48 



swamp and were all found to be those of A. crucians. This species 

 was not found in any other spot for several miles round, though 

 with warmer weather both A. punctipennis and A. quadrimaculatus 

 were found throughout the territory examined up to within a few 

 feet of the swamp. It was therefore evident that these waters 

 possessed some peculiarity favourable to A. crucians but repellent 

 to the other two species. The chemical wastes emptying into 

 the upper end of the swamp were at once suspected, and the 

 results of subsequent investigations leave little doubt that they were 

 the determining factor. An analysis of this water showed it to be 

 acid and to have a high content of sulphates, sodium, potassium, 

 iron and aluminium. Fish were entirely absent, and experiments 

 showed the water to be poisonous to them. Since A. crucians 

 was propagating in enormous numbers it is evident that the 

 food-supply was suitable and at least sufficiently abundant. The 

 water contained a scanty microscopic fauna and flora, but a large 

 quantity of minute particles of organic material that was apparently 

 disintegrated plant-tissue was present. This was presumably derived 

 from the mass of dead leaves covering the bottom of the swamp, 

 its abundance probably being due to chemical action caused by the 

 substances emptied into the water from the factory. It would 

 appear from these deductions that A. crucians can subsist in nature 

 on a diet made up prunarily of non-living vegetable matter, and this 

 has been proved in the laboratory. It is not known whether the 

 chemical properties of the swamp waters possess an attraction for 

 the egg-laying females and the abundance of A. crucians is due to 

 this factor, or whether the females lay their eggs indiscriminately 

 and their development in the contaminated water is due to the latter 

 being especially suited to this species. Experiments showed that 

 larvae of A. punctipennis were able to live side by side with A. crucians 

 in the swamp water, while A. crucians could also develop in water 

 outside the swamp, development being in that case considerably 

 retarded ; only one adult was obtained from about 20 larvae placed 

 outside the swamp, and many larvae failed to pupate. It is con- 

 sidered probable that the determining factor in development in nature 

 is either the selection of favourable places for egg-laying or else a direct 

 effect of the water on the eggs, unfavourable conditions destroying 

 them before they hatch. More investigation on this point is desirable. 

 A. crucians, which is generally known to breed mainly in the autumn 

 and winter, was observed breeding prolifically in the swamp early 

 in April and continued to do so until late August, when the observa- 

 tions were discontinued. |jy 

 Experiments were made to determine the range of flight of A. cru- 

 cians from the swamp towards the neighbouring city of Montgomery. 

 The data obtained are given in a chart and tables. It was found 

 that A. crucians, when breeding in large numbers, becomes distri- 

 buted over an area within approximately 7,000 feet of the breeding- 

 place in numbers sufficient to be of sanitary importance. This is 

 in agreement with the conclusions previously reached by Le Prince. 

 From 7,000 to 9,000 feet, the danger would be questionable, probably 

 depending upon circumstances,5while 9,000 feet seems to be the limit 

 of the range of flight. This species showed very little tendency to 

 choose dwelling houses for shelter in the daytime. The bulk of the 



