101 



investigation begun in the previous year [see this RevieuK Ser. B, 

 iii, pp. 78, 79]. 



During the June survey on the Coosa River, Alabama, Anophehne 

 larvae were abundant in the running water of small streams and in 

 marshes and pools adjacent to them, but were practically absent in 

 the main pool, except under conditions which suggested that they had 

 been washed there from above. Of about 600 bred imagines, two 

 only were A. quadrimaculatus, the others being A. prmctip&miis. In 

 the latter part of May 1915 some female A. quadrimaculatus were 

 taken in a house at the power plant. A careful search was made for 

 their breeding places, but only A. pimctipennis were bred from the 

 larvae collected, and it is thought that possibly these individuals had 

 hibernated from the previous year. In August and September, even 

 more so than in June, the running water and side pools and marshes 

 adjacent to it were swarming with A. punctipennis, though the pond 

 water showed a preponderating proportion of A. quadrimaculatus. 

 A. crucians, in addition to these species, was found in holes made by 

 digging for angle-worms in one part of the bay at the mouth of a 

 creek. The breeding in the pond at this time was decidedly irregular 

 and apparently depended on factors affording protection to the 

 larvae from wave action and fish. The reason why they were so 

 rarely found in the backwaters of creeks is unknown. This zone, 

 except the upper part, was usually barren, although A. punctipennis 

 might be above in running water and A. quadrimaculatus below in 

 the bay. It seemed desirable to see if A. quadrimaculatus again 

 disappeared in advance of A. jnmctipennis in the autumn, and on 

 26th-30th October places which had earlier shown a preponderance 

 of A. quadrimacuhtus were examined, and 71 per cent. A. punctipennis, 

 19 per cent. A. quadrimacuhtus and 10 per cent. A. crucians were 

 found. 



The occurrence of a considerable number of A. quadrimaculatu.s 

 \nthin range of habitations might be a decided factor in producing 

 the malaria] fevers prevalent near this pool in 19 i 4. The first general 

 recommendation made to the power company owning the pool was 

 to keep the level of the pond as high as permissible during the season 

 when it is not producing A. quadrimaculatus, i.e., from about 

 1st October to 15th or 30th June, and to lower it for the three summer 

 months. This will kill the brush and land vegetation up to the high 

 level and land the floatage and much of the drift on the bank above 

 the low level, thus ensuring clean banks during the breeding season and 

 rendering them open to both wave action and fish. It was also 

 recommended to save the summer rises as much as possible so as to 

 create the greatest variation during this season, as this would land 

 the summer floatage and render the larvae more accessible to fish 

 by compelling them to move from their shelters. In certain places 

 the introduction was advised of mosquito- eating fish such as Fundulus 

 notatus and Gambusia qffinis, especially the former. 



In the June and July survey of a pool on the Black Warrior River, 

 Alabama, which was dammed in December 1914 for the improvement 

 of navigation and has steep banks for the first 20 miles above the 

 dam, there was but little accumulation of drift, though there was 

 a great deal of leaf- floatage in which Anopheline larvae were found. 

 Of the 100 larvae bred all were A. punctipennis except four individuals 



