322 THE DOLABELLINAE. 
merge into the zone 6, in which the homogeneous secretion increases in amount’ 
and staining quality, appearing almost black with hematoxylin at the point 
of discharge, a. In other sections this secretion, evidently of a mucous nature, 
is seen issuing between the epithelium-cells, and is especially abundant in the 
deeply infolded groove shown in Plate 9, fig. 7, near the free margin of the flap. 
Here the secretion accumulates in considerable amount, and issues between the 
thick cuticle of the spine-bearing area and the epithelium just beyond it. 
Other very characteristic gland-cells are scattered just below the epithe- 
lium of the dorsal surface of the fold. Many of these show no connection 
with the surface, while others are united by more or less slender neck-like pro- 
longations. A group of these cells is shown, Plate 8, fig. 6. Deeply staining 
nodal granulations of varying size and shape are abundant in them. Unfortu- 
nately the fixation of the limited amount of material at hand does not permit 
any trustworthy conclusions to be drawn concerning them. 
The epithelium of the palatal folds rests upon a narrow layer of dense con- 
nective-tissue, below which are found bundles of smooth muscle-fibres, running at 
various angles, and surrounded by a large amount of fibrous connective-tissue. 
The central area of the fold is occupied by looser connective-tissue, in which seat- 
tered bundles of longitudinal muscle-fibres are located. At the base of the sec- — 
tion, Plate 9, fig. 7, is shown the duct of the salivary gland in cross-section at d. 
In Dolabella hasseltti (Fér.) similar relations of the palatal spines to the 
epithelium and the cuticle were made out, with some differences worthy of note. 
In this form the spines are slender, thick-walled, tapering tubes, the thickness 
of the walls increasing and decreasing at regular intervals in coincidence with 
a system of parallel ridges and depressions of the outer surface. Each spine 
is the product of a single matrix-cell, the central free end of which is prolonged 
into the tubular base of the spine, while the wall itself is secreted by the zone 
of the cell-surface around this projection. Thus the spine fits down over the 
prolonged tip of the cell in the form of a conical cap, the successive layers of 
chitinous substance being in the main secreted by the sides of the distal end. 
The central portion, however, takes part in this secretion at intervals, the 
layers then formed appearing as several curved dome-like sheets, bridging across 
the entire lumen of the spine and continuing downward as part of its sides. 
No indication of the participation of any of the surrounding cells in this seere- 
tion could be found, nor is anything here present similar to the small central 
cell which determines the cavity in the spicule-development in the Solenogastres, 
as described by Heath (1911, p. 28). 
