338 BANKS. 



motion on the part of the observer, or a shadow caused by any object 

 passing near, would send tbe full grown larva? to the bottom, while the 

 smaller ones would hide themselves in the entanglement of the algce. 

 The most practical way of obtaining the larva? was to dip up algce and 

 water in a porcelain plate whereupon each larva, however minute, was 

 at once thrown into silhouette against the white background. 



No difference, either in general or in minute appearance could be 

 observed, even upon microscopic examination, between these larva? and 

 the ones which breed in salt water. The breeding places were in every 

 case exposed to the direct rays of the sun; often no vegetation was 

 present within 10 or 15 meters of the streams. Culicine larva? of a 

 species not yet determined were associated with those of Myzomyia 

 ludlowii Theob. This former species has been found breeding in open 

 fields in the city of Manila in water containing alga?, but has never been 

 seen in salt water at places on the sea where the larva? of M . ludloivii 

 Theob. have been encountered. 



The mosquito larva? were found breeding at Sagada and Bontoc under 

 conditions identical with those at Cervantes, so that the question of 

 altitude with reference to the influence upon the species appears not to 

 enter into the discussion as to why these mosquitoes should be found 

 thriving under such different environments. 



DEDUCTIONS AND CONCLUSIONS. 



I am not aware of any other record of a species of mosquito propa- 

 gating naturally in both salt and fresh water. Previous laboratory ex- 

 periments along this line made by me, both in America and in the 

 tropics, have always demonstrated that mosquito larva? removed from 

 fresh to salt water or vice versa, have perished within a short time, 

 either as a direct result of the difference in media or because of the 

 effect of either fresh or salt water upon their food. No attempt has been 

 made by me gradually to transfer M. ludloivii Theob. from salt to fresh 

 water and it was likewise impossible at the time of my visit to Cervantes 

 to bring living larva?, from there to the coast for experiment. 



There are two theories which might be advanced to account for the 

 different environments of M. ludlowii Theob. One is that the insects, 

 originally salt-water breeders, have gradually been migrating by way 

 of the Abra and Cagayan Rivers from the mouth of the latter up toward 

 the higher regions, adapting themselves' slowly first, to the brackish 

 waters near the sea, then to the tidal fresh waters and finally extending 

 up as far as Sagada and Cervantes. 



The other and more natural deduction, in view of the fact that most 

 mosquitoes are fresh-water breeders, would be that originally all mos- 

 quitoes breed in fresh water. This would appear the more reasonable 

 since very many species breed in isolated water and in forests far inland. 



