June, '16] PARKER: DISPERSION OF MUSCA DOMESTICA LINNAEUS 351 



to a very small breeding ground. In this case the period for the flies 

 to become generally distributed is slightly longer according to the data, 

 but probably not as an actual fact. Distances of over a mile were 

 noted within a five-day period. 



The summer's results have suggested that, if it were possible to 

 liberate varying quantities of flies at stated intervals during a long 

 period of time and to have a large number of permanent recapture 

 stations from which flies were collected at daily intervals, results could 

 be secured of some importance from the standpoint of disease trans- 

 mission. The stations should be equally spaced and divided into 

 successive zones of 100 yards width. The data secured would concern 

 the following points: the average length of life of a fly, the number of 

 flies liberated, the number recaptured at each station, the number 

 recaptured in zones (each 100 yards more distant from the point of 

 liberation), the time required to reach given stations, and the time 

 to reach given zones. This data would permit averages to be figured 

 that would indicate the probable distribution of any given lot of flies 

 at any given time after liberation. Knowing then the length of life 

 of any disease organism on the outside of a fly or in its intestine, it 

 would be possible to establish a theoretical limit to which flies would 

 be likely to carry virulent organisms from any point of contamination. 

 Also, if at any given station the number of flies captured was in 

 excess of the theoretical number expected, it would imply that this 

 excess was explainable by unusual attractiveness or some other cause 

 and hence would indicate the character of places at which the proba- 

 bilities of contagion would be greatest. 



Conclusion 



The writer wishes to emphasize the fact that he by no means con- 

 siders the preceding discussion to be any more than an addition to 

 previous knowledge of fly dispersion. The suggestion that the stimuli 

 emanating from breeding and feeding areas are the factors mainly 

 responsible for seasonal dispersion under city conditions finds con- 

 siderable support in the data presented and is not at variance with 

 data previously published. It harmonizes better than any other 

 suggestion with conditions as studied in . the cities and towns of 

 Montana and finds some support in the known habits and life history 

 of the house-fly. 



In spite of the fact that the number of flies used was greatly in 

 excess of that of all previous experiments combined, it is felt that far 

 larger numbers could be used to great advantage. Much remains to 

 be learned concerning fly dispersion under various conditions and 

 whatever evidence can be found to support different view points, the 



