280 
BULLETIN OF THE UNITED STATES FISH COMMISSION. 
soft strips between them, and partly slip over one another, so that the size of the 
head-sheath is greatly reduced and the anterior opening nearly or quite closed. It 
is possible to withdraw the head, at least partly, in most species without causing the 
complete folding of the head-sheatli; evidently a supplementary contraction of the 
fine transverse muscle fibers is necessary to bring this about. 
In some species (notably Diurella tigris Muller, pi. I, figs. 3, 4; Diurella rous- 
seleti Voigt, pi. iv, fig. 37; Rattulus multicrinis Kellicott, pi. vi, fig. 58; Rattulus 
capucinus Wierz. & Zach., fig. 59, and Rattulus cylindricus Imhof, pi. vn, fig. 62) the 
head-sheath falls when contracted into very regular folds. In D. tigris Muller, 
D. rousseleti Voigt, and D. intermedia Stenroos, and perhaps in other species, the 
number of these folds is nine. In some other species, as, for example, in Rattulus 
gracilis Tessin, pi. v, fig. 48, the folds are very irregular. In still other species no 
such folds are present, and the lorica may remain widely open when the head is 
retracted. This is the case, for example, in Rattulus scipio Gosse, pi. v, fig. 52. 
On the anterior dorsal margin of the head-sheath there are in certain species 
of the Rattulidce a number of teeth. In Diurella i rousseleti Voigt there are nine 
well-marked teeth; in other species there are but one or two. Leaving out of con- 
sideration for the present the case of Diurella rousseleti , we may classify the teeth 
in other species into two categories: 
(a) In Rattulus multicrinis Kellicott (pi. vi, figs. 55 and 58), Rattulus capucinus 
Wierz. & Zach. (pi. vi, figs. 59-61), and Rattulus cylindricus Imhof (pi. VII, fig. 62), 
there is a single nearly median projection of the dorsal lorica edge, extending over 
the head. In Rattulus cylindricus Imhof this is prolonged into a long hook, curved 
downward over the anterior opening of the lorica. In these cases the tooth seems 
to be nearly or quite in the middle line. 
(b) In a number of other species there is either one tooth (Rattulus gracilis Tes- 
sin, figs. 45-48; Rattulus scipio Gosse, figs. 50-52; Diurella tigris Muller, figs. 1, 3, 4; 
Diurella tenuior Gosse, figs. 7, 8; Diurella weberi, figs. 12-14 and 116-117; Diu- 
rella intermedia Stenroos, figs. 108, 109) — or two teeth ( Rattulus longiseta Schrank, 
figs. 67-70; Diurella ins ignis Herrick, figs. 15, 16; Diurella porcellus Gosse, 
figs. 19, 20; Diurella stylata Eyferth, figs. 27—30), which seem of a different char- 
acter. These lie distinctly to the right of the dorsal middle line (so far as that can 
lie defined), and form prolongations of one or both edges of the “striated area” of 
the lorica, hereafter described. When there are two of these teeth they are usually 
unequal in size, the right one being longer. (Only in Diurella stylata Eyferth are 
they nearly or quite equal in length.) In most species they are merely short teeth, 
but in Rattulus longiseta Schrank and Diurella stylata Eyferth they are long spines. 
The position of these teeth on the right side is one of the markedly unsyni metrical 
characters of the Rattulidce. A further account of these teeth may best be deferred 
until the “striated area” has been described. 
Many of the species have no teeth at the anterior edge of the lorica. The ante- 
rior opening of the lorica is usually oval, with a slight notch near the ventral middle 
line. In some few cases the edge of the head-sheath projects farther on the left side 
of the opening than on the right. This is notably the case in Diurella weberi n. sp. 
(pi. xiii, figs. 116-117); it- is slightly so in Diurella tenuior Gosse and Diurella 
brachyura Gosse, and perhaps in other species. 
In some cases three or four or more teeth have been described by different 
authors at the anterior edge of the lorica. In many cases this is due to the optical 
