xin, b, i Barber: The Transmission of Malaria 9 



only a study of considerable numbers of the two types will 

 show. 



Of the types of rossi included in the table, A. rossi var. inde- 

 finitus and A. rossi type Giles refer to type indefinitus and type 

 Giles compared above. "Known" means that forms were iden- 

 tified by examination of both larva and adult, and "probable" 

 refers to such specimens as were identified by adult characters 

 alone taken in connection with the character of the breeding 

 place. The coast indefinitus all came from brackish water, and 

 the character of the adult, of the breeding places, and of such 

 larvae as were examined makes it practically certain that they 

 were all of the indefinitus type and not mixed with type Giles. 

 This is the more likely, since practically all came from one 

 locality where type Giles was not found. The inland "probable" 

 is also almost surely of the indefinitus type, since all were col- 

 lected in the small muddy pools where, as stated above, type 

 Giles was never found. The Giles inland "probable" is less 

 certain, but of a large series of larvse subsequently collected 

 from the same habitat, practically all were of type Giles on 

 microscopical examination. The mixed group needs no comment. 

 The less certain forms were included, since all were rossi. 



No other species needs special comment except hunteri, a form 

 recently described by Strickland s and closely allied to A. sepa- 

 ratus of Leicester. 



The columns in Table II under dissection of salivary glands 

 include practically none dissected before the tenth day after 

 feeding on the gamete carrier. The percentage with infected 

 salivary glands is based only on specimens with infected gut. 

 Practically every specimen with sporozoites in the salivary glands 

 had oocysts, empty or otherwise, in the mid-gut, and it is believed 

 that the percentage of gut-infected specimens that showed spo- 

 rozoites in the glands gives a more definite idea of the tendency 

 of a species once infected to form sporozoites than would a per- 

 centage based on all dissections, many of them negative. From 

 the percentage of dissections with ova well developed (the last 

 column of the table) , dissections of specimens caught in the adult 

 stage are excluded, a large percentage of which already had ova 

 well formed at the time of first feeding. The rossi ty$e»indefini- 

 tus and type Giles compared in the small table appended at the 

 bottom are included in the body of the table as well. 



3 Indian Journ. Med. Research, 4, 2. 



