2 77 
week or more after the first attack of fever. Strong confirmation is 
also given by Case 6, in which Dr. Fantham noted a sudden 
appearance of sexual forms (P. vivax) on the i ith day, seven to ten 
days after the original asexual rise and after the patient had been 
taking thirty grains of quinine daily for a week—suggesting that 
the same phenomenon holds for P. vivax also. 
Other observations are as follows:—(i) By no means every 
asexual rise is followed by a sexual one. In seven cases (13, I 9 > 
25, 26, 28, 31, 33) no crescents at all were seen. In Case 13, a 
negro, 173,400 asexual forms per c.mm. failed to produce any; 
and none was observed in another negro (Case 28). (2) I he number 
of crescents never exceeded five per cent, of asexual forms found 
in the generation which (hypothetically) produced them, and were 
generally in much smaller proportion. (3) We could find no 
relation to youth, vigour, amount of haemoglobin, duration of 
infection (from three to thirty weeks), or season of first infection. 
(4) The daily counts of crescents generally show marked daily 
variations, but we think that after considerable doses of quinine 
have been given for some days these daily variations tend to be 
smoothed out (Cases 16, 18, 22, 24). (5) In Cases 15, 14 and 16, 
particularly, a distinct tertian tendency was observed in the daily 
variations of crescents. This appears not to have been previously 
noticed, but is just what would be expected in the case of a tertian 
parasite—though, of course, the curve would be confused in a 
double tertian. (6) Sometimes the crescent curve suddenly rises 
with great rapidity, reaches its maximum, and then begins to fall 
on the next day with equal rapidity, though, later, the fall often 
tends to be much more slow (Cases 14, 1 5 » 22 > 2 3 )> especially when 
the numbers are very small. Apparently, the greater the maximum 
the quicker fall (Cases 14, 15); and the fall is sometimes jagged 
with a tertian tendency (ibid.). (7) Sometimes the maximum is 
irregularly maintained for a number of days (Cases 16, 18, 24), 
especially when quinine has not been previously given. (8) The 
crescents remained detectable for as many as 31, 32, 32, 35 and 44 
days in Cases 20, 22, 23, 24 and 18 respectively; and finally 
disappeared (in 1 c.mm. of blood) in Cases 1, n, 12, 16, 21 and 23. 
From these data we think (1) that a varying percentage of 
asexual forms (sometimes very few or none) are constantly 
