2 73 
These figures, together with the numerous ones given in the table 
of cases, show a very marked correlation, but there are, of course, 
many minor deviations. In numbers of the specimens there were 
certainly two sets of parasites of different ages, which were often 
counted together, though only one set could have been concerned 
with the fever. It is probable also, though by no means certain, 
that the resistance to the toxin of the Plasmodia varies, not only in 
different persons, but in the same person at different stages in the 
course of his infection and under different physiological conditions. 
Our new methods will, we hope, enable us shortly to undertake 
very detailed studies on these points. 
;. The -pyrogenic limit. It is quite evident that if the parasites 
fall below a certain limit they are no longer numerous enough to 
cause fever. Thus with P. vivax, an average of 125 parasites of 
all ages and forms counted during sixty-eight days in eight cases, 
and with P. falciparum an average of 460 asexual forms counted 
during eighty-three day in nineteen cases, failed to produce fever. 
The actual limit above which they become pyrogenic probably varies 
in different cases. With P. vivax as many as 1,500, 852 and 540 
were found in Cases 2, 6, and ; without fever, and as few as 150 
and 50 in Cases 5 and 9 with very' slight lever (986 F.), so | at 
about 200 to 500 may perhaps be taken as the usual limit. it 1 
P. falciparum , as many as 1,620 (asexual) were found as an average 
of four non-febrile days in Case 26, and 1,196 as an average o 
six non-febrile days in Case 25; while in Case 18, from 16 to 1,8 o 
were present for twenty-two days without causing any mar 
fever, so that 600 to 1,500 may perhaps be adopted as the usua 
limit. With P. malariae (Case 1), an average of /9 were present 
for thirteen days without fever, but about 140 twice caused sligi 
rigor. 
8. The asexual forms between the relapses . It is generally 
thought and taught that these forms tend to ‘ disappear between 
the relapses—though of course exceptions, especially in c 1 ren, 
are noted—and to reappear during them; and several hypot eses, 
such as that of parthenogenesis (F. Schaudinn) have been advanced 
explain the supposed phenomenon and have led to what appe 
be much waste of work. We find that the asexual forms do no 
