178 WILLIAM BRODBECK HERMS 



8:50. The lamp was placed about one metre away, causing 

 an immediate retreat of the larvae toward the flesh. 



8:55. They were again collecting at this time as at 8.35. 



8:56. The lamp was moved to a distance of two metres. 



8 :57. The entire mass was moving toward the flesh and away 

 from the light. 



8:58. The larvae were again drawing nearer to the light. 



9 :00. Moving the lamp to a distance of three metres, resulted 

 as before. In this case it was necessary to place the lamp some- 

 what higher than previously, to which change the larvae responded. 



9 :04. The lamp was moved to a distance of three and one-half 

 metres. 



9:06. The larvae were retreating toward the flesh slowly and 

 in a rather scattered manner. 



9:10. The mass was very close to the flesh. 



9:11. The larvae were moving away from the flesh, butr in two 

 groups : one massed, the other rather scattered. 



9:12. The scattered larvae had assembled and the individuals 

 in the mass were moving toward them. This new aggregate actu- 

 ally proved to have collected at the point nearest the light. 



9:16. The lamp was moved to four metres. 



9:25. Out of about 200 larvae in the vessel all but about a 

 dozen had returned to the flesh on the side away from the light; 

 there are usually some stragglers under weak illumination. 



9:27. The lamp was placed once more at three metres dis- 

 tance. The response was very slow. 



9:38. The entire mass of larvae had a^ain assembled just at the 

 edge of the flesh. (The experiment was interrupted). 



Though only an ordinary kerosene lamp with a small (no. 1) 

 burner was used in these experiments and no intensities were 

 calculated the results are nevertheless interesting. It may be 

 said with assurance that the larvae when aggregated, react posi- 

 tively to lamplight; even at a very low intensity; and that the 

 experiments afford examples of positive phototaxis acting more 

 strongly than normal positive chemotaxis, provided the experi- 

 ments cited in my earlier paper (Herms, '07, pp. 78-79) are 



