HISTORICAL SKETCH 13 



Lahille proposed in 1904 a classification founded on the same characters 

 formerly used, the proboscis and palpi, also adopting some of Theobald's scale 

 characters. He introduced in addition a new character, the narrowing of the 

 wing and the consequent shortening of the forked cells, as indicating specializa- 

 tion for rapid flight. He adopted two tribes, based on the presence or absence 

 of a long proboscis, the Corethrinae and Culicinae. The latter he divided into 

 three groups, founded on the length of the palpi, the Macropselaphes, Hete- 

 ropselaphes, and Micropselaphes. Each of these groups contained two sub- 

 tribes, characterized by wing differences; these differences he expresses in the 

 terms Protopteres and Teleopteres. Thus the Lynchiellina (Lynchiella pro- 

 posed as a substitute for Megarhinus) are supposed to be derived from the 

 Anophelina, the Toxorhynchitina from the Culicina and the Uranotgeniina from 

 the Aedina. 



In the same year Lutz (in Bourroul) offered a new arrangement in which the 

 order of importance of the old characters was shifted. After removing the 

 Anophelinse and Ankylorhynchge (Megarhininse) he, for the first time, em- 

 ployed consistently the character of the presence or absence of setae on the post- 

 notum of the mesothorax, formulating two groups, the Metanotopsilse and the 

 Metanototrichffi. In consequence the character of length of palpi is employed 

 secondarily and repeated in the two groups under the names HeteropalpjB and 

 Micropalpa?. In addition a number of new genera and two new subfamilies, 

 Haemagoginge and Dendromyinae, are proposed. His general scheme was the 

 nearest approach to a natural classification so far reached, and a great advance. 

 It was adopted by Theobald in the fourth volume of his monograph (1907), 

 without, however, coordinating his many previously established genera. 



In 1904 Felt redefined certain genera, largely on the basis of the male geni- 

 talia, and proposed some new generic terms. In 1905, Dyar continued the classi- 

 fication of Felt and proposed some additional names. Early that year Dyar 

 separated the forms with short proboscis as a separate family, Corethridas, and 

 included in it the genus Dixa, the position of which had, until then, been con- 

 sidered uncertain. Independently Eysell, also in 1905, considered the Culicidae 

 and Corethridge to represent separate families and further elevated the Ano- 

 phelines to family rank, as Anophelidaj. The corethrid forms were already 

 excluded by Blanchard in 1900, and the family Culicidae restricted to the forms 

 with piercing mouth-parts; in this he was followed by several authors. It is 

 said that Eondani, already in 1856, established the family Corethridae. 



In 1906 Dyar and Knab published a classification founded upon larval char- 

 acters, the adult not being taken into consideration. A number of new generic 

 and specific names were proposed. The paper marks a departure from the old 

 methods in that the historic adult characters were first definitely abandoned. 

 In the same year Christophers published on the larvae of mosquitoes, however, 

 without a definite classification. He proposed a new genus, Jamesia, for certain 

 predaceous larvae. Mitchell, in a paper on the mouth parts of mosquito larv^, 

 gave great importance to the modifications of the larval mandibles and maxillae 

 and raised the genera Deinocerites and Psoropliora to subfamily rank. 



