414 HENRY C. TRACY. 



"fluttering" movements mentioned above, a somewhat arbitrary 

 method is necessary. These "flutters" were assigned a value of 

 10, which seems about the minimum estimate of their value. 

 At the free swimming stage these "flutters" sometimes continue 

 several seconds which, of course, cause an extended swimming 

 movement; a value of 10 for each second Avas given for such 

 movements. The results thus arrived at are partly based on 

 estimates and hence cannot be exact. But they will serve to 

 give the general order of magnitude of the movements of the 

 larvae under different conditions. Most of these movements are 

 undoubtedly greater than the minimum assigned value; hence 

 the error favors a negative result. Since the method of estima- 

 tion of numerical value was the same for the same larva under 

 both control and experimental conditions, errors due to the 

 method of recording tend to neutralize each other. 



The records of each experiment were afterward transferred to 

 chart records by recording the total numerical value of the 

 movements for each minute on cross-section paper. The records 

 of all specimens for each concentration of CO2 were then averaged 

 for each minute; then, in order to smooth the curves somewhat, 

 averages for each two minutes were taken and recorded as 

 shown on the charts accompanying this paper. 



Observations were made at two periods of the larval develop- 

 ment: 



(1) Within the first five or six days after hatching and before 



external respiration had begun (larva? about 6 or 7 mm. 

 in length). At this stage there is no response to light, 

 rotation, jar, or vibrations and little or none to tactile 

 stimuli. 



(2) At the end of the larval period when the yolk sac is nearly 



absorbed, and the larvae are on the point of swimming 

 free (18-20 mm. in length). All the reaction mechan- 

 isms, as far as known, are now functionally developed. 

 The behavior of the respiratory mechanism in the free swim- 

 ming larva? (second period) was observed under the different 

 conditions of CO 2 and the record superimposed on the charts 

 recording the body movements. The time in seconds necessary 

 for the completion of 10 respiratory movements was taken with 

 the stop watch for each individual (whenever possible the 



