96 



Dr. J. F. Gemmill. 



[Jan. 17, 



supports, but also to some extent as active pushing agents. Should they be 

 defective or inactive or not straightened out over any part of the walking- 

 surface of the shell, that part will tend to lag behind at each lurch unless it 

 happens to be already right behind or right in front. The lagging of one side 

 will produce an apparent rotation of the shell towards thab side, and will also 

 cause the line of progression to be deflected in a corresponding direction. 

 Here the amount of deflection and rotation cannot possibly exceed two right 

 angles, and will usually be much less. 



The very marked relation between progression and rotation shown in the 

 tracks which fall under class (b) seems capable of satisfactory explanation on 

 the above lines. 



If one injures a group of spines on the walking surface, so much irritation 

 seems to be caused that purely passive or mechanical factors are superseded. 

 At least I found it quite unsafe to predict that a recently injured part would 

 lag behind. The handicapping effect of the injury seems to be overridden by the 

 screwing action of the lantern and spines, and this action in turn must depend 

 on nerve reflexes, the analysis of which falls outside the scope of this paper. 



Ratio of Botation to Progression. — To give an idea of amount of rotation 

 relative to progression, I adopted, in the case of each experiment, the 

 ratio obtained by dividing the total rotation in degrees by the amount of 

 progression in inches. The average value of this ratio in groups (a), (b) and 

 (c) is given below, and it will be noted that the table only includes instances 

 in which the amount of progression in 20 minutes was upwards of 4 inches. 

 From what has been said before, it will be evident that, if all instances be 

 taken into account, the E/P ratio may vary from nothing to 180 or more. 

 In a single lurch there may be as much as 20° of rotation. 



Table showing Direction of Progress and Eelation to Eotation in 40 instances 

 in which the Total Amount of Progression exceeded 4 inches (100 mm.) 

 in 20 minutes. 



f Without rotation, 2 instances, 

 (a) Progression along a line which is straight | With rotation, 12 instances ; average 

 or shows only small variations from side -{ E/P ratio, 16. 



to side ; 14 instances Dextral* rotation, 6 instances. 



(J) Progression along a line curving more or J Both dextral 1Q instances . 



less uniformly towards one side ; 23 m- j Both sinistral 8 instances . 



stances ; average E/P ratio, 11 Curvature of progression does not corre- 



| spond with direction of rotation, 

 5 instances. 



(c) Progression along a line showing one or more sudden changes in direction ; 

 3 instances ; average E/P ratio, 10£. 



* In the watch-hand direction as viewed aborally. 



