2/6 C. M. CHILD. 



the diameter at the middle in some cases where it differed from 

 the oral diameter ; and finally the diameter about half way 

 between the aboral end and the middle, designated as aboral 

 diameter. At the same time diagrams were drawn showing any 

 pecularities or special features of the regeneration which might 

 occur. All the figures of this section were drawn from these 

 measurements and diagrams. In cases where the tentacles 

 differed markedly in length on different parts of the circumference 

 the two tentacles of the figure represent the extremes. 



No one can be more fully aware than myself that these meas- 

 urements are not and cannot be more than rough approxima- 

 tions. Direct measurement of the labial tentacles was often impos- 

 sible, but an estimate of their length was made in such cases by 

 placing the measure as near to them as possible and making 

 careful comparison. It is probable, on the other hand, that these 

 measurements are as exact as the instability of form in Ccri- 

 anthus will permit us to obtain. While I do not believe that 

 extended investigation along this line is profitable in this case in 

 consequence of the absence of stable form I have proceeded far 

 enough to convince myself that such measurements are unable 

 in a general way. Marked differences can certainly be detected. 



The results obtained from the several series of measurements 

 made agree in general though they showed some differences in 

 detail, and I think it is possible to draw certain conclusions from 

 them. The following series is selected as an example from 



among those made. 



Series j6. 



On October 20, 1902, disc and tentacles were removed from a 

 large specimen by a transverse cut, and the body was then cut 

 into four pieces, A, B, C, D (Fig. 3). These pieces were meas- 

 ured at intervals of two or three days during about six weeks. 



The following table is an abstract of the measurements. The 

 proportions of the whole animal' before section, the estimated 

 lengths of the pieces cut and the dimensions of these pieces on 

 five different dates during the six weeks are given in millimeters. 

 These are sufficient to show the general trend of the changes 

 and measurements made 'in the intervals between these dates 

 have therefore been omitted as unnecessary for the present pur- 

 pose. 



