212 DONALD WALTON DAVIS 
no. 54, table 6, showed ten orange stripes and (internally) six 
pairs of complete mesenteries. On one side the boundary 
between old and new evidently lies in an incomplete endocoel of 
the second or lower grade. This accounts for the lack of one 
orange stripe between two adjacent pairs of complete mesenteries. 
On the other side an undisturbed old incomplete endocoel of 
apparently the first order is present, and an orange stripe would 
be expected there. In the absence of an orange stripe in this 
region, one is driven to consider the possibility of this pair's 
belonging really to the second order of incomplete mesenteries 
rather than to the first cycle of which it is apparently a member. 
In that case four cycles of incomplete mesenteries must have been 
present. If this is so, division must have occurred in an exocoel. 
Both the presence of four cycles of incomplete mesenteries and 
divisions in exocoels are uncommon occurrences, and this com- 
bination is consequently improbable and the explanation un- 
satisfactory. 
In the specimens just described, orange stripes were not found 
on the boundary between old and new regions when this lies 
between members of the second cycle of incomplete mesen- 
teries. If this always held true and the incomplete mesenteries 
were invariably regularly placed, we should never find orange 
stripes associated with two pairs of incomplete mesenteries not 
separated by a pair of complete mesenteries. It has been 
shown previously (p. 196.) that irregularities in the first cycle of 
incomplete mesenteries are occasionally found. Thus two pairs 
of incomplete mesenteries of the largest size with no complete 
mesenteries separating them were found in three specimens on 
which counts of orange stripes gave odd numbers. Regeneration 
was apparently complete in these cases, and the number of 
stripes is such as to lead one to suppose that in each case one 
occupies the additional incomplete endocoel. In these instances 
a reduction from the typical number of orange stripes, which is 
double the number of pairs of complete mesenteries, or an 
excess over the typical number, corresponds with a departure 
from the normal number of incomplete mesenteries of the first 
order. 
