518 ) Charles Zeleny 
error have been removed in the present experiments in the ways. 
indicated above. The individual differences that remain are 
partly at least due to inherent differences. ‘This source of error 
can be eliminated only by the use of a large number of individ- 
uals in each group. In the following data after throwing out 
all cases coming into corflict with the disturbing factors given 
above the number remaining for comparison is sufficient in some 
cases while in others it is unfortunately insufficient for conclusive 
results. 
Tne data for each experiment are treated separately. New 
observations are given in full and the results of former observa- 
tions in brief. ‘The data are presented in two divisions, the first 
including animals without and the second those with a molting 
habit. 
EXPERIMENTS ON ANIMALS WITHOUT A MOLTING HABIT 
Animals without a molting habit, as compared with those posses- 
sing this habit have a distinct advantage due to the absence of 
this complicated factor. They however have in most cases the 
disadvantage of less constancy in the level of the injury from 
which regeneration takes place. The level of the cut greatly 
influences the rate of regeneration and special attention was. 
therefore paid in all non-molting cases to its elimination as a 
source of error. 
I The Opercula of Hydroitdes dianthus® 
The Serpulid worm, Hydroides dianthus, has a large, so-called 
functional, operculum on one side of the head and a small rudi- 
mentary operculum on the other side. Regeneration of the small 
operculum when it alone is removed is compared with its regener- 
tion when the other or functional operculum is removed at the 
same time. In the first instance a new, small operculum gradually 
grows out from the cut surface. In the second instance the small 
operculum rapidly regenerates into a large one in case the func- 
*© Roux’s Archiv, 1902 and Journal Exp. Zodl., 1905, pp. 18-77. 
