CLIMATE AND DISTEIBUTION" 293 



this area, perhaps here and there and not always in the same place every winter. 

 From these foci the summer dispersal is no doubt much facilitated. 



The temporary distribution is determined by the means of carriage that hap- 

 pen to be available. The insects do not fly far, as we have shown, but are readily 

 carried in conveyances. Thus the species may continue to exist for a time 

 wherever it may be carried while the temperature remains about 80°. 



These facts have not been generally taken into account in discussing the distri- 

 bution of Aedes cdlopus, and so the temporary and permanent habitats of this 

 insect have not been separated. It is impossible to do so from records of cap- 

 tured specimens only, consequently these distributions can not be exactly 

 mapped. 



When Theobald published the first two volumes of his Monograph of the 

 Culicidae of the World, in 1901, he stated roughly that this insect, which at that 

 time was not known to him as the yellow-fever mosquito, ranged from 38° south 

 latitude to 38° north latitude, and his map upon page 292, vol. 1, indicated a 

 general distribution throughout eastern Australia, western New Guinea, all of 

 Celebes and farther India, southern Japan, eastern Hindostan, the Seychelles, 

 southeastern Africa, the African west coast, including Senegambia and Lagos, 

 Spain, southern Italy, the east coast of South America from British Guiana to 

 the La Plata Eiver, Panama, Belize, all of Cuba, Jamaica, Haiti, Bermuda, 

 and all of the southern United States. In vol. iii of his monograph, published 

 in 1903, he adds certain other localities. 



In No. 46, vol. xviii, of the public health reports of the Public Health and 

 Marine-Hospital Service, published November 13, 1903, and subsequently re- 

 vised to September 10, 1905, one of the writers (Howard) published a long list 

 of additional localities. All of these are covered by the general statement of the 

 distribution given above. All the exact localities observed by us will be found 

 under the heading of Aedes cdlopiis in the systematic part of this work. 



The foregoing consideration of the temperature conditions governing the 

 breeding of Aedes calopus explain why epidemics of yellow fever have occurred 

 on the Atlantic coast of North America even as far north as Montreal, and 

 may again occur, whereas no epidemic has occurred on the Pacific coast, nor 

 is it possible for one to occur. The summer temperature on the Atlantic coast 

 is for long periods at or above 80° both day and night, so that the mosquito, once 

 carried by ship or otherwise from regions of its permanent occurrence, may breed 

 in large numbers in our cities. It needs then only the introduction of cases of 

 yellow fever to start an epidemic. 



On the Pacific coast, on the other hand, the nights are so cold that the mos- 

 quito can not survive. It is as regularly imported into Pacific coast ports as into 

 Atlantic ones. We have records of specimens taken at San Diego and San Fran- 

 cisco. It breeds permanently in all the west coast Mexican seaports and must be 

 frequently brought to the Californian coast. Yet it has never been known to 

 breed there. This seems at first sight strange, since the mean annual tem- 

 perature of southern California is much above that of eastern cities where epi- 

 demics have occurred. In San Diego and Los Angeles one sees tropical vege- 



