274 
necessarily disappear at all between the relapses, but are generally 
still to be found in small numbers per c.mm. on most of the davs. 
The apyretic periods preceding the thirteen relapses in our cases 
(. 7 > * 7 > *8, 19, 20, 23, 24, 26, 32) lasted 114 days altogether (nearly 
nine days each on the average). On eighteen of these days no 
observations were made; but on the remaining ninety-six days the 
parasites were found in small numbers in 59, or 61 per cent., and 
were not found only in 37. Moreover, the general trend of the 
curves suggests that they were not found on these days only because 
their numbers were a little too few for detection. The parasite 
curve, at its height during a pyrexial period, generally falls very 
rapidly at first and more slowly later, and tends to reach its lowest 
about half way between two apyrexial periods. At this point it 
may or may not remain above the detectable limit (by thick him 
methods). After this it was observed (by D.T.), especially in 
Cases 7, 17, 23 and 24, to begin mounting slowly at first, until 
when it reached the pyrogenic limit, another pyrexial period 
commenced. All this is scarcely compatible with the speculation 
that the apyrexial periods are due to the abrupt death of most of 
the asexual Plasmodia , or to their conversion into ‘resting stages,’ 
etc. j\or do sucli speculations appear to be at all necessary. It is 
easy to see that the survival of comparatively small numbers of the 
asexual forms will suffice to keep the infection alive, not only for 
t ie short periods observed by us, but for ‘ relapses of long interval; 
and for months or years. Obviously, if only a few parasites per 
c.mm. are present, they may easily be overlooked in the small 
amount of blood (say o'l c.mm.) usually examined by thin film, 
may then be reported as being absent: yet many millions may 
1 exist in a patient who contains say 3,000,000 c.mm. of blood, 
r s f )ec ^ at ^ on regarding parthenogenesis in malaria and tire case 
au mn, supposed to support it, have been criticised elsewhere 
y one o us (R.R.).* j n two Q f our re i a p S j^g. cases ^ 2 6) no 
xua orms at all were found during thirty-five different thick 
/ ™ 10nS - ^ ven ^ the y were present in small numbers 
narlT c C ° lleCt in the s P Ieen )> the numbers of asexual 
canrmf ° Un , m ° r JUSt ^ efore relapses (24 and 400 per c.mm., 
__ e explained by parth enogenesis unless we suppose that each 
# Prevention of - - 
