168 DONALD WALTON DAVIS 
may easily be mistaken for an orange stripe (cf. fig. 3, 4). In 
the course of weeks this regenerating part increases both in 
absolute size and in proportion to the old tissue until finally it 
forms, in some cases, much the greater part of the polyp. Also 
the newer region becomes darker until it is no longer distinguish- 
able in color from the older. Meanwhile tentacles and a faint 
white bar are formed in the new region of the oral disc, and pale 
orange stripes appear in the new area of the column. A new 
white bar is at first fainter and narrower than an old one. This 
is illustrated at 3 in my figure 2, and better by Hargitt's colored 
figure ('14, fig. 2). The distinction persists for months at least. 
The new orange stripes, besides being paler, are closer together 
than the old ones. In most cases these features, especially the 
closer grouping of the stripes, enable one to distinguish a re- 
generated region for many months (figs. 36 to 41; also Daven- 
port, '03, figs. 8 to 11, 14), the exceptions being chiefly small 
and faintly colored individuals or those in which one or more 
regenerating regions constitute almost the whole of the specimen. 
Sections show that a new siphonoglyph is associated with the 
new white bar, and that in time new mesenteries are formed in 
the sector regenerated in the region of fusion of the torn edges. 
Even when the external distinctions between old and regener- 
ating tissue are absent, the mesenteries, as seen in stained 
sections, may give perfectly conclusive evidence that regeneration 
has occurred and even locate the precise position of the boundary 
between new and old portions (p. 169 ff). 
No obvious change occurs in the older regions in consequence 
of fission, except in the parts adjacent to the torn edge, which 
are subject to injury in the course of the division. No old 
incomplete mesenteries become complete, except perhaps in the 
rare cases where probably such a mesentery is itself injured and 
fuses with the esophagus or where the uninjured mesentery unites 
by chance with a torn edge of the esophagus. Similar unions 
between other parts sometimes occur, more often in artificially 
cut specimens than in normally divided ones. Abnormal ad- 
hesions of the latter sort commonly fail to persist, whereas the 
more nearly normal adhesion between a mesentery and the 
