110 PAUL S. WELCH. 



Extrusion of Excrement. The voiding of excrement occurs in 

 connection with the respiratory trips to the surface but since 

 these trips are the direct effect of the demand for oxygen the 

 extrusion of wastes is only an accompanying feature. The ratio 

 of the number of respiratory trips to the number of extrusions is 

 variable, depending upon the feeding activity of the larva. 

 The following records will serve to illustrate the variation: 



Larva no. i 2 extrusions in 14 trips to surface. 



Larva no. 2 i extrusion in 1 1 trips to surface. 



Larva no. 7 i extrusion in 41 trips to surface. 



LOCOMOTION. 



Dissemination is accomplished in three distinct ways, viz., 

 swimming, floating, and crawling. 



Swimming. The larva usually swims on the surface although 

 when forced to do so it can swim to a limited extent below the 

 surface. The specific gravity of the larva in all stages is very 

 slightly less than that of water, thus making it easy to remain at 

 the surface. Furthermore, the larva at all stages has an oily 

 surface which prevents a "wetting" of the exterior and constitutes 

 another means of remaining on the surface. This form of loco- 

 motion is accomplished by means of a series of rather vigorous 

 horizontal undulations in the execution of which the whole body 

 takes a part. The caterpillar lies in a trough in the surface 

 which partially surrounds it and is propelled by the characteristic 

 sinuous movements. The efficiency of this kind of locomotion 

 varies with the age of the larva and the amount of disturbance 

 of the surface of the water. Very young larva? do not progress 

 very rapidly but the full-grown ones are efficient swimmers, 

 showing noteworthy speed and endurance. Larvae, 4 cm. in 

 length, were removed from their burrows and placed in the open 

 water in a protected bay where the surface was perfectly quiet. 

 They immediately began to swim and records were made of the 

 rate of progression. In one instance a larva made a continuous 

 swim of 250 feet in 19 minutes and at the end of the journey 

 showed little or no signs of weakness. A number of similar 

 records were made with similar results. Ripples, diminutive 

 waves, and currents interfere seriously with this form of loco- 



