xiii, b, i Barber: The Transmission of Malaria 7 



but when both are found together in ponds they appear in- 

 distinguishable to the naked eye. When occurring in ponds, 

 in the younger stage both much resemble A fuliginosus 

 macroscopically. 



Type Giles appears to be rather common in the Federated 

 Malay States. It was found by me in seven different places 

 within 6 or 7 kilometers of Kuala Lumpur. In one of these 

 places — certain fish ponds — they could be collected in large 

 numbers throughout a period of several months, and they were 

 abundant in two other ponds. I also found them in fresh water 

 at Tronoh, Perak. In brackish water many were collected in 

 some pools well supplied with algae at Port Weld, Perak. The 

 brackish-water larvae seemed a little darker in color and the 

 adult slightly darker than in the fresh-water type, though prob- 

 ably this was only a slight local variation. The occurrence of 

 this type in brackish water, the ordinary breeding place of 

 A. ludlowi in Malaya, is noteworthy, since the larva of type Giles 

 and that of ludlowi appear identical. It is evident that adult 

 as well as larval stages are necessary in distinguishing these 

 forms. 



SUSCEPTIBILITY TO PARASITES OF MALARIA 



In comparing the two types of rossi in regard to susceptibility 

 to malaria, five pairs of caged anophelines were fed on gamete 

 carriers, one member of a pair type Giles, the other, type inde- 

 finitus. All specimens were examined in the larval stage and 

 the types separated before emergence. Each pair of cages was 

 fed at the same time and on the same gamete carriers. In two 

 of the pairs both types were collected from the same pond and 

 at the same time. The members of a pair are almost exactly 

 comparable, except that in two pairs some mosquitoes that 

 emerged late were introduced into cages after the cages had 

 been once exposed to a carrier. The majority of those intro- 

 duced later were of type Giles, so that any error from this 

 source, if such exists, would tend to reduce the percentage in- 

 fected of type Giles. In every pair but one type Giles showed 

 at dissection much the higher percentage of infected mosquitoes. 

 In the exception 3 out of 4 indefinitus dissected were infected 

 and none out of 3 type Giles. The results of this comparison are 

 summarized in a small table appended to the bottom of Table II. 

 It will be seen that of type Giles 37.2 per cent were infected 

 of 94 dissected, while of type indefinitus only 11.0 per cent were 

 infected of 73 dissected. Type Giles gave on the average twice 



