CHOICE OF BREEDING PLACES 221 



difficult of examination, since the plants for the most part grew upon the forest 

 trees thirty feet above the ground. Those that had fallen with the trees had lost 

 their water. Later, however, many plants were found in places which could be 

 reached, and in the water accumulations of their leaves were found breeding a 

 number of species of mosquitoes, and among these the larvae of Anopheles cruzii. 

 He considers the species to be a typical forest insect breeding only in such places, 

 and that it is responsible for the malarial attack of a mild type from which the 

 railway constructionists had suffered. The genera of Bromeliacese most com- 

 monly met with are said to be Bresia, Nidularium, BiUbergia, ^chmea and 

 Bromelm. In the water in these plants, he found bits of vegetation and a humus 

 formed from them, and living in it small crustaceans, tadpoles, larvae of Culi- 

 cidae and Chironomidse. Among the other mosquitoes inhabiting this bromelia- 

 water he found Megarhinus violaceus which was feeding upon the other culicid 

 larvas. The relation of Anopheles cruzii to this outbreak of malaria was not 

 positively established. 



On the supposition that Anopheles punctipennis is not a carrier of malaria, the 

 control of malaria in certain regions of the United States, through the destruc- 

 tion of the Anopheles larvae or breeding-places, may be simplified if a difference 

 in the character of the breeding-places of A. punctipennis and A. quadrimacu- 

 latus (the only other common inland species in the more northern United 

 States) can be ascertained. This consideration led Jordan and Hefferan {loc. 

 cit.) to carry out the following observations : 



" In the present instance it was found that the favorite breeding-places of A. 

 maculipennis * and A. punctipennis, although close together, were of quite dif- 

 ferent character, as shown by the following facts relating to the distribution in 

 Western Michigan. 



" Dipping for mosquito larvae was carried out in the springs and spring-fed 

 pools of the northern ravines, along the river shore, in the bayou, and in rain- 

 v.-ater barrels of the village. In all of these places, with the exception of the 

 bayou, Anopheles larvae were found in abundance at some time during three 

 consecutive summers. The distribution was more extensive in 1902, merely 

 because the early part of the summer was wet, and water stood longer in the 

 springs and ravines. In 1903 the early part of the summer was dry, so that by 

 the middle of August two springs and the river shore only were left as breeding- 

 places in the immediate vicinity of the village. In the river the larvae of A. 

 maculipennis * are found regularly at certain places along the north shore, 

 where abundant food and quiet are ensured by the wreckage and masses of river 

 weed which lie a few inches below the surface of the water. Here, with the river 

 running almost due west, the larvae are exposed to direct sunlight. They were 

 not found along the more shaded south shore. A curious instance of choice of 

 breeding-places occurred in the summer of 1904. A small stream running to 

 the river had, during the spring, a course of a mile or so down the ravine, but 

 by August 1 it had dried to a few pools of the following character: (a) River 

 inlet, 40 feet long and 7 feet wide, shaded by willows, bottom sand and mud, 

 no larvje; (b) Ten feet above this inlet a pool 40 feet by 4 feet in area, some 15 

 inches deep, entirely without shade or vegetation, no larvae; (c) Only three feet 

 from this second pool another pool of clear water, 5 feet by 3 feet in size, and 6 



*— Anopheles quadrlmacul^tus of this work. 



