222 MOSQUITOES OF NORTH AMERICA 
inches deep, shaded in the morning. Duckweed abounds here, and a quantity of 
green alge cover the sticks and stone of the bottom; the pool swarms with 
Culex and Anopheles punctipennis larve; (d) Thirty-five feet above (c) is an 
‘iron’ pool with a deposit of iron; it is about 9 feet by 9 feet, 10 inches deep, 
partly shaded by bushes and a huge log. It contains no vegetation and no larve. 
Of these four pools the smallest one, in the center of the row, is the only one con- 
taining Anopheles larve. These larve are exclusively those of A. punctipennis, 
while in the river only a few feet from here, A. maculipennis is abundant. 
“ Another noticeable feature of the distribution is the entire absence of all 
mosquito larvee from the bayou, although the water is currentless, the vegetation 
abundant, and the light and shade favorable. It is probable that the duckweed 
and lily pads cover the surface too closely in the shallow water near the shore. 
As pool (c) above described became choked with duckweed later in the season, 
the larve disappeared. 
“The following table shows the undoubted selection of different breeding- 
places by A. maculipennis and A. punctipennis: 
ANOPHELES BRED FROM LAaRv AND Pups, 
A. maculipennis. A. punctipennis. 
Males. Females. Males. Females. 
Collected from spring-fed pools........-. 0 0 30 35 
Collected from river......... ccc cece eee 46 47 1 4 
“ The fact that the breeding-places are only a few feet apart renders especially 
remarkable the specific preference displayed. The predominance of A. puncti- 
pennis in the other two regions that were examined is in accord with the obser- 
vations in Michigan. A. punctipennis shows a predilection for spring-fed pools, 
and, in localities where these abound, is the chief species. Hirschberg and 
Dohme observed a rather definite geographical distribution of the two species 
in the vicinity of Baltimore, A. maculipennis being found on low ground in or 
near large bodies of water, while A. punctipennis was found in small, clear 
streams or springs on higher ground.” 
Smith in New Jersey says of A. punctipennis, “ On the whole it breeds most 
abundantly in clean water along the edges of ponds or swamp areas or in the 
eddies of shallow streams.” Concerning A. quadrimaculatus, he says, “The 
breeding places are similar, but this form also occurs in brackish water on the 
salt marshes, hence has a somewhat wider range and adds the positive danger 
of disease to the disadvantages of an undrained marsh.” 
J. K. Thibault, Jr. (in litt.) states that at Little Rock, Arkansas, Anopheles 
punctipennis is the most abundant species in the city and that it is scarce in the 
country; but that A. quadrimaculatus is the more abundant mosquito in the 
river-bottoms but is scarce in the city. 
An interesting point has been brought out by Dutton in his report of the 
malaria expedition to the Gambia (Liverpool School of Tropical Medicine— 
Memoir X), to the effect that the number of mosquito breeding-places present 
in compounds was found to vary with the social position of the occupier, and in 
this statement he includes Anopheles mosquitoes. The following is taken from 
his report: 
