xiii, b, i Barber: The Transmission of Malaria 17 



exposed to a crescent carrier on two successive days and on 

 each of the three following days to a benign-tertian carrier. 

 Nine days after the first exposure they were dissected. The 

 mid-gut of one specimen showed 9 oocysts not segmenting and 

 apparently about 9 days old. Sporozoites were present in the 

 salivary glands. In the wall of the gut beside one of the im- 

 mature oocysts apparently mature sporozoites were found. 

 There was some evidence here of a superimposed infection of 

 the same insect. 



Much evidence may be found in these experiments to indicate 

 that the probability of infecting a mosquito depends on factors 

 other than the number of gametes present in the carrier at the 

 time of feeding. In Table VI, under A, we have 22 positive 

 carriers. In fifty-eight feedings with these carriers the per- 

 centage of crescents averaged 14.8 per hundred leucocytes. In 

 the always negative carriers, column C, thirty-four feedings gave 

 an average of 23.2 per cent of crescents. However, 4 of these 

 later feedings were done on insects apparently less robust, 

 since reared from larvae kept for some time in the laboratory. 

 If we omit these four feedings we have an average of 15.1 per 

 cent of crescents for the negative carriers, a percentage still 

 slightly above that of the positive carriers. 



This subject is approached in another way in Table VIII. 

 Here dissections of the controlled lots are summarized with 

 respect to the relation that the percentage infected and the 

 average number of oocysts per infected mid-gut bear to the 

 number of crescents present in the carriers at the times of feed- 

 ing. Crescent percentages are arranged in groups, and as many 

 dissections as could be brought into these groups are included 

 in the table. Some dissections had to be omitted, because the 

 mosquitoes had been exposed to carriers of such widely varying 

 percentages that they could not be brought into any group. 

 Anopheles umbrosns, barbirostris, and sinensis had many nega- 

 tives, and these were nearly all found in two crescent groups ; so, 

 for the sake of comparison, a second series of columns is given 

 in which these species are omitted. 



It will be noted in Table VIII that the percentage infected 

 does not increase with the increase in percentage of crescents, but 

 comparing groups in which there was a fair number of infected 

 specimens, we note a steady rise in the average number of oocysts 

 as the gametes increase. The 75.1-100.0 group is scarcely com- 

 parable, since there were only two infected mosquitoes in that 

 group. 



1522-12 2 



