276 
without quinine, in Cases 5, 17 and 26; while, on the other hand 
small numbers of asexual forms remained in some cases for five 0: 
more days in spite of considerable daily dosage. In Case 23 a 
severe relapse began six days after the stoppage of quinine tannate, 
which had been given for nine previous days in ten to thirty grain 
doses daily. No other relapses occurred after continuous quinine 
treatment; but twelve occurred without it. 
10. Observations on the sexual forms. It is, of course, 
generally held—and probably quite rightly—that these are 
developed (by a cytological process not yet clearly seen) from the 
asexual forms. But at the same time no one has been able to note 
any correspondence between their numbers—many of the former 
may be present when the latter are very scarce, and vice versa. This 
lule, which is fully confirmed by our figures, has always been 
difficult to reconcile with the accepted theory of origin. On 
comparing our curves, however, the remarkable fact was observed 
(by D. T.) that there often seems to be indeed a 
correspondence, but that it is delayed for from eight to ten 
days that is, that the sexual curve tends to rise that number 
of days after a rise in the asexual curve. Out of our twenty-five 
infections with P. falciparum eighteen showed crescents, and of 
thebe cases, 14, 22, 17, 18, 20, 23, 24 and 30 suggest this 
phenomenon, while in 17 and 20 the rises are sufficiently isolated to 
show it most distinctly. In Case 17, asexual rises existed on the 
ist~3rd days, the nth-i2th and the 2ist-22nd days, separated by 
apyiexial periods during which few asexual and no sexual forms 
were found; but on the 30th day (eight days after the last relapse) 
crescents appeared for the first time. In Case 20 an asexual relapse 
commenced on the 17th day and reached its height on the 20th day. 
A very few crescents had been found all the time; but a marked 
ise in their numbers commenced on the 26th day and reached its 
t on the 28th day. The most ready explanation is that the 
crescents require eight to ten days for development , during which 
ley remain undetected, probably in the internal organs of the host, 
s is confirmed by the fact that numerous small ones are seen at 
madT^ enCement ° farise; “ d aIso b y the observation (frequently 
at m fresh infections the crescents seldom appear until a 
