 
to now. 
Physiology. — “Contributions to the study of serum-anaphylaais.” 
By Dr. J. G. SrreeswijK, Foreign Member of the Pasteur 
Institute at Brussels. (2.4 Communication). (Communicated by 
Prof. C. H. H. Sproncx). 
(Communicated in the meeting of February 27, 1909). 
To experimentally bring about the phenomenon of serum-anaphylaxis 
(Ricuet) — which with other related forms of changed power of 
reaction of the organism is denoted by the general name of Allergy 
(v. PirQuer) — consequently at least two serum-injections are neces- 
sary. For the first, the sensitizing injection, spores of serum are 
sufficient, whilst, however, at the end of the incubation-stage a larger 
dose is wanted to bring about intoxication (which, however, is 
different again, according as the injection is given intraperitoneally 
resp. subcutaneously, or into the circulation resp. into the brain). 
It has of course been asked, whether these two functions, the 
sensitization and the intoxication, have to be attributed to the same 
material or to two different constituents of the serum. Now without 
