ASEXUAL REPRODUCTION IN SAGARTIA 213 
For the sake of completeness, two specimens showing still 
wider variations from the typical conditions should be described. 
One of these on which ten orange stripes were counted shows 
mesenteries that would account for but eight. The other two 
stripes may have been situated in a broad expanse of the body 
wall between two pairs of complete mesenteries, in which are 
seen one pair of large incomplete mesenteries and an irregular 
lot of small ones, of doubtful history. They vary in number 
and in size at different levels. I have no explanation to offer 
for the presence of this group of mesenteries or of the stripes 
associated with them. The other specimen showed a single 
stripe in its older part and none in the newer region. Sections 
revealed five pairs of complete mesenteries and three complete 
mesenteries whose mates were not yet complete. A specimen 
with mesenteries present in such numbers and degrees of 
completeness would ordinarily show at least seven orange 
stripes. 
In the preceding paragraphs are described all of the specimens 
I have observed and sectioned which constitute exceptions to 
the general statement that the number of orange stripes in an 
undivided or fully regenerated specimen is twice the number of 
pairs of complete mesenteries. Of the nine such exceptions, 
three, showing one less than the typical number of stripes, are 
explained on the ground of division in an incomplete endocoel 
of second or lower grade resulting in elimination of one endocoel 
normally occupied by an orange stripe; three with an extra 
orange stripe show, each, an additional unexplained pair of in- 
complete mesenteries of apparently the first order in whose en- 
docoel the supernumerary stripe probably lay; while three show 
wholly unexplained deviations from the normal relations of stripes 
and mesenteries. Over against these exceptions must be urged 
the significance of forty-nine specimens on which the number of 
orange stripes as determined before killing was exactly double 
the number of pairs of complete mesenteries as ascertained by 
examination of sections. Undoubtedly in these cases the orange 
stripes occupied the complete endocoels and incomplete en- 
docoels of the first order only. These forty-nine specimens show 
