22 MOSQUITOES 



hundred yards of tlie surf. As it is absolutely isolated 

 from trees, and as the mosquitoes do not appear during 

 a lull in the breeze, but only after several days of land 

 breeze, it appears to him that the popular view of the 

 influence of land breezes upon spreading- mosquitoes sea- 

 ward is not wholly without foundation. 



All the evidence so far adduced refers to more or less 

 normal conditions and, in fact, such conditions are the 

 only ones which need practicall}^ be considered. Under 

 very exceptional conditions, however, it seems perfectly 

 plain that mosquitoes may migTate for considerable dis- 

 tances ; two such instances are described in the following- 

 lines quoted from a recent letter received from the Hon- 

 orable J. D. Mitchell, of Victoria, Tex., a reliable ob- 

 server and a naturalist. 



. . . I have witnessed, in my life, two migrations of mos- 

 quitoes, that will always be fresh and vivid in my memory. 



Where the Colorado River empties into Matagorda Bay, a 

 marsh is formed over the lowlands, by a raft in the river ; this 

 marsh contains about eighteen square miles, and is a magnifi- 

 cent breeding-place for mosquitoes. 



I used to ranch it on the peninsula-like body of land, formed 

 by Carancalma Bay on the east, Matagorda Bay on the south, 

 Keller's Bay and Creek on the west, and the Calhoun County 

 north line on the north (refer to coast maps of Texas), The dis- 

 tance between my ranch property and the marsh above described 

 is, by bay route, about forty miles ; the way the crow flies, about 

 thirty-five miles. My ranch was situated on Carancalma Bay, 

 near the north end of land. 



The first migration occurred in October, 1870. There had been 

 an overflow from the upper country, which tilled the aforesaid 



