SPONTANEOUS MOVEMENTS IN I.ARV.l.. 413 



(41 Stimulation of movements which might result from the 

 shaking the bottle containing the specimens, after the 

 addition of the acid to the sea water. 



To control these sources of error, the following routine pro- 

 cedure wa- adopted. The larvae were taken from the stock in 

 the aquarium. -< raped off from their attachment, and placed in 

 a j.so cc. lla-k <.f fresh sea water at 25 C., and allowed to remain 

 undisturbed from one half hour to an hour. The bottle con- 

 taining the larva? was then shaken a definite number of times 

 (jo uniform back-and-forth movements). This procedure -ei -\ed 

 control for the shaking after addition of the acid; se\eral 

 ' -rimem- - emed to show that stirring of the sea water in 

 tlii- manner had no appreciable effect on the movements of the 

 lar\ a . but nevertheless, at the beginning of each record of normal 

 movement- the bottle was shaken as described. The movement ? 

 of tin- larva- were then recorded for a period varying from 15 

 minutes to one half hour; this furnished the record of the 

 mo\ements under normal conditions of each individual at 25 t 

 At the end of this time, a little sea water was poured out of the 

 bottle, the proper amount of HCL w/2O added and the content- 

 o! tin- bottle made up to exactly 250 cc. with sea water and the 

 .-topper inserted. These operations were carried out as rapidly 

 a- p. i--il.li in order to prevent the escape of CO; into the air. 

 The bottle was then stirred (as described above for the normal 

 iol in order to secure as rapidly as possible a uniform < < ' 

 nd 1 1 ion concentration throughout the water in the 

 bottle. The movements were then recorded, usually for 30 

 minim- to .ne hour. At the close of this period, the acidulated 

 uater ua- poured off, and replaced with fresh sea water of 

 the -a me temperature. The slight- agitation of the larva? whieh 

 \\a- unaxoidable in pouring off the acidulated water is probably 

 a -ui'ticieiit i ontrol for the stirring at the beginning of the previous 

 -tep- in tin- experiment. A record of the movements \\as 

 then made. 



In making the record, the time, character and extent of the 

 movement- \\ere noted. A single coil to one side was taken as 

 the unit. The number of separate movements can be recorded 

 in this way with considerable accuracy; but in the case of 

 movement- following each other in rapid succession, as in the 



