278 
generating crescents, which, after about eight to ten days, appear 
in the peripheral blood. Hence, when the asexual forms remain 
sufficiently numerous for a period, as when no quinine is given, the 
stock of crescents is being constantly replenished. Thus the 
mortality among the older crescents is constantly being compensated 
for by new arrivals, and the total numbers appear to remain 
constant—an appearance which has given rise to the generally 
accepted hypothesis (perhaps erroneous) that the crescents survive 
for weeks. On the other hand the sudden rise and fall of their 
numbers seen in Cases 14, 15, 22 and 23, suggest that they really 
survive only for a short time after their appearance in the peripheral 
blood. In Cases 16 and 24, however, a few crescents were still 
found after quinine had been given continuously for sixteen to 
eighteen days, suggesting that some of them may be able to survive 
foi longer periods. In Cases 20, 22 and 23, they appeared in 
increasing numbers for four to six days after continuous 30-grain 
closes of quinine had been commenced, suggesting that the dmg 
has no effect upon them when they have once been generated. But 
quinine appears to affect the numbers of crescents indirectly by 
cutting off the source of supply. Though many researches upon 
the effect of quinine on crescents (a subject of the greatest sanitary 
importance) have been attempted, so far as we know the possibility 
of a constantly continued supply has not previously been taken into 
consideiation. We are continuing researches on the point. 
d The leucoc ytes were counted frequently, but not always 
ai y, at first by the haemocytometer, but in the later twenty 
cases y the thick film method. The latter method has the 
vantages of putting the leucocytes closer together and of 
but T'f US t0 CStlmate tllem at the same time with the parasites; 
for’ rrr ^ haemoc ytometer, it is not very appropriate 
Hl - e y entlal counts because of occasional uncertainty in 
fin ,^ U1S ln £ the various kinds. Accordingly these were made 
occasion CaSCS/ ^ ^ ms > 5 °o cells being examined on each 
iesu ts are, as shown in all the cases, that during the pyrexial 
low 11C t0tal leucoc y tes comparatively few, being frequently 
' as 2 to 3.000 per c .mm. (Cases 17, 18, 23, 24, etc). But as 
Periods 
